How a 10% Price Increase Doubled This $50M Founder's Profit

Why is the quality more and most important? I'm stating that like it's a fact. So imagine in a marketing funnel if I put nothing but perfect ideal customers into the top. Every single visitor is a perfect person who has exactly the problems that you solve, who uses the language that you use on the homepage, who already has the budget allocated that is big enough to fit your pricing page, right? What happened is you get a lot of conversions. Even if your site isn't that good, even if your sales team's not that capable, whatever, they'll just take the—we call them order-taking sales people—they'll just take the orders. If quality was very high, then even if every other step in the funnel was poor, you would still do really well. Good quality means like you can't help but win.

Hey, listen. A lot of people. I checked the other day. Last year there was a million uniques. So that's millions. That's legit, a million. The average time on site for literally just dragging everything, the average time on site is four and a half minutes. That's not bad for 100% of the traffic that ever hits it, right? So that's not bad. Millions is fair, it's fair. Maybe it's billions by the end of your life, you know.

Right. Yeah, one of... Right. But if you're getting... It could be, it could be. We'll find out. Could be people are confused and it's taking them too long to find something. This is true. You've got to be careful about metrics. They tell you something, but you have to be very careful what you extract from that.

If you're trying to get to the truth, spin's not useful.

Yeah. What I found is that there are a few special areas in a business where small changes in those areas have a large impact on something you care about, like revenue or profit, whether you have great employees, whether customers stay—you know, these kind of really important things. Most are not. Most things you do are incremental. They may have an impact that you may not even be able to measure, but maybe in total they do. Most stuff is just you just grind it out and you get some kind of possibly unmeasurable benefit. But there's a couple areas where that's not true, where it magnifies or multiplies the effort. There's not many of them; most things are not like that. And so if you can find which ones those are, that's usually worth a whole lot of your time and prioritization. You can imagine spending 50% of your time, for example, on the stuff you have to do, but 50% on these things. Not that they're easy—they're not—but if we change them a little bit, they have a big impact. Another nice thing about them is they don't have to be perfect. If a small change to the positive makes a large impact, they don't have to be perfect for it to be impactful. So that's great because being perfect is hard, and maybe not even a good ROI. But changing it somewhat is a really good ROI. And so those are really interesting. Ideally, you would find these things empirically. Ideally, you'd have some incredible mathematical model of the whole company. Every important thing would be a number, which of course it's not in reality, but imagine if it were. Then you'd have this incredible model and it would all be right, with all the things pointing at all the other things and all the crazy cycles and things that happen. Then you could use sensitivity analysis or some other math to say, "Oh, look, if I change this part, then I get a big change in revenue or profit," or something like that. And then that's what they would be, and you would go focus on that. That would be smart. If you could, you would do that. But you can't. Actually, there's a couple of really special companies like Amazon where they kind of do, and it's amazing what they're able to understand and invest in. But for the 99.999% of us where that's not the case, still you know that they're there. What I found is that there's a couple of areas—you listed some of them—where very typically it's one of these special multipliers or leverage points where a little change has a big impact.

So if we say it's usually this, this is a pretty good idea. If you're wondering, "How do I change my business a lot?" these are some pretty smart areas to try, or at least try one of them. The other thing I find is that they're often hidden in plain sight, by which I mean, like you said, funnel. So let's take that marketing funnel. You think you have a handle on the marketing funnel. You say, "I know what it is. Traffic comes in the top, hopefully customers come out the bottom. There's these different steps in the funnel. I will measure the conversion rate of each step, and wherever I see opportunity, whichever step I think I can affect, maybe with A/B testing or whatever, I will do that and I'll measure it and try to do that. Maybe whichever step is the worst in terms of people falling out, that's the one I can get." And so you're like, "Look, I get it. I get a marketing funnel, there's nothing more to see here." But none of those steps I just said are the most important part of the funnel. None of those change the quality of what comes out of the funnel the most. So you look at the funnel and you think you get it, but you don't. So it's hidden in the sense that it's like, "I'm looking at this funnel, I don't get it." Here's how you can see what this hidden thing is. You said it was the quality of the stuff coming to the top of the funnel, like traffic coming to the top of the marketing funnel, the quality of applicants coming to the top of a hiring funnel, the quality of startups coming to the top of an angel investor's funnel. There's a lot of funnels in the world: product ideas that come in the top of the product, there's a lot of funnels. Why is the quality most important? I'm stating that like it's a fact, but I better justify that. Imagine in a marketing funnel if I put nothing but perfect, ideal customers into the top. Every single visitor is a perfect person who has exactly the problems that you solve, who uses the language that you use on the homepage, who already has the budget allocated that is big enough to fit your pricing page, right? Just imagine it's the perfect customer, perfect fit. What will happen is you get a lot of conversions. Even if your site isn't that good, even if your sales team—if it's leads—even if your sales team is not that capable, whatever, they'll just take the—we call them order-taking sales people—they'll just take the orders because they're already ready to go, right? So if the quality was very high, then even if every other step in the funnel was poor, you would still do really well. Good quality means like you can't help but win.

Now let's take the other end of the extreme. Let's say I put nothing but junk traffic at the top. Well, they don't want to buy. It doesn't matter how good your A/B tests are and your language. If this isn't the right people, you're not going to convert very many and you never are. No matter how good all those steps are, it's just not going to happen. So if you take a step back, what did we just learn? High quality on the top means I succeed no matter what, and low quality means I fail no matter what.

No, wait. We just said "no matter what" both times. The top-of-funnel quality determined our success or failure regardless of what the other steps did. Now, I don't mean to say the other steps are not important. I'm not saying that. What I'm doing is emphasizing how important the quality at the top is. It's the most important. That's what I'm saying—not that they're unimportant, but this is the most important. But it wasn't part of the normal funnel discussion, because the normal funnel discussion is: here's all our steps, here's all our conversion rates, how do we make it go up? But you never said the quality at the top; you're not even measuring the quality at the top probably. So this is what I mean by the quality at the top is this leverage or multiplier of power. But you didn't see it because you thought, "I understand the funnel math, leave me alone," and so you didn't see it. That's why it's hidden and a multiplier—hidden and a point of leverage. And actually, it's even worse than that in marketing, which of course we can get into. It's even worse than what I just said, which is actually quite interesting. But this general point is common across all these funnels: hiring funnels, product idea funnels, investment funnels if you're into that sort of thing. In general, this is true of funnels, and we're not usually paying attention to that.

Yeah. How do you do it? To focus, the first thing is you say, "Wow, we should improve that. We should measure it somehow." But let me give you some specific ways to do this, because it's not obvious. In fact, it's especially hard because it's when you know the least about the traffic. It's the most anonymous, you have the least data about it before it even shows up; it's not even in your analytics software, even if it's trivial. So what do you do? Here are some really specific things you can do. The first thing is you want to have the people who are about to come to the website to self-filter the useful traffic versus non-traffic before they show up. How would you do that? By having really pointed, crazy specific messaging in whatever you're bringing them into—an advertisement or social media, whatever the marketing is that you're doing to bring them in, whatever that is. Because usually advertisers are asking, "What hook is most likely to get people to click this ad?" Right? I want click-through rates to go up, so whatever makes people click the ad, that's what I want to do. That's the normal mentality. But when you think about quality, you realize I only want the right people to click the ad, which means I am not optimizing for maximum click-through rate. See, this is different. I want clicks from the right people, which means I need hooks that cause the right people to click more because I'm so specifically talking to them, and also have the bad people go, "This is not for me," and not click—that second part. And also reject the other people, repel the other people. That's not usually what you're thinking, but that's what you do if you want quality. A great example is if you're selling a product to technical software engineers, let's say: you should use crazy software engineer talk in the ad. It should use some inside joke and whatever. Software engineers will be like, "Oh, this feels like it's for me." Obviously they're not literally thinking that, but this feels like something that's for me. And other people do not think that and do not click the ad. That's what we want. We want you to go, "What the hell?" and not click. That's what we want. One of the best examples of this is Twilio. Twilio used to have these billboards that just said, "Ask your developer." That's all it says, right? So the developers are like, "Yeah, we're in charge." It's very clear what's going on there. Besides commiseration with the person, you could put things that people consider dangerous, like putting the price right there. "Won't that cause people not to click?" Yeah, if they don't have the budget. But if they don't have the budget, they shouldn't come, right? So price, or for teams of a hundred or more, you might say, "Oh, I only have a team of 10, I won't click." Or a certain vertical. There's different kinds of qualifiers you could put in there. All this advice goes under this one category of self-filtering: putting this pointed kind of a hook where you're thinking about filtering in the hook in the first place. That's just one idea. I can give you more.

Click-through rates will go down.

But what should go up? Things like conversion from the homepage to anything else, or conversion from the landing page they get to, even sales, which is what we wanted anyway. So it's not that it's impossible to measure, it's just a little trickier. And actually that's my second thing: to widen or deepen what you consider the conversion rate. What I mean is, normally again, you think of the ad as a click-through rate; that's my conversion, the higher the click-through rate the better. But again, that's not really measuring quality. If you could move that one step forward and say, "Well, I'm going to consider the ad a success or converted if they click and then also click past the landing page that they hit, then I'll count it." Well, then you'd be optimizing the ads for a better, deeper part of the funnel where they're a little bit more engaged and real. So ads that got a lot of clicks but not other engagement, i.e., bad traffic, those ads would not perform well according to your conversion. We all know a lot of these ad software have these kind of things, so really maybe this is motivation to really do it. Then the question is, how deep can you go? The ideal is it goes all the way to sale, and you ask how many sales does this ad make? That would be ideal, all the way to the end of the funnel. Now, in some cases you might be able to do that, like if it's an impulse buy and it's a one-session buy and all that, you might be able to do that. In other cases it's too hard and you can't, understood. We're just talking about a vision statement. What's the ideal? The ideal is it's measuring sales. In practice, the real question would be: how far down the funnel can I push my definition of a conversion or an ad success so that I'm measuring my ads as best as I can, with the ideal being a sale, right? Another nice thing, because you're like, "Oh my gosh, but that's going to be so few conversions," and I would say yes. But imagine how much you can pay for each one of those conversions.

Right. I normally think 50 cents a click, a dollar a click—yikes, that's a lot. Wait a minute, I can spend $50, $100, $200 per conversion, of course. So I'll have fewer of them, etc. But if I find ads which do that, I can spend crap tons of money on those ads. Multi-attribution is a problem, when buyers and users are disconnected that's a problem—there's a lot of challenges. I never said this was easy. Just saying if we can move it a little bit, though, it helps a lot. While it's not easy, it's worth it. Maybe we can only do 20% better than we're doing now, but what I'm saying is 20% better on this is massively impactful compared to other things you could move by 20%. So it's worth it, is my point. That's the second thing: pushing the definition of conversion rate or success down the funnel. That's the second thing. I got more if you want more.

Okay. The next one is filtering—I call it channel filtering. This is kind of like the self-filtering, but this is filtering using the tools of the channel. Again, you're used to thinking, "How do I get more traffic out of this channel?" But you really want less traffic that's higher quality. That's what you really want. So you imagine in keywords, rather than spreading the keywords as much as possible, it's like, well, what are the keywords that go with the highest quality traffic, and maybe I shouldn't be chasing some of those other keywords? Or keyword exclusions. I've had this in my career a lot where there's the keyword for something, but it can include a lot of other stuff. By excluding, you're just narrowing or filtering ahead of time. Filter out to keep the quality; these are all different techniques for how do I do that. Google has other kinds of audience filtering about affinities and what market—we don't need to get into that. But there's sometimes tools and stuff. Also, which channels? You mentioned LinkedIn. People go to LinkedIn and they talk about work, so if I'm selling a B2B software, LinkedIn feels like a good place, TikTok doesn't. There's other kinds of stuff, like selling a hoodie that costs $7 where it's probably TikTok and not LinkedIn. So there's filtering in the channel, but also I shouldn't be in 20 channels. I should be in the channels that make the most sense for me where the traffic will be higher quality, even though there are more views in some kind of absolute number on Instagram or something. But those aren't my views. Again, quality, not quantity. So that kind of channel filtering, whether it's which channel or within a channel, that kind of filtering is another one. I'll just give you one more of this funnel stuff, which is to segment the traffic that comes to the website by the channel. So if I come from LinkedIn, then what? If I come from LinkedIn with this particular ad campaign, then what? And so on. Again, this is not news that we should have landing pages, or that we should maybe tag things with cookies, or use something like Segment.io or whatever. I'm not trying to shill for any company, I don't care, but using these things to tag or understand the traffic so that we can be more specific. Now, that doesn't actually increase the quality of the traffic. If you go get a bad campaign and send it to a landing page, then it's bad traffic; that didn't actually help. But what it does do is it prevents other landing pages and tests from being poisoned by this bad traffic.

So if you have some bad traffic somewhere, okay, but at least it's cordoned off. Hopefully you can figure that out and try to optimize it separately, or stop, or whatever you're trying to do. It's cordoned off from the good traffic so that as you experiment—because of course you're going to experiment with ads and channels, you have to—of course sometimes you're going to pull in a lot of garbage, and you want to keep that garbage away from the not garbage. While that doesn't totally prevent garbage, you can see how that makes your processes better. That kind of keys off another point which I didn't make before: remember I said that it's worse than you think about the garbage traffic? Here's a really cool way that it's worse than you think, which this segmentation solves. Let's say you didn't do anything of what I just said. The advertising team is told, "Get more traffic." They're like, "Okay." And so they go get more traffic. Of course, most of it's garbage because the low-hanging fruit is taken; of course, it's usually worse, right? So they do, and more garbage is coming through. But you say, "That's okay, my A/B test will filter it out." Here's why that's wrong. I pull in more garbage. Now I'm looking at my A/B test, but all I'm doing is saying, "I want to convert the most amount of whatever is coming in." Yes, but that means I'm convincing the garbage because now there's more garbage. The only way for me to make a better, more successful quote-unquote A/B test in isolation is to appeal to garbage and pass more garbage along to the next step, which is the last thing I want to do. It's even more expensive if sales gets to it. It's even more expensive if the customer signs up and they're a shitty customer, and they bang us in tech support and they leave anyway and leave a bad review—that's even worse, right? The further the garbage gets, the worse and more expensive and all that.

So the garbage comes in, but the A/B testers are just like, "Oh, I got to improve this." And so it's just so bad. It's multiplied by so many people, so much expense, so bad. So that's why it's even worse than what we were saying before. It's like, oh my gosh, you realize the extent of the pain, right? This final thing, this segmentation of the traffic, helps with this issue. Because if you have a channel that's pretty high traffic, pretty good, you have A/B tests going, everything's happy. Of course, you're going to experiment with new traffic and some of that's going to be garbage. Don't put it into those funnels that are working and pollute those A/B tests and break them. Don't do that. Have it over here so that these new A/B tests are broken or whatever they are. Don't pollute it. It doesn't totally prevent garbage, but preventing it from poisoning things that are working, that is good. That is better than what you were doing before, so it is progress in a certain sense. Although segmentation by itself might be an obvious thing to state, hopefully now it's clearer that it's not just like, "Oh, we could do that because it's more personalized," which is true, but hopefully you can see: oh my gosh, the ramifications are way more than I thought. So it's not just kind of a good idea that maybe we should get around to. No, no, no. This is huge for you. This is enormous.

It is like that. Because you don't realize how those cuts ripple through everything else in the company, and therefore it's just way more costly and powerful than you figured. It wasn't part of your original funnel diagram, so maybe it was hidden altogether, much less how important it is and how much of a multiplier it is. That's also hidden.

Well, and again, the good thing to remember is—it's like, "Oh my gosh, how do I make this so much better?" It's not going to be perfect. Of course, how would it be perfect? It isn't going to be perfect, but you don't have to. If you shift it a bit, you'll help a lot. So you can roll the angst back a little and go, "Oh, I just need to improve." There's so many areas of the company I could improve a little and only get a little benefit. Here's an area I could improve a little bit and get a big benefit. Yay. Your job isn't to be perfect. You could set some sort of directional vision statement, of course, fine. But that's not your goal, that's not your commitment. If you can improve this a bit, it has an enormous impact. So that's why it's worth it.