How Julian Shapiro’s Bell Curve does marketing for some of the best tech companies in San Francisco

Andrew: Hey, there, freedom fighters. My name is Andrew Warner. I’m the founder of Mixergy where I interview entrepreneurs about how they built their businesses, and I do it for an audience of real entrepreneurs. And my friend Courtland Allen was telling me that he doesn’t do poker at his house.

I said, “All right, let’s do poker together.” So the two of us did poker at his house, and he had a few friends over who work at Stripe like he does, he sold his company to Stripe.

He had a few friends who are in the Y Combinator world there. And they were talking about this one guy that Y Combinator brings in to do marketing. And I said, “Kind of interesting, maybe I should interview him at one point.” And then Courtland said that for his podcast, Indie Hackers, the number one guest, I think by far the most popular guest on the podcast was this same guy.

I said, “Maybe, I should stop filing this stuff away. So what’s his name?” He said, “Julian Shapiro.” I said, “Interesting, marketing for Y Combinator and the most popular guest. Could you intro me?” He said, “Yeah.” I said, “Okay, but I will tweet something out publicly to make sure that we don’t forget about this,” and I did. And that’s how we ended up getting today’s guest on here to do this interview.

Julian Shapiro is the founder of Bell Curve. They’re kind of like the Chief Marketing Officer for many tech companies. They kind of outsource Bell Curve to do the stuff that a CMO, a Chief Marketing Officer might do for them. And he’s got a lot of fans here in the chat, which I can see.

I invited him here to talk about how he does marketing for some of the best tech companies here in San Francisco, in Silicon Valley, and to learn from him also, frankly, if I could, Julian, what I could use from his techniques and what’s worked for him at my company.

And this interview is sponsored by two phenomenal companies. The first is called Ahrefs. If you’re doing any online content creation, especially if it’s based on and you want to get traffic based on search engine optimization, you got to check out Ahrefs. I’ll talk about them.

And the second is a company you’ve heard me talk about a lot. They are the best place to go hire developers. It’s called Toptal.

Julian, good to have you here.

Julian: Hey, it’s so nice to be here.

Andrew: Julian, who are some of the companies that you work with?

Julian: We’ve worked with Perfect Keto, Imperfect Produce, Envoy, Framer, now Microsoft. We run their LinkedIn ads for a few products. And a lot of YC companies in general.

Andrew: So wait, what do you do? You work for Microsoft too or for a company that works with Microsoft?

Julian: Yeah. We now run Microsoft’s own LinkedIn ads for their own products.

Andrew: Really?

Julian: Yeah, super fun.

Andrew: How many people at your company?

Julian: So we’re just nine people.

Andrew: Wow. Why am I not seeing your video, dude? Everyone else here, today we’re doing live series of entrepreneurs coming online with me with a live audience. You’re the only one who’s not coming on video. Why?

Julian: That is a fair question. Short answer is I have no clothing, and I know I’m going to regret trying to tweak my webcam to be just on the upper part of my face. I’m like, “I going to regret this if I look at that in a year from now.” So, unfortunately I’m going to be a black screen for now.

Andrew: Like, you’re totally nude?

Julian: Well, no, I do have clothing but no upper body clothing.

Andrew: Because?

Julian: Got perfume on it and I’m super, super allergic to perfume, makes me feel like my skin’s burning. So I threw out almost all my clothing I had ready to wear this morning.

Andrew: Wow, that is such a . . . It’s a like San Francisco thing to do, to not actually say, “I’m going to go and find a way to get all this perfume off,” but just say, “I’ll toss it out. I’ll figure it out later.”

Julian: Yes. So hard to get it out in the wash.

Andrew: Wow-wee. All right. I want to get an example of what you do, and one of the things you told me to do before we started is, you reverse engineer. Reverse engineer, what?

Julian: The patterns that work for types of companies. So like the first thing I’ll do, when we work with a company, is I’ll try to figure out who are they most similar to. So what is the business model? Is it B2B, B2C? Who are they selling to? What are they selling? What’s their price point? And is it subscription or one time, or services? And once I bucket you into something, then I try to figure out, “Okay, well, everyone like you, how did they succeed?” And then I pull those learnings from the Rolodex I’ve been accruing and then try to use that to get growth working for you pretty quickly.

Andrew: I told you that I think better when I have concrete examples, and I asked you if you could come up with one company that you did this for. What’s the company we can go over here today?

Julian: For sure. So they’re called Tovala, T-O-V-A-L-A.com. They’re a food subscription, so they deliver food to your door. And it comes paired with an oven that cooks the food for you automatically.

Andrew: So I buy the oven and then I pay monthly to get the food delivered to put in the oven and the oven is like a smart oven that will cook it up. Am I right?

Julian: That’s exactly right, yeah.

Andrew: Hey, I’m on there. I’m not like guessing this or remembering it. I’m on tovala.com, so I can see. And in fact, there’s like a barcode on the food, so if the oven scans that barcode, it knows exactly how to cook my food properly. That’s what we’re looking at

They came to you first and then you said, “Before I start working on them, I want to reverse engineer their best competitors and see what’s working and what’s not”?

Julian: That’s exactly right. So it’s twofold. One is I’m not afraid to do some competitive analysis. And the other thing is I’ll go ask people who specialize in this space. So I actually just put my ego aside and I spoke to other agencies that exclusively work on e-commerce conversion rate optimization.

So they make it so if you’re an e-commerce business, you’re selling physical goods, they will figure out what part of your website is the bottleneck. And so I just asked the guy who’s the best at this and said, “Give me your list of 12 things I’m supposed to be doing.”

And then I sanity check that against, “Have I seen these actually work for other people I’ve worked with?” So one, competitive analysis. Two, talk to people better than you. And three, look at the data you’ve already accrued yourself.

Andrew: So who did you do that with? Let’s break it down.

Julian: Yeah, for sure. So Dev Patel, he runs an agency called GrowthRocks, and he is who I got this initial idea of what are the no-brainer commonalities among e-commerce businesses. And then, in terms of competitive analysis, fortunately there are some pretty direct huge competitors to Tovala. You have Blue Apron, right, you have Purple Carrot. I’m sure you guys have seen those ads on Facebook. And among my own clients, this is pure coincidence, we’ve just happened to have worked with enough e-commerce businesses for me to find those patterns there as well.

Andrew: So what did you find? When you were looking at the patterns, for example, for Blue Apron, what’s relevant about them?

Julian: Yeah, exactly. So the first thing to realize is it’s hard to actually know what is worth replicating. When you’re looking at a competitor’s website, you don’t know which of those things is working, which of those things is actually hurting them, right? And so looking at a snapshot can be very dangerous and misleading. You might actually be bucketed into an A/B test they’re running or the A/B test you landed on is a variant that’s then just screwing around. It might be the worst variant possible. So if that’s what you’re cloning, you might be in for some surprises.

Andrew: And also, would I even want to clone in their homepage? If I’m looking at marketing, wouldn’t I say, “All right, Blue Apron, what seems to be working for them?” I’ll go to SimilarWeb, see where the traffic’s coming from and then see if I can run some small tests that way. But that’s not even accurate enough, right?

Julian: Exactly.

Andrew: So what did you do instead?

Julian: Exactly, so you want to take one step back. First, you never want to work off a snapshot. You want to work off a trend or a pattern. So let me give you an example. If I go to archive.org, I can look at historically how has their website changed over time. So it’ll give me these snapshots of a month ago, two months ago, and I can see what are the constituent parts of their site that are rock-solid. They know they work. They’re not yanking them off. And then conversely, what is currently being tested? What’s the new thing?

And I’m going to first start with what’s rock-solid, what’s been there for half a year, right, because it’s not going to still be there if it’s doing poorly.

And the second thing I’ll do, as opposed to using Blue Apron, for example, as my snapshot, I’m also going to look at a bunch of their competitors and find the commonalities between all those pages.

Andrew: So let’s get concrete. When you did that, what did you see that was concrete and what did you see that was similar within their competitors too?

Julian: Yeah, absolutely. So in no particular order of importance, let’s just start with aggressively A/B testing pricing. So one thing you learn quickly when you do grow for e-commerce businesses is pricing is perhaps the highest leverage thing to change.

So for example, if I’m charging $50 a month for my subscription package, bringing that down to $40 a month will oftentimes have more of a positive effect on conversion rates than just about anything else you could possibly do on your website.

So they realize this. I know they realize this because I can see going back through the archives how aggressively they’re trying different pricing display formats, like, “Oh, no, you know what? We’re going to give you three months.” This will reduce the price point or, “We’re going to try six months but throw this special deal in.”

Andrew: Right. And so you saw that with Blue Apron. They were doing price tests that recently?

Julian: I saw this not only for a Blue Apron and their competitors, but almost all e-commerce businesses. So right now I’m talking commonalities across e-commerce, but we’ll go into some specifics regarding food as well. But what’s interesting is you don’t necessarily have to change your price to actually take advantage of this fact.

What you could do instead is if you look at Blue Apron’s squeeze pages, which your audience maybe familiar with, so like a special web pages that you send people to after they see your ads, and on these special squeeze pages, they might be pitching a deal, a limited deal, a special discount, or they might display the price in a different way where they do or do not exclude certain extra bonuses.

Andrew: So one example that I see with Purple Carrot, for example, they’re not just saying that it’s $72 a week. They’re saying it’s $12 per plate, and yes, it’s $72 a week, and that’s what you’re talking about, not necessarily changing the price, but changing the way they’re displaying it. Am I right?

Julian: That’s exactly right. And if you look at Tovala, if you go to the website right now, let’s take a look because they’re changing all the time because they should be, because they know how important it is to A/B test this. And so right now, they’re framing it as $250 upfront for the oven and $12 a meal.

Now, here’s how they used to do it. So this is the process I went through with them when we started. We first figured out what are all the possible ways we could display pricing, and it turned out they happened to do a huge upfront price, like you’re going to pay $300 for the year, all your meals are going to be covered. But it turns out, when you use a lower price point and anchor it on a per meal pricing, then you can compare it more easily to all of the competitors and it’s more obvious how good of a deal you’re offering, right?

Andrew: Okay, I’m with you.

Julian: So framing is very important. So that’s one thing. Let’s move on to a different one. Let’s take a look here. Free shipping. So this is a small one, so don’t spend too much time on it, but it just turns out, when we do a free shipping test across most e-commerce businesses, that is one of the top five little things you can do that has a huge impact, very high leverage. You just add that notice. Now whether or not you can actually afford it is another question as the business, but it does meaningfully move conversion, sometimes by up to 25%.

Andrew: Okay.

Julian: So that’s just like table stakes, like get that out of the way. Increase your base price if that’s what you have to do to just offer free shipping. People are very spoiled due to Amazon Prime.

Another very, this was a surprising one, you actually want to A/B test your product images. So people overlook this. When they A/B test things, they’ll A/B test the copy on their web page, meaning the words, the structure, the order, all that stuff. What they won’t actually do is go back to the photographer and say, “Hey, can you put this on a different background?” Or, “Can you get a different angle of the product?” Or, “Can you use a model in the photography of a different gender or ethnicity?”

Let me put it this way. If you find the right gender, ethnicity, background, product angle versus the worst combination of all of those criteria, the best and worst versions compared to each other can up to 50% improve conversion rates. And all you got to do is go back to the photographer and just spend a few more hours teasing out different ways to visualize this. And people will overlook that. They usually shoot it one time, set it and forget it.

Andrew: But is it about . . . I’m surprised actually about ethnicity, but, okay, I get that. What I would imagine is . . . I’m looking at Tovala’s website. One of the product shots is someone cutting into the food, and it looks delicious and it’s paired with wine. I get that that would work. But it feels like there’s another one where you’re taking the plastic wrap off of the metal container, the aluminum container that the food comes in, and I would think that that would be something that I wouldn’t want to show, unless people told me that they were not buying because they’re not sure how this thing comes and how they put it in the oven.

And it feels like you want to do more than just test different ethnicities and different things that you think would work like background and understand people’s questions, problems and confusion, and then go get photos taken. Am I right?

Julian: Exactly, that’s step one.

Andrew: And then how do you do that? How do you know what they care about? How do you know that too many people are confused by how this will work and think that it’s just going to be some kind of powder that they put in their microwave? I don’t know.

Julian: Yeah. So now we’re getting into the realm of basically value prop identification, so what are the things you should be talking about in general. And we’re getting to the realm of landing page, copywriting, and structure. So what should you be showing them and in what order? So in the case of selling anything complicated, and that’s one thing about Tovala compared to competitors, is not self-evident how Tovala works nor why.

With Blue Apron you get some packaged meal, ingredients. You make it yourself, done. With Tovala, they first send you an oven. Then later on, you get sent these specially par-cooked packages that you’ve never seen anything like before. And you put them in the machine and it cooks them in a special way.

So the primary challenge of Tovala is trying to figure out how do you balance on a landing page, how much do you explain what it is you do because it’s not self-evident, and how much do you just actually put your product on display and make it look awesome?

So in the case of the Tovala, if you look at this, I mean, the date, I’m sure this landing page will change a lot because they have a really good rigorous A/B testing cadence. But at this point in time, when I’m looking at the top half of the web page, there’s a row with four images, and they’re showing you get a delivery in a box, you open the package, you put it in the machine and you eat your food. That’s really powerful because if I just showed four images of the food that you could eat, well, it’s kind of like self-evident that, yeah, I’m going to get a variety of delicious foods. That’s what all of these services have in common. How does Tovala differentiate? And in their case you get something very special in the first two images.

First, you realize you’re getting something delivered. You’re getting food delivered. Okay, it’s in these special boxes, but then they show you the actual inside and you start realizing, “This is a special kit that fits into an oven that knows how to cook them uniquely.” It’s scanning the barcode on the box. It knows what it is. And that is the third image where it crystallizes. You’re like, “Wait, what’s going on here?”

And so you start asking questions like, “How is it cooking these meals automatically?” And that’s the narrative that kicks off the page.

Andrew: Okay, and so how do you know that this is what people need?

Julian: Right.

Andrew: Is that part of what you do? As a CMO for these companies that hire you, do you try to understand the frustration, the needs, the desires of the customer so that you can create this?

Julian: Two ways, the first thing you do is you survey. So if you have enough existing users, you don’t have to go off hunches. You send broad surveys and you ask things like, “What are the things that you love most about Tovala? What are the things that were most unclear to you when you first heard about Tovala?” And lastly, “What were the things that almost stopped you from buying altogether?” And then you can take those answers and proactively address those skepticisms on the home page and you can figure out what to say based on that.

And the second way . . .

Andrew: Julian, do you actually do that? You actually will send out surveys and get responses like that? I find that people don’t respond to surveys like that. It’s not worth their time.

Julian: We do actually, and the trick is it’s not cheap because we offer Amazon gift cards for responses. And so maybe we’ll plan a budget of like $500 to $2,000, and we’ll give up, maybe $25 gift cards or $20 gift cards. And you might think, “Well, people will just answer with bogus answers to get the gift card.” In fact, the answers are incredible, super valuable, and we don’t have to give gift cards to people who were just winging it. So we get phenomenal responses.

So if you can put aside the budget for that, it is budget well spent. And here’s what, let’s put that expense in contrast. So let’s say I spent $500 up front. I get proper data. Now I use that data to inform my landing page. Let’s compare that to the alternate that everyone else follows, where they just wing a new landing page based on a few hunches, they then blow over $500 on ad traffic sent to it trying to figure out where their hunch is correct. And it’s not sensible to be too conservative with money in the survey stage.

Andrew: Okay. What’s the other way then? You said survey is one. What’s the other way that you would know this stuff?

Julian: So the other way is the least romantic, but probably the best answer, which is you just A/B test it. You come up with 10 ideas, you A/B test them all, and you just find out what the data actually tells you performs best.

So you’ll have a few hunches like what should I be talking about on the landing page, in what order should I talk about those things. But to go in assuming you know the right way to do it is usually a recipe for failure. And so let the data tell you.

Andrew: Okay. So you are looking at all these sites, you’re starting to reverse engineer. You have a list of things that you’ve learned about e-commerce sites, which is pricing always has more importance than people think, and photos are more important than most people think. What else? What else did you do before going to work with Tovala?

Julian: Yeah. So this was a surprise learning that I wanted to test on their behalf, and they let me indulge. And then I learned this and then applied this to all of our other clients. And that is it turns out a very long landing page counterintuitively outperforms a short one. So it’s drilled into our heads, right? You have to be concise, you have to get right to the point and get out of the way. That’s not really true.

Concise doesn’t mean being as short as possible. It means conveying as much information as it can in the space you have. And so your goal is to just say all that needs to be said to get them to a point where all their objections are gone. And they have enough of an ingrained motivation now that when you send them to the checkout page, they’re actually going to want to buy because you prepared them for it. That is the key differentiator between how we do landing pages today and how we did them worse last year.

Andrew: Okay. I’m looking, by the way. I’m trying to do this for other people. So what I might say is, “You know what? Who’s got a top podcast? I’m running Mixergy. I want more people on my site. Well, Tim Ferriss does.” I went to timferriss.com and he hasn’t touched that in a while. It’s still powered by Weebly and has some broken things on it. So I realize, “Okay, that’s not where he’s doing it.”

I just remember, it’s Four Hour Work Week. So I go to fourhourworkweek.com. Let’s see what I can see at the top saying what it’s about. And then there’s a start here, collecting an email address. Even though he’s a podcaster, he’s going for the email address first. That’s interesting, right?

Is that what I would do and then go in and see how he’s changed the site over the years to get a sense of how his promotion is changing?

Julian: It’s a good question. So it’s worth me teasing out that the long over short rule of thumb isn’t universally applicable. It is for most companies. Most B2B, SaaS, most e-commerce, but for mobile apps and for things that are very self-evident, like, “Hey, I have a podcast,” like Mixergy, for example, you don’t need to, because the point of going long is to proactively address objections and proactively explain what it is you do, if it’s not self-evident. And a lot of SaaS software is very confusing, right?

A lot of novel e-commerce products, like Tovala, are actually kind of confusing. But if you’re sending someone to the podcast, they get what it is. You don’t have to waste their time. So if you’re going to do a long homepage, maybe make that in service of having a gallery of all of your past episodes, but you don’t have to go on and on for a podcast.

Andrew: What I mean is that if I’m trying to see what’s working for other podcasters to reverse engineer them, it doesn’t seem as easy as going to see their site and watching what’s changed over the years because Tim hasn’t changed his site for years. I mean, literally, I think five years, it hasn’t changed.

It feels like there’s something else and it feels like a lot of times you can’t tell what people are doing, unless you’re deep in the site, because as a superficial view of someone’s site, you don’t know why they’re changing or what they’re changing and is the homepage and the site even really where the action is. Does that make sense?

Julian: It absolutely does. So first thing worth asking yourself is, are you looking at a website where the website is actually the bottleneck to conversion? Here’s what I mean by that. People don’t need to go to tim.blog to listen to his podcast. They just go right through iTunes. They go through Spotify. So Tim has no extreme motivation or incentive to go and waste time updating his website endlessly when no one’s even using his website to get to his episode.

Andrew: I’m sure he’s not, and I don’t know that we would even know that. Wouldn’t we just say, “Hey, look maybe it’s that”? And then we’d walk away with this impression and then completely miss that what he might be doing is buying a bunch of ads to landing pages that collect email addresses in exchange for giving you his 5-Bullet Friday or his checklist from what he learned from interviews, and then from there he gets people to subscribe. But none of it happens on his website. It all happens on ads that lead to landing pages that lead to an email, and then from there there’s a podcast.

What I’m trying to say is it doesn’t seem like it would be obvious by just looking at somebody’s homepage and reverse engineering it how they do their business.

Julian: Right. So, first, it is worth reiterating though that for podcasts, this is not the best example. But, that said, you could go to tim.blog/sitemap.xml. That’s the sort of auto generated file that’ll reveal every page on their website. You can just go look at them and see, do they actually have a special landing pages, and then if you pop one open, something like try.tim.blog or try.tovala.com, which happens very often, then you can go look at those using that site map file.

But the real answer is you did the correct first step, which is, either based on your experience or based on archive.org, looking at the history of his website. If you first know or see through archive.org that his website hasn’t changed in ages, then that is a sign you shouldn’t be looking at it because this is not someone who takes A/B testing and growth marketing seriously on their homepage.

So what you really want to do is go for other companies who do have showed progression over time on archive.org and most importantly are actually succeeding wildly. So obviously, Tim succeeds wildly, but most products that you’re trying to benchmark off of in this process that I’m sharing should be ones that are actually at the top of their game, are crushing it.

You could test this through a couple of means. You can go to alexa.com and see how heavy their traffic’s spiking over time, and you can use other tools to audit, “Are they actually growing and how much traffic are they getting?” And if they’re getting a lot and their website’s changing over time, it means they’re incentivized and actually acting on the fact that they need to A/B test their homepage.

Andrew: Okay. All right. So you did this, you got some of the things that you brought back. Some of the understanding that you got, which is pricing needs to be important in the way that you explain it, not just the price itself, you understood . . . Well, what else did you get as you were breaking down other sites? And then I’d love to know what you did to Tovala and how that worked out?

Julian: Yeah, sure. So long over short, on the homepage, A/B test product photography, be very aggressive on A/B testing pricing. There’s also a few little things I’m reminded now as I look at my old notes. This is a fun one. The hamburger menu. So when you go to tovala.com and you shrink the window down so it’s mobile sized. I don’t know if Tovala does this, but something I learned afterward is those three lines, which are called by web designers as the hamburger menu, because it looks like hamburger, aren’t actually understood by a lot of people throughout America.

They have no idea what three lines signifies. They’re like, “Where did all the links go?” And so they assume the links have disappeared on mobile, because they’re not a tech-forward designer who gets what this icon means and they never navigate. So when we actually just change those three lines to the word Menu, suddenly clicks spike up 20%, people get to where they want to go.

There were so many small learnings like that that were so interesting, and here’s the framework for validating that you’re not just chasing a rabbit. When you find something like that, you don’t put it in your notes and say, “Boom, I just got a growth hack.” You then try it on other businesses, other websites and if it uniformly works, then you know you’re on to something.

Andrew: Okay. Yeah. You know what? I’m looking at ours, when I shrink down my site, it does give you the hamburger menu, and really, for what reason? It’s not like it takes up that much space to have the word Menu, and we clearly have that available to us. Yeah, that makes a lot of sense.

Julian: A few more things. Those images that represent the product in the product page, very often you’ll get a conversion boost if you switch those to GIFs. So even if the GIFs are meaningless, even it’s literally just a static image of a meal moving up and down for no reason, just kind of dancing around in the background, that catches your eye and is distinguished enough from every other e-commerce product page you see in your life, that you start getting a bit more attention.

And I’m not a huge fan of gimmicks like that, that are sort of distracting for the sake of being distracting, but I would be lying if I didn’t say it does help with conversion. So I wanted to share that as well. I have more. I can keep sharing.

Andrew: I’m curious about the process for knowing that. Would you know that because you looked at other people’s sites and you started to see it or . . . ? How do you get to that?

Julian: Yeah. So that’s a good example of, “I’ll see it. I won’t take it for granted.” I’m not going to say, “Oh, Blue Apron’s doing that. Therefore, I must.” Instead, I’ll see it, I’ll put it in a note, a spreadsheet of ideas I’m going to test one day. And then eventually, when I’m queuing up my next A/B test for Tovala, that might be something I try. And I’ll let the A/B’s results validate whether it actually works or not.

Andrew: Okay, I get that. I do notice that as I’m looking at these different sites, a lot of them have video at the top. So it’s basically something that could be an image, like somebody eating salmon, but they don’t. They actually show a video of somebody eating salmon, or they show a video of something else happening.

Julian: Exactly. I have a couple little . . . These are complete hunches. So there’s no data here. A couple of sort of suggestions as to why that might be helpful. One is it’s just novel. It doesn’t look like every other website, so you pay more attention, like I said, about the GIFs.

Another one is it conveys you actually have money. To produce a video because they’re note a fly-by-night operation, right? Especially if it’s live production. Anyone can do animated, which are very overplayed, right? And so those don’t score you many points. But this is actually a really good segue to the biggest wins for Tovala.

So the biggest wins for Tovala resulted from me doing competitive analysis off of the website, meaning I was doing the competitive analysis on the ads themselves. And now what Facebook does is it reveals all of the ads being run by any company running ads. Period.

Andrew: Yeah. That’s one of the things I learned here today, that if I just go to say, facebook.com/blueapron I can go into Ad . . . under Info, I can see the ads that they’re running. There it is, on the side menu, and I don’t know how long this will last. But it says, Info and Ads, and then I get to see all their ads. Right?

Julian: Exactly. So here’s the kicker. There’s two tricks here. If you want to do competitive analysis on a competitor’s ads, first make sure it’s a company that appears to be doing pretty well, right? If you do this on someone who’s a fly-by-night operation, you might chase the wrong rabbits.

But if you see them running an ad, log that in your spreadsheet of ideas of things you are going to test yourself one day, if they also pass your sanity check. If they seem crazy, don’t do them. If they make sense, if they resonate, consider them in the future.

But another trick is go to their website for no other reason than to get retargeting by them. And so turn off your ad blocker so that when you go to Facebook, you actually want to see their ads for the first time ever, you want to see Facebook ads.

And here’s why this matters more than the previous technique Andrew was sharing, where you go look like all of their ads in a database that Facebook now shows you, the problem is, when you look at that database of ads, there’s no indication of which ads are doing well. You have no idea which have spent $1 million and which have spent $1. And you do want to chase the wrong one.

So when you get retargeted, you get to see how much engagement there is. If there’s tons of hearts, likes, comments, you know it’s been running for a while and you can look at all the dates those comments were left, and people don’t leave ads running for a year that are garbage performing ads.

So that’s the real trick is, rather it’s a trick, but that was the real win for Tovala was leveraging the insights taken from competitors.

Andrew: From seeing the ads that they were doing, and then as a result, can you take me to what happened? Because they worked with you, what happened?

Julian: Yeah, for sure. I don’t want to claim ownership over these results. I worked as a team with them, and their team is truly the best growth team we worked with. But within about, I believe within eight weeks, revenue increased 3 or 5X and stayed that way indefinitely. So it’s just like we just tapped, it was like potential energy that we just released into kinetic. And it just took two months. And that was putting these processes into place.

And let me categorize what those two processes were very briefly. One category was the website. So most of what we’ve discussed so far. The other category was going very heavy on the ads, and the wins probably came, if I had to throw a number out there, maybe 75% from the ads, 25% from the website.

Andrew: All right. We didn’t even get into the ads. We’re going to talk about them in a moment. But first, I’ve got to tell you and everyone who’s listening to me, that if you need to hire a developer, you got to go check out Toptal. Julian, have you ever hire them or hired from them?

Julian: Haven’t used Toptal yet, no, but I do want to.

Andrew: What are you thinking about using them for? If you would use them, who would you hire through them?

Julian: Salespeople and engineers, which are notoriously hard to find.

Andrew: Yes. So salespeople, they’re not great. They don’t do sales yet, but engineers they do. Why would you go to Toptal for engineers? What kind of engineers would you even need?

Julian: So we want frontend engineers and we want them to be remote. And because we’re a remote company.

Andrew: To do, what? To create . . . ? Well, actually, I’m even trying to think of what Bell Curve would need an engineer for.

Julian: That’s actually a really good question, because there is a huge part of what we do that people do not realize has to do with engineering — conversion, tracking and debugging. So it’s just getting the pixels on websites to make sure we’re properly tracking every step on the website in the funnel so we can fix what’s broken.

Andrew: Oh, and . . . Wait? But would this be some software that you’re using that’s off the shelf or are you looking for an engineer to actually create your own analytics package?

Julian: So it’s less about building software and more about the implementation.

Andrew: Oh, got it.

Julian: So [we get the 00:31:12] pixel from Facebook but it is much more complicated than you’d think to get it integrated into someone’s complex mobile app, for example.

Andrew: Okay. All right. That’s the type of project where if you’re looking to hire somebody to do that type of one-off project, I’d urge you to go check out toptal.com/mixergy. One of things that people have told me they used Toptal for is the specialty things that are really hard. They want someone who’s experienced, who is expert at it to come in and do it, and then they’d be okay with that person even leaving after the project is done if the team can take over. But they want that experience. They don’t want someone to figure it out. They don’t want to go through a long hiring process to find someone who’s just okay. They want to just go through a quick hiring process and get someone with experience and who’s already phenomenal at it. If that’s you, I urge you to go check out, toptal.com/mixergy.

I’m looking at Julian. You just sent out a text. What is this? “They sure did.” Oh you’re responding to people. Get back into your responses to them in the chat.

Go check out, if you’re listening to me, to toptal.com/mixergy. You’ll get 80 hours of Toptal developer credit when you pay for your first 80 hours, which is amazing in addition to a no-risk trial period, toptal.com/mixergy.

There’s someone who’s really a fan of yours, Julian, who I don’t know his name or her name, but they’re coming in, I guess as web visitor. I’m going to type, “What’s your name?” What’s the question that they asked?

Julian: Yeah, so they wanted to know, “Did Tovala actually see positive ROI on the ads?” Because it’s notoriously competitive to try to push e-commerce products on Facebook or foodstuff, right? And the answer is yes. And the way the ads work, this actually I can summarize the whole ad thing in just one quick blurb, which is focus on video for Facebook, mobile, and Instagram ads. Video usually outperforms static images ads hugely, particularly if you’re selling physical goods that lend themselves to like savory, enticing imagery, like food does.

And the second key trick to getting the ads to work is to point out how you differ from competitors. So if you’re just another food service, people don’t understand why they should pay attention. But if your ad itself or the resulting landing page you send people too from the ad has the confidence and the brashness to compare to the competitors, people really appreciate that. You’re doing the competitive research for them and saying, “Blue Apron does this, we do this. This is how we’re better, this how they’re better,” and for people who Tovala is the best fit, that resonates very deeply.

Andrew: All right, that makes a lot of sense. I think, before we move on, was there something else you want to close up of this Tovala conversation with?

Julian: No, nothing.

Andrew: Okay, I just don’t want to cut you off. It’s still kind of weird for me not to see your face, but I have to say it reduces some of the pressure to that when I can see someone’s face in an interview, I tend to focus on everything and that becomes a lot of mental work.

How do you end up with the name julian.com or with the domain name julian.com?

Julian: Yeah, sure. So I think I was 14 when I learned what domain names were. And I was like, “I wonder if they can flip this around. What if I could like speculate?” And didn’t even know what the word speculate meant, but I bought a domain for my friend for $1,000, deleted.com. And a week later, as a 14-year-old, I sold it to his friend who didn’t know he owned deleted.com for $22,000, and it was through this forum called dnf.com, domainnameforum.com.

And I was like, “Okay, how do I make more money? What do I do here?” And so the first thing I thought of was like, “Wait, can I buy Julian.com? Does that have value?” And so I did a Whois look up. And I saw, which is how you figure out who owns a domain name, and I saw that it was owned by like Percy, Samuel and Julian.

And it looked like it was a law firm or something. But the website was gone. No one was alive. Like Percy was dead. Percy had died. And I was like, “How do I [inaudible 00:35:20]?” So I called the number. Nothing, no ring. And then I just forgot about it for a few years.

And then at some point around, I think I was 19 or something, 19, 20, I was like, “You know what? I’m getting this domain name, I don’t care what it takes.” And so I actually I just aggressively researched, I like private-eyed, who were all the people that knew Percy?

His name was actually Julian Percy or Percy Julian, one or two, something like that. And so I found out his old business partner who was still alive. And I contacted this guy, the sweetest guy on the planet. And he’s like, “Listen, I don’t know what you’re talking about. I certainly know Percy Julian or whatever the guy’s name was, but I don’t know what Julian.com is.” And I say, “Listen, if I go and track down the company that owns the domain, it’s like the registrar, like the GoDaddy behind this domain, can you show them legal papers that you’re now the entitled owner being his old business partner of julian.com?”

So the end of the story is I handheld like an 80-year-old man into figuring out how to put together legal documents to convince GoDaddy to send him the domain. It was actually a different registrar. And by the time he got it, he had no idea how to give it to me. And that was a three month journey of getting a domain from him. So it was a long story, but long story short, he was very happy. He got thousands of dollars and now I’m julian.com.

Andrew: Oh, what year was that?

Julian: I guess it was like 10 years ago, so like 2009.

Andrew: That’s fantastic. That’s a late in the game here, actually, to be getting that.

Julian: Yeah, I think that day at . . . The week, maybe, right after I bought it and switched the contact details on it, someone was like, “Hey, can I buy it from you?” And he was like for the Julian City or County in California, some marketing purpose, someone wanted to buy it for.

Andrew: You said that you had some insights for growth from the last month. What are some of the insights that you’ve got?

Julian: Yeah, I’ve been writing down some notes so I could share with you guys. So these are just things that I was keeping in a log and I haven’t fleshed out, but I just thought they were really cool. I want to show.

So the first thing I’ve noticed is something called . . . I’ll try to keep each one of these at like 30 seconds. So this thing is called accelerated retargeting. This has probably doubled or tripled performance for all of the accounts I’ve tried this on. What it means is, instead of retargeting someone with ads, meaning you show them ads after they’ve been to your website, the normal way of doing it is you just kind of trickle and trickle down and show them ads for the next six months and they see it here and there. That’s not the right way to do it.

Accelerated retargeting means you blast them with retargeting in the first four days after they hit your website. And then from the fifth day and onward, for the next half year, however long you continue retargeting, you drastically scale back the frequency so they only see like once every three weeks.

But if you blast them in the first four days, while you’re still very top of mind, that’s how you reach the perfect intersection of profit and revenue volume and conversion volume.

So it’s pretty awesome. It is reliably working for our clients.

Andrew: Okay.

Julian: I’ll pause it in case you want to dive into anything, otherwise I’ll just shoot.

Andrew: No, keep going.

Julian: Okay. Another one, this was interesting. So a CTA, a Call To Action, which you guys probably know, can be a button, right? So like at the top of your website, it could say, “Click here to contact us,” or, “Click here to go to iTunes,” whatever. Try to avoid making them too prominent in your hero, meaning the top of your website, what people see before they scroll down. Everyone thinks your hero has to say, “Buy now. View product. Shop now.” It doesn’t, actually.

Here’s what happens when you have a huge call to action in your hero. People click it reflexively. I know this because I look at heat maps. People click it and they never scroll down the rest of the page to actually learn what it is you do. So by the time they get to your product page, they just bounce because they’re like, “Well, I don’t know what the point of this is. I don’t want these products.”

But let’s say you remove those buttons and you force people to keep reading the homepage slowly, and at the bottom of the homepage you have a CTA. At that point, before they click the CTA, they’ve now read why, how, why they should care, how you compare and that’s when they’re better informed to buy.

Andrew: You know what? I do see on Tovala, there’s a clear shop button, but it’s this they are to be available whenever you want it. And let me scroll all the way . . . Yeah, at some point in the middle there’s another more prominent shop button when they’re showing me the prices. I could see the difference in that.

Yes, it’s available, but that’s not what they’re trying to drive me towards.

Julian: Exactly. Basically, think of it this way, use this rule of thumb. A more informed visitor onto your product page is more likely to buy. And so make them informed before you drive them to where you want them to go. There are always exceptions, and I’m not saying, “Don’t have a nav bar with links to all the key parts of your website.” But don’t try to make it the thing you have them do within four seconds of [page 00:40:10]. Otherwise, what is the point of your homepage, right?

Andrew: By the way, from earlier when you’re talking about retargeting, Alban Brook in the chat is saying, “For retargeting, what type of ads are you seeing that work best for your clients?”

Julian: Retargeting is interesting. I like the segmented. So in the first two days out of those four days, I give broad ads, like here’s a broad reminder of what we do, why, why we’re cool. The next day I try to up the urgency. So the third out of the fourth day, I start giving them specific ads with specific products. So maybe we sell clothing, I’ll show them my most popular T-shirts. And on the very last day, I increase the specificity and time pressure even more. On that fourth day, there’s a time limited offer.

And so you want to escalate reasons for why people should click if they haven’t yet. And I’ll round out this answer by saying, as with most ads on Facebook and Instagram, video tends to outperform images. So really try to focus on producing some videos.

Andrew: Wow. I’m so surprised because I don’t tend to watch videos on these platforms, or maybe I do and I don’t notice it, because of the subtitles.

Julian: Yeah, it’s interesting. I make sure that videos are designed to be visual first. So if you require audio, if you have like an explainer video that goes into a 60-second pitch, you’ve already lost, unfortunately, because these videos are muted by default. So they’d have to opt in.

And so the trick is if you’re going to rely on explanations beyond the visuals, then do them Buzzfeed style, where you just like animate text on to the actual image and animates in and out in a very pretty eye enticing fashion.

Andrew: What’s your revenue? What are you guys doing in your business?

Julian: So we used to be a very low seven figures, and then we’ve just pivoted half the business. And so we’ve taken a big risk, the actual founding team, cut our salaries drastically in pursuit of not having to raise funding. So to bootstrap toward building something called demandcurve.com, which in short is a way to train people to become growth marketers like I am without having to pay anything upfront. So that’s what we’re focusing on now.

Andrew: And so, when you say low seven figures, you’re talking like $2 million a year, is what you guys were doing?

Julian: Exactly, right. Yeah, around that area.

Andrew: And so now you’re saying, “I’m going to cut back on my clients. I’m going to raise awareness with what we’re doing, and I’ll create a course where people can sign up and learn from us.”

Julian: Yeah. So what we want to do is we were inspired by these programs called ISA, so Income Sharing Agreement. And my co-founder, Asher, started the first one called App Academy, where you come, you learn coding, you pay nothing upfront. You only pay if you get a job after.

And so we’re doing the same thing with marketing. We train you for free. You only pay us if we get you a job. Otherwise, you pay us nothing. So it aligns incentives in such a way where we have to be really good at teaching, and we have to be really good at knowing growth, otherwise you will fail your interviews.

Andrew: I get it. Is that the Lambda model too?

Julian: Exactly, right. Yeah, so ISA, Income Sharing Agreement. And I think you’re going to see a lot of this. You’re going to see it for sales, you’re going to see it for a lot of verticals and because of the expertise and network we have in growth, we felt now was the right time for us to do this.

Andrew: It looks like you just closed your application on the 23rd of April, and you’re about to start in a few days?

Julian: Yeah, first batch. We’ve been training companies for six months, but we’ve not yet trained individuals to become hired elsewhere. So right now, founders bring us in to train them to do what we do but they already work for a company.

Andrew: You mean the founders get trained, or is it their people?

Julian: No, you’re exactly right. We actually train the founders. Where mostly YC companies that go through the program maybe 70% and basically they graduate YC. And they’re like, “Hey, we understand the importance of growth, but there’s a tactical depth that we’re currently missing. And can you just teach us the specifics of how to run Facebook ads, how to do conversion rate optimization?” And so we go super deep, boringly dry deep, and we make sure they learn it through deliberate practice and sample projects.

Andrew: Okay. Yeah, it looks like Demand Curve is not fully up yet, but you’ve still gotten so people through the program. Am I right?

Julian: That’s right. It’s the first batch, and the second one will be ready . . . It will be kicking off in July. So there’s a bunch of cohorts coming up.

Andrew: All right. You were going through some of the things that you learned recently. What’s the third item on the list?

Julian: Yeah, I’ll hit you with a couple more. This one I love because it’s a frame shift, and I feel like these are needed more in growth, which is don’t dismiss break-even paid acquisition. So that means, is paid acquisition, the marketing jargon for paying money to buy people online or by any means. So ads are a great example of paid acquisition.

If you have an ad channel, like Quora ads, which kicked up a couple of years ago, which we like, it is okay if it’s only break-even. As long as it’s not the primary thing that you’re depending on to grow your business, break-even paid acquisition, meaning you’re only earning back as much as you spend on the Quora ads, is free brand awareness. All those eyeballs now know who you are, and in the future, it’ll steamroll and accumulate to be worth it. So keep that going. Don’t turn it off because it’s break-even. But I’m not saying that’s your goal. Your goal is still first and foremost get a channel profitable.

Andrew: You know what? That’s actually something we’ve been considering internally. We’ve been buying ads for this chatbot course program that we’ve got. And my frustration with it is it is just breaking even. It’s breaking even over like three months. I thought, “Let’s just get rid of it. One less distraction.” You’re saying, “Keep it going, Andrew, because people are getting to know you and some of those people will end up buying something in the future or being in your world”?

Julian: Yeah. Consider this. The amount of money companies spend purely on brand awareness where they get nothing back tangibly is huge. And now you’re getting brand awareness for free. Why shut it down? Now, there are some like maybe side effects you’re not accounting for. For example, if you’re blasting them on Quora and it’s breakeven, but you’re annoying them, right, and they’re also seeing your ads everywhere else on the planet, that can have some negative externalities that you might want to avoid. But generally speaking, as a principle, don’t shy away from something if you’ve done everything you can and it’s not profitable. As long as it’s breakeven, consider leaving it running.

Andrew: Okay, what’s next?

Julian: Let’s see what else we got here. Okay. So geo rollouts. So this is conveying ubiquity by focusing on one tiny region at a time. So here’s how most people do ads. They boot up Facebook, Twitter, whatever, Twitter ads that is. And they say I’m going to target all the United States. And what they’re doing is they’re taking like $1 million or whatever the spend is $1,000 and they’re dividing it into tiny little crumbs and distributing it throughout the entire United States. That’s not super effective, because part of what makes ads work, part of what makes people click your ads is accruing a level of familiarity that they know who you are after a few months and they finally feel comfortable clicking you or buying from you.

So if you want to take advantage of that psychological phenomenon, what you should actually do is take that budget and spend it in a very tiny geographic region so you look like you’re the biggest dog in the playground and so you’re everywhere. They think you’re everywhere, but they don’t know that you’re just targeting San Francisco. And the way you do the trick here is you do it at the same time as you’re advertising on Facebook, unlike every other channel, so you do a cross-channel blast. No matter where they go on the web, they think you’re the biggest thing in the world because they don’t know that you’re not spending that same amount of money throughout the whole U.S. They don’t know it’s just San Francisco.

So you can accelerate the rate at which people become familiar with your brand, and this helps in particular for brands that really need a credibility boost in order to get business.

Andrew: Julian, wouldn’t I, even if I was advertising all across the U.S., I’d still be targeting the same type of people and so to that group of people, I would feel ubiquitous. Why does it have to be geographic? Why can’t it be psychographic? Say, “I’m going to pick everyone who’s in their early 20s who happens to have a tech startup.”

Julian: Absolutely, no matter how you limit it, it’s okay as long as you entertain limiting it. So you’re totally right. You can still do all of the U.S., but maybe the way you limit it is by doing just 20 to 25 people, people who like, you know, delivery and you can narrow it by other means, so the same principle applies. It’s a great point.

Andrew: Okay. You have one more and then I want to talk about my second sponsor, and I want to come back and just be more practical for myself to see what I can use from everything you said?

Julian: Sounds great. Sounds great. So how about as a last one saturation.

Andrew: Okay.

Julian: So this is the biggest pain point for all of our clients is they come to us and they say, “Julian, I want to spend $X per month on Facebook ads.” And I say, “Well, where are you coming up with that number?” Like, “Oh, I don’t know. It’s just part of our investors. These are the targets we set internally.” And then I have to remind them that Facebook doesn’t care about what your targets are.

There’s only a maximum amount of value you can get out of Facebook ads, and it’s capped in part by how big is your addressable audience on Facebook. So if you come to me and say, “Spend $1 million this month on Facebook ads for my audience,” that is a meaningless statement. The correct question is, “Hey, Julian, how much can I spend given my audience on Facebook? And what’s the ceiling I should stay under so that I don’t spend so much money that I hit saturation?” To wrap up this point, let me just quickly define saturation.

If you spend too much money on a limited size audience, on any ad channel, they’ll see your ad so often, so actually it’s like the extreme funhouse version of what I just said a moment ago, where you want to get that frequency. If you push it too far, people hate your brand after two months because you’ve just been spamming them endlessly and then your CAC, your Cost to Acquire a Customer, plummets after like four or five, six months when people have seen your ads too many times and they start getting burned on also a lack of ad variety.

Andrew: Okay. And so the way that when we’re talking to our agency and saying, ” I want you to buy ads,” I shouldn’t tell them, “Here’s the budget,” I should tell him instead what?

Julian: Yeah, I appreciate you recapping this. So you should go to them and say, “What’s the best we could probably push per month without ever hitting saturation, without ever annoying our audience so we can keep ads running forever?”

Andrew: And that’s what I would tell them, even in the beginning?

Julian: Oh, yeah. So you never want to exceed that, because if your goal as the company, if you want to get more listeners for your podcasts, more customers for your business, so “more” is the key word here. If you want to increase the volume of customers, the onus is not on Facebook to get you those customers. The onus is on you to diversify your channels.

So if you want more than what you can get out of Facebook without saturating, the next thing you do is you get another channel to work and you complement it to Facebook and you increase total volume overall through multiple channels, not by squeezing Facebook for everything it’s possibly worth.

Andrew: One of the things that I liked about what you said to Courtland was, you said, “You want to keep looking for the new networks, the new social networks, new sites that are starting to sell ads and be one of the first people on there before people have figured it out.” And so, if Pinterest is going to sell ads, if Quora is going to sell ads, you want to be one of the first advertisers on there. And you say, “All of these guys are going to have to sell ads because they can’t make money in any other way.” So just keep an eye out for them.

Julian: Yeah, exactly. I do want to caution though, that it goes a little bit into growth hack territory where it’s not sustainable because after you’re early on it on a new ad network like Quora released ads two years ago, we jumped on it, benefited hugely. We got clicks for like two cents. After a while, everyone jumps on the bandwagon and then you’re, you know, diminishing returns.

And it doesn’t happen very often. There aren’t huge social networks, like every week that are doing this, but it does happen almost once a year. If there’s a Snapchat or Tinder or some social product that releases their own ad network, jump on it and it could be really exciting. Or here’s the funny sort of inverse, it could be a complete bomb and you can waste all your money experimenting hoping it has potential. Just like we did with Spotify.

Spotify came out with ads, maybe a year and a half ago. We tried it. It was the worst performing ad product we used in our lives, truly, and that’s saying a lot, because some others are really bad. It might be fantastic today, but in the first month, we basically wasted like $3,000 or more.

Andrew: Wow, I would’ve thought that . . . First of all, it’s not that much to lose, and I would’ve thought their ads would be better.

All right. Let me talk about my second sponsor. It’s a company called Ahrefs. You’ve used them. You still use them. What do you use Ahrefs trust for? And I’m going to tell you in a moment why they’re never going to sponsor me again. But why do you use Ahrefs?

Julian: So we use Ahrefs to track how all of our content marketing does over time so we know we’re doing it properly. And we also figure out what are the keywords that are leading to most of our traffic. And then we reverse engineer how do we better address those keywords going forward.

Andrew: What do you mean, you want to know the keywords that are doing well for you in search engines?

Julian: Exactly.

Andrew: In Google I should say, of course.

Julian: Ahrefs would show me, “Here are the thousand keywords that are currently driving my articles traffic.”

Andrew: I didn’t see that.

Julian: And then I’ll able to see which keywords am I getting the most traffic from and which keywords lead to the content marketing, meaning the blog posts that actually lead to purchases for my product. So I know which are most valuable and which are highest volume.

Andrew: Wait, they’ll actually help you know which ones are closing sales?

Julian: They won’t tell you that, but that’s my job to go into Google Analytics and tie that together.

Andrew: Got it, got it. Wow. So what’s a keyword that’s doing well for you on julian.com, for example, or do you do it for Julian.com?

Julian: Yeah, I do actually. I use it for personal stuff too. So if you type, “How to build muscle,” I should be the second result in Google. If you type, “Bodybuilding diets,” I think I’m in the first one or two maybe results. It’s funny. I wrote a guide a couple of years ago on bodybuilding and Ahrefs showed me how it took off.

Andrew: And so you said, “This is taking off. It’s sending traffic to my site. Why not? Let’s just create more about it.”

Julian: That’s exactly right. It gave me data to power decision making for what else should I write, where should my energies go for content marketing.

Andrew: That’s really the best way to put it. That’s what they do really well. What content should you create? They help do it. It’s really inexpensive. If you go to ahrefs.com, it’s going to cost you $7 to get started with them for seven days. If you’re not happy with it, cancel. It’s just like an interesting thing to go and experiment. Type in your domain name and see the results. Type in other people’s domain names and see the results. And then start to type in keywords that you’d like to get and see, what could you do. What are other people writing? And you’ll probably come across a bunch of junk that other people have written, that you could improve on and get traffic from them.

All right. It’s great service. I really like them. Here’s two reasons why I think I’m never going to get them as sponsors again. And Julian, as a guy in the ad space, you tell me if you’re right. I think I’m never going to get them as a sponsor again because they didn’t give me a unique URL. So how are they going to know if my people are sending traffic over?

And a unique URL means, a unique URL like ahrefs.com/mixergy, and I should be telling people, “If you go there, it’s actually going to cost you $1, not $7.” You know I mean? That’s number one.

And number two, I think these ads should not be confined to the ad spot. It is such a good tool, I should be using it throughout the conversation and then I happen to have them as a sponsor.

So when you’re talking about Tovala, if you happen to mention that Tovala is great for content marketing, I should just be typing them Ahrefs and then say, “Hey, when I look them up on Ahrefs, here’s what Ahrefs is telling me about their traffic.” Use that to inform my conversation with you. And then people see the utility of it and then they’ll sign up. That’s I think the answer. What do you think, Julian?

Julian: I think you’re right. I’ll play devil’s advocate on their behalf, which is it’s true not having a link is a little bit dangerous, flying blind. However, when you’ve done enough podcast sponsorships, you start to see that it’s kind of hard to track traffic through links anyway for a bunch of reasons. And so often what you’ll do is you’ll just go to your Google Analytics for the day the podcast aired, and you’ll just try to look for a correlation for a spike, and then you can see what that spike is over the baseline and assume that’s what came from Mixergy. So maybe they’re doing that, but I don’t know though.

Andrew: You know what? You might be right. I usually am against that because I think podcasts last forever. People are not just listening on day one, etc. But you know what? The truth is the majority people do listen on day one and they definitely will, the vast majority will listen within three days and, yeah, if you seeing a bump in traffic, that makes sense.

I also wonder, these guys know traffic really well. I wonder if what they’re saying is, “At some point in the process, we’ll ask them if they’ve heard about it on Mixergy. And if we get some number of people we’ll know this works.” So I don’t want to outsmart them, but I do have enough advertising experience with podcasts that the things that I never would have wanted to do that I have to recognize do work and one of them is a unique URL with a special offer that they can’t get anywhere else.

Still, I really wanted to work with Ahrefs. They’re such a good brand to be associated with, don’t you think?

Julian: They’re one of the few products that are very well respected among marketers and agencies.

Andrew: So I heard you with Courtland Allen, and I thought there were a few things that I took away that were interesting, like you said, “I’m reaching out to Courtland because I like his community, and by helping him, something good will happen.” And it seems like that’s what you did, you helped them think through his product, and then he had you on his podcast. Then you got a bunch of traffic.

And so it seems like that’s something you’ve done before. You’ve given people, reporters Tovala to have them try it out and that’s like a give up front way of working. You’ve written articles for people when you were working on Velocity JS to get them to care about it, right? This is part of what you do, and that’s something that I could learn from. Am I right about that?

Julian: Well, if you have the ability to write well, and you can sustain the motivation to write something to completion, it is undoubtedly one of the highest leverage ways to grow anything you’re working on. Your personal brand, products, clients, anything.

Andrew: I do, I think I do write well. I don’t have the energy to do it all the time, and I need it to be like a machine. I want to start using what you’ve said and not just say, “This is really interesting,” and have it be brain candy. I want to use it.

So one of things you did was you did write content for other sites. I don’t want to write content for other sites. I don’t think I could sustain that. I want to have a process where I pay someone else to do that. Is that possible?

Julian: Yeah, absolutely. So my friend, Nat Eliason, has an agency called growthmachine.com, and that’s exactly what he does. And there are other people like him. And so essentially they’ll find a network of writers, people from Upwork, people from the Blogosphere, and those figure out a framework for sourcing a lot of articles on your behalf, Andrew. And then they’ll also source editors who will sanity check the writing, how you want them to write. So that there’s multiple layers of sanity checking and sourcing doing a crowd-sourced approach, and the end result are quality controlled articles. You can pump out any volume you want to pay for it.

Andrew: To other sites?

Julian: For other sites, for your own site. You can use this in service as guest posting and others or for writing for yourself. I actually don’t write for other sites. I mean, I write a TechCrunch column on Growth, but that’s very high leverage. I do it for TechCrunch because everyone reads TechCrunch. But generally speaking, I’m not going to spend time writing for small blogs, and I recommend you don’t either because you already have a brand.

Andrew: Really? So then, what should I do?

Julian: If you want to pump out content for the sake of getting more eyeballs . . . well, let’s first distinguish. Do you want to write content to get more eyeballs, to get more authority or to build greater affinity? So let me tease those out really quickly.

Eyeballs is self-evident. Authority means people read your stuff and go, “Wow, this guy actually knows what he’s talking about.” And then affinity is writing stuff that makes people fall in love with you. So the stronger the affinity toward Andrew, the more people will, I don’t know, listen to your recommendations, right, the more they’ll be retained throughout your podcast episodes.

So what are you writing for? What do you want to write for?

Andrew: I would say to get more people to listen to my interviews, but then maybe that falls into the affinity bucket.

Julian: Yeah, a little bit of affinity, a little bit of volume. So you want to intersect those two. So you don’t want to do volume where you’re a content mill pumping out garbage. And you want to do volume that is also high quality content that builds like hooks. Like when I read this thing from Andrew, I’m hooked to read more. And the real affinity comes from, “I think this guy’s amazing,” or, “I think this girl’s amazing,” and you can convey that through a whole bunch of things. You can convey your personality. You can be extremely useful upfront, get out of the way, remove the fluff, don’t be a salesperson, just be a breath of fresh air. All sorts of things you can do.

But to answer the key question like, how do you increase volume through content marketing if you’re already large like Mixergy? It’s probably one of two things that come to my mind. The first is, maybe three, but let’s focus on two.

One is find the big sites that are big enough to move the needle on your own website — TechCrunch, Huffington Post, Forbes, Fortune, Inc., whatever. Write for them. Don’t write for small blogs.

And maybe number two is focus on an SEO component here. So write content for the sake of getting access to the long tail of some Google eyeballs that otherwise have never heard of Mixergy. And that’s probably the right answer for you, because that doesn’t spike like TechCrunch does where you get a little blip and then it disappears. It’s also incremental, like it builds on itself. It’s a really valuable asset.

And so my recommendation is what are the intersection of things people are searching Google for that are the things you actually deliver on. So if you’re going to write a blog post that caters to a keyword, like through Ahrefs, and that’s where I get this data from, when I see a keyword there, I’m writing a blog post in pursuit of servicing that keyword. I don’t just choose the keyword that has the most volume and the biggest potential. I choose the keyword that has the most volume, but also I can write something amazing for, or I . . .

Andrew: So I might just say, I might go into, let’s say Ahrefs or any other tool, to find the stuff that I should be writing on, and it might be something like the word “entrepreneurship.” It might be find a business idea. I’m just sitting down and I’m just searching. What’s the thing that our audience cares about it and that happens to have a lot of search volume.

And then, since I’ve got content anyway, I might be thinking, “Well, let’s just see if I could write something or have a writer write something based on the content that’s already on the site that will get access to that volume of search.” Am I right?

Julian: That’s exactly right. So because you already have content, you already have the goods. Most people don’t starting out. So in your case, you go through your content library, you look at your GA, your Google Analytics, and you identify all the videos that when people watch, not only do they watch it to completion, but they then love you, and as you can measure, by how many people watch that as the first video and then go and watch a ton of videos after that.

And you want to find those hero videos in your analytics. And those are your goods and then what you do is you back into that. So now you go and find the keywords that are most appropriate and highest volume for those killer-hero videos you have. That’s the framework I would bring to the table here.

Andrew: Okay. What else would you recommend to somebody who is not a funded startup, that does not have Tovala money in the bank, do to start to get some results?

Julian: For content marketing in particular, you don’t have to have any good or any money to start. So that’s the beauty of it. That’s why it’s like the ultimate acquisition channel for customers, because anyone with the hustle and motivation can do it.

So first, you want to use something . . . I mean, Ahrefs is not paying me. I know it seems that way. And I really joined in and went heavy on that pitch there because I actually do love them. I used them this morning. So use a tool like that to explore the keywords available in any space that you care to write about, or more appropriately, if you already have a business or a product, go research that space on Ahrefs. They have a whole bunch of tools like content gap and keyword explorer and whatever to go figure out what is you could be writing about.

And they’re going to rank those keywords and they’re going to show you volume and they’re going to show you keyword difficulty. And those two, they both matter the same. So keyword difficulty, if it’s really high, it doesn’t matter how much volume it has, you’re probably not going to be able to rank very well for it. So figure out where the balance lies. You can Google, like, literally Google for the phrase, something like, “Comparing keyword difficulty versus volume,” and figure out the formula that other people advise for knowing which keywords to triangulate. And then what you do is you start writing really good content that addresses those keywords.

And I’ll give a couple quick pieces of advice, because I don’t want to say that and then provide no actual tactical knowledge here. So think of it this way. Google’s goal with their algorithm is to deliver someone from Google to your website and not back again. You have to be the final destination, because if they go to your website, hit back in their browser and then click onto another result on the search results page, it means you didn’t give them what they wanted. The searcher was not satisfied. And Google measures that. There’s a million ways they can measure that, Google Chrome, Google Analytics, Google Search, everything. So they know what people are doing in the journey.

And so how do you get people to stay on your website and end their search on your website. This is the highest leverage advice I can give anybody for content marketing. And the reason this is so, and I’ll explain a few tactics in a moment, but the reason this is so important is because if Google sees that people stop at your site as the final destination, that’s how they quickly rank you at the very top of the first page. That is what they’re gunning for, assuming you’re a relevant topic and so forth, you’re relevant to the keyword.

So here’s the tactical advice very briefly. If you want someone to finish and end on your page, ask yourself, if they’re entering this key term into Google, what are they really looking for?

So if you have a blog that sells, I don’t know, tea, for example, loose leaf tea, and someone’s searching for, “Buy tea,” and they come to your tea post, but you have no links back to Amazon saying, “Here’s where you go and buy the tea,” or, “Here’s some tea we sell ourselves,” you’re just talking about types of tea, they’re going to click back because they’re like, “Well, great, you taught me about tea. Thanks for that. Where do I buy the tea?” They bounce, they don’t come back. And now Google’s like, “Well, people don’t like this page for this keyword.” And then Google down ranks you.

So think about what are the things you could do. And a couple quick tidbits there, like try to satisfy all of these sub-topics they could be thinking about when they arrive at your page. Try to exhaustively address why did they come here. And then also make that information easy to navigate, place a table of contents at the top. And the principle there I’ll end out on is, when you have a table of contents at the top of a blog post, you reassure visitors that there’s at least some section of your post that gives them what they came here for so they don’t leave prematurely. So again, it goes back to the theme, don’t let them bounce. Don’t make it hard for them to bounce through creepy UX stuff. But give them what they want.

Andrew: I use SimilarWeb. For some reason you use Alexa. Why don’t you use SimilarWeb, by the way?

Julian: Back when I was comparing these tools, there used to be three, Compare, Similar and Alexa, Compare was really relatively bad, Similar was relatively bad, and Alexa was relatively . . . I’m using the word relatively meaningfully. Like between websites, Alexa is pretty good at knowing which website is actually how much better than the next. Now, that might have changed, but they’re relatively good.

Andrew: SimilarWeb has gotten really good. And so I’m using them and I can see that, “How to build muscle,” is the top phrase that sends people to your site. Number two, “Julian Shapiro.” Number three is, “meal plan to gain muscle.” So I said, “Let me see how Julian does this whatever he’s talking about.” And so sure enough, as I scroll through this, not only, like if you say oatmeal as part of the meal plan, you actually link to Amazon using some tool . . . Let me see what the tool is. You’re using aforward.com to send people to Amazon where I could buy the instant oatmeal. You’re talking about Soylent powder, and so you link me to Soylent to go buy that. And now I get to see how to do this whole meal prep. And it’s part of this whole, basically online book that you’ve written and the table of contents is always visible so I could always see how this fits in with everything else. I get how you’re doing it.

Julian: Exactly, and actually I should point out one distinction, which is, SimilarWeb might be awesome for a lot of things and I’m pretty sure they are. I’m just referring to them for comparing traffic volume. But yeah, for what you’re describing right now, keyword research, they’re pretty cool.

Andrew: Yeah. I also like to see like what sites are sending traffic to people. There’s always some random thing. Did you buy ads to promote your podcast with Courtland?

Julian: No.

Andrew: Why is it doing so big? Why was it so hot?

Julian: No, I didn’t buy ads, didn’t push it to it. I don’t know. I mean, anyone who’s listening to this now, if they like this, if they’re still paying attention, they’ll probably figure it out, maybe they like the specificity of the advice. I don’t quite know, but I guess one thing I try not to do speak in vague terms. I’ll start vague just so I can set the lay of the land and then I try to go tactical very quickly. I think people appreciate that. That would be my best guess.

Andrew: All right. Just trying still figure you out. And I think I’ve got it. I think I have a little bit more of it. What you do is you do marketing for tech companies in a way that is based on what’s worked for other clients of yours and what you’re seeing work for others. You’re very data driven. Am I right about that from what I’ve seen about you?

Julian: Yeah, I try not to wing stuff too much. Although a lot of my advice is sort of off the cuff, but it’s coming from a bunch of data points.

Andrew: You seem kind of obsessive, just sitting there working all the time, right? I don’t see you having much of a personal life.

Julian: That sadly is not far from the truth.

Andrew: Really? Do you think that’s going to go on for a long time?

Julian: No, I think it’s been a really fun year pivoting Bell Curve into this school where we train people, or this bootcamp rather. But no, I think now it’s time to rebalance things.

Andrew: All right. I’m glad to hear that. I hope that you and I will get to see each other at some point in the future. I understand that you’re still keeping, like staying at home. I hope you get more shirts. Are you going to have someone deliver it?

Julian: We’ll see, we’ll see, we’ll see where it goes in my Asana list of priorities. I might be shirtless for the next couple of months.

Andrew: Really? Honestly?

Julian: No, I’m just screwing with you.

Andrew: Okay.

Julian: I’ll have something in an hour.

Andrew: Okay. I can’t deal with being shirtless just sitting at home and working at all. I know there are people, I do like Skype calls with the . . . or face, a video Facetime or something and they’re shirtless. I just can’t deal with that. It’s just too distracting. I mean, I don’t mind them, I mind me.

Julian: Right now every listener is just thanking me for not turning on my video cam.

Andrew: Yeah, I think it worked out even without it. I think it worked out without it because you’re someone who just does your homework. I’ve seen that in you in other podcasts. I just get a sense that that’s what you are. So you had a bunch of really short tidbits. They lasted 30 seconds, which you know people’s attention is capable of handling. And you told me before we started, “Andrew, you could not insult me,” which means that I was able to just say whatever I wanted and not worry about it. So that helps.

All right. If anyone wants to go check you out, actually there’s two places, they could go check out bellcurve.com, which is another awesome domain or julian.com where they could see pictures of other shirtless people who have learned from you how to be healthy.

Now I understand why building muscle is a big section of your site because Ahrefs and Google have told you that that’s what people want.

Julian: That’s right.

Andrew: I like the little, like the buttons and stuff on your site. Sure enough, I could actually switch from, “I’m a man,” to, “I’m a woman,” and all the images change to go along with who I am. That’s great. I love the details of the way you work.

Julian: Thanks, man.

Andrew: I want to thank the two sponsors who made this interview happen. The first, please, guys, if this did not work for you guys for this ad, let’s talk. But it’s Ahrefs, I really love your site. I love your software and more importantly my guests love it. So Ahrefs, if anyone in the audience is looking to do some research on what kind of content to create, that’s the site to go to. And if anyone at Ahrefs is listening to this podcast at all, let’s talk about how to make this ad even better. I know you guys are, like you’re into improving and I am too.

And I want to thank the second sponsor, the one that is way into maximizing everything. Julian, you would love the growth people at Toptal. The ad, by the way, was doing well for them for years, the Toptal ad on Mixergy. They weren’t happy. So why aren’t you happy? They said, “Well, we want to understand how many people are not using the URL but are still buying,” I go, “It’s working. Why don’t just like stop it?” They go, “We need to know. Maybe we should be buying other ads from other podcasts. If we know exactly how much money we’re making from you, then we’ll know in general, how podcasts could work for us.”

So they were doing this intense analytics about time of day when we publish and how many times people were coming over, and they realize, it’s not just about podcasting, is just who they are. These people are to a level where they care about data, that you wouldn’t believe. They’re obsessive.

I shouldn’t say you and Julian would believe it. I didn’t believe it, but I’m grateful to them for sponsoring. If anyone wants to go check them out, it’s toptal.com/mixergy.

Thanks, Julian.

Julian: My pleasure, Andrew. Thanks for having me.

Andrew: Bye, everyone.