Make Money Online

So draft is the agency, draft revise is the service. And draft is I basically call it a consultancy. And I do interaction design for a living. Interaction design is the process of making something easier to use, more effective at communicating, and generally better to complete tasks on. That can be a lot of things, right? That involves fixing the usability of something. That involves interviewing customers and seeing what their motivations are, making sure you’re making something that they actually want to buy. It involves fixing some of the graphic design, involves fixing some of the functionality, involves addressing a lot of edge cases, like what happens when I paste Copied JPEG from a text editor into the credit card field. What happens? Oh, gosh. Where is your God now? Right? I do that sort of stuff. So it’s kind of a QA-ing of the product. It’s a lot of things, right? And Draft has had that under its umbrella for the entire five years that it’s existed. For the past three years, I’ve been running a service called DraftRevise, which basically acts as a A B testing, conversion rate optimization type service, and I operate on a retainer basis. What is A B testing? It is the process of you have control, the existing website, and you want to test a change and see what happens. So you run that change for about 50% of your customers, and you measure the success of the change. Is a way of being very careful and deliberate about your design process. It allows you to measure the impact of a design change in a way that is far more careful, less cavalier than if you just made the change and assessed the step function on Google Analytics, which who does that in practice? Nobody does that. It’s a lot easier to execute. So you have an A-B testing framework that’s usually a WYSIWYG editor, or you can redirect people to a specific page if it’s a more ambitious change. And that’s what I’ve been doing for the past three years. It’s been really interesting. I’ve changed the service in a lot of ways. But overall, that’s kind of the core conceit of it.

So I’ve been doing a more and more involved research process over the past few years. It started out with DraftRevise, where I would basically go in and do what’s called a heuristic evaluation, where I had a bunch of different metrics that I would go and vet your website against. And That was done in a vacuum, right? It was done, I’m an expert, I know what’s going on, but I’m still not matching it to your customer needs, which was bad. It didn’t actually work out well. I ended up beating the industry-wide success rate for A-B testing, but only by a little. So it was 12% is the industry-wide success rate for A-B tests. And I was at like 18% or 19%, something like that. And that’s fine. I mean, you’re paying me to a 50% increase is cool, but that’s still kind of dismal in the grand scheme, right? Like you’re going to be dealing with more failure than success. which sucks. So more and more I’ve been going off like the heat maps that come in as I run A B tests and seeing like, oh, where are people actually like working on the site. Oh, they’re not clicking at all here. So why are we A-B testing it? Right? Like they’re just going to ignore the new thing too. How far are they scrolling down the page? Oh, nobody’s even scrolling there. Okay, well, that’s a problem. Where are they being lost on the page? Oh, they’re being lost up here. Now we need to change this element. And then we find the core thing that needs to be changed. I’ve started looking more and more at Google Analytics because everyone has a horribly broken Google Analytics install. And so just trying to fix that and deal with the Byzantine horror that is that Stupid web application that I’m very grateful is being offered for free. You get what you pay for. And Also, talking to customers. I get people on the phone. I have them write surveys. I have them provide support inquiries. And I listen to your support team and figure out what’s going on that’s wrong with your product and do everything that I can to Capture revenue, reduced churn, and make it so that people are actually buying and getting the benefits of your product.

Yeah, I mean, one thing that I used to actually recommend, which I no longer recommend, was, oh, A-B testing can settle a debate. And so Carl over here wants the button to be blue, and Wendy wants it to be green. And they don’t have any reason for it. They just think that blue and green, one of them, is going to generate more revenue. Okay. Let’s run an A-B test. Blue and green. Well, the button was red, so we have to have the control in there also. So several outcomes can happen, right? Red can win. Blue and green can do dismally. Okay, well then everyone is unhappy. Blue wins. All of a sudden, Carl looks like the better designer. And then you’re dealing with internal politics that are actually really toxic and problematic. What happens if nothing wins and it’s completely inconclusive? Everybody throws their hands up and blames me. Right. There’s no positive outcome, even if you have a revenue generating outcome from an A-B test that exists to settle a debate. So, what I do, I always, always, always justify it with research. I’m saying, like, oh, people are already hitting this button. Like, a lot of people are already hitting this button. Maybe we need to look further afield and figure out ways to get more people to hit the button. Maybe there are a ton of outbound links in the header that other people are hitting that we don’t want there. That could actually be potentially problematic. And so let’s look more deeply at this and not focus on the stupid color of the stupid button because you read something from a blog from five years ago that said that tested, interestingly. So when people come in with an agenda for testing, I’m like, great. So I’m going to take this, I’m going to look at it, and then I’m going to print it out, and I’m going to go on a Skype call with you, and I’m going to Set it on

I would I would uh fricasy it. No, um it’s such a funny word. Uh so I no, it was I would actually take your opinion seriously, right? I’d be like, oh, yeah, we could test that. That’s an interesting idea. But now it’s like, okay, well, that has to go through the research process. So you have this idea. How are you justifying it? And sometimes they do justify it, and then we’re good, right? But if I have to come in as a consultant and justify every design decision to you, it goes both ways. You don’t get a free pass. I know it’s your business, but you don’t get a free pass. And the reason for that is because both of us are human and we’re both experts and we both know what we’re doing. I come to you with respect and integrity and I inspect expect you to do the same. And so, you know Another thing that matters significantly in the design process, I come at this from a design background. I am not a marketing guy or growth hacker at all. I believe that you are making design decisions when you’re A-B testing, which you are. Changes to the copy are design decisions. They affect the way that your site communicates. Changes to the coloration, obviously design decisions. Changes to the layout, also obviously design decisions. So when you’re coming to me, we’re doing a design process. And I do everything in my power to convey that DraftRevise is a design process and that you’re buying design and it’s in the guise of revenue generating conversion rate optimization, right? I found that works. It took three years to settle on a process that works, and I’m very, very confident about it right now. So, one thing about the design process, the reason I’m spending this whole time harping on this, when you’re in a design critique, it doesn’t matter who ventures the winning idea. It doesn’t matter who critiques the idea in the best way. It’s not a contest. You need to everybody needs to work in the service of the design. There’s another person in the room, and you’re there to serve them, and that person is the design. Okay, I’m not serving the client, I’m serving the design. The design serves the customer, the customer serves the client. And I know that’s a long way to get to that point, but it generates business value and it results in a better outcome for everybody. It also takes your ego out of the process, right? So you’re not just fighting with the designer about matters of taste. And you’re taking it away from like small trivial stuff. So, one thing I love saying, the logo is purple. I hate purple. It’s my least favorite color. I’m like, well, I hate purple too. But does purple fulfill the business goals? And then we have a conversation about the business goals, right? And I’ve immediately just pulled you over to my side of the table. And we’re both attacking purple from a business standpoint. And then it turns out, oh, we have a premium offering. Purple has these royal connotations. Oh, it’s kind of like. A feminine color. Oh, it’s like more of a luxury product. Oh, great. It’s like a beauty product. Now we can use purple. And you’re just going to grin and bear it every time you drop purple on that website, right? Like. And it’s I’m just completely spitballing an example here, but that that’s an example of some something where like, you know I don’t care whether or not the color is purple, only in as much as it fulfills the business goals.

I mean, I talk to the customer. I’m like, who are you? How old are you? Where do you live? I get demographic maps. When do you shower? When do you shower? Where were you on the night of January the 16th? Great user experience interview question. No. I do what’s basically a jobs-to-be-done interview. And my colleague Eric White actually wrote a really great thing about this recently that I saw at a conference that I went to in Stockholm with you. Basically, you’re sitting down and you’re asking, well, why did you buy this? Right? Okay. You needed it. Okay. Well, why did you need it? Okay, great. How did you determine to buy this one? I’ve vetted a bunch of competitors. Okay, which competitors? Okay, why did you settle on this one? Well, because it was the prettiest. Okay, well, what what about it is pretty? Okay, this. And then all of a sudden, you have this giant map of the customer’s motivations, right? And you’re sitting there and I’m just asking questions and shutting up. Like I’m not saying, oh, well You know, you’re a, I don’t know, a single mom in New York City, and you just bought this, and you completely love it. And it helps your life in New York City so much. And you really love it as a parent, and it’s really great. Everybody else thought that people in trucker hats in Alabama really like this. It’s like, oh, crap. You know. Why were 85% of your hits coming from Manhattan? You know, like, right? Like. So you get this all the time. And I try and provide, I mean, this is a classic UX research thing. You provide a persona, right? You say, okay, well, here’s Jane. She’s basically. You know, as if Sex in the City were a documentary, and these are her motivations, and these are what she wants, and this is what, you know, et cetera, et cetera. And, um, You end up getting a lot of insights that are really, really valuable as far as how to communicate, right? And how to change the design, right? So maybe it should be purple at that point. Maybe it should be like peach or pink or I don’t know. I hate changing something to pink because feminine, but consider alternate things that aren’t the like red and black and the impact font. That you’re using to pitch to Joe Q public in Alabama, that’s not landing, right? This is the drastic example, but you see the point.

Everyone is just on track, right?

Yes. So the businesses I initially created DraftRevise for SaaS businesses. And there’s still a plurality of my customers, but I found a lot of people come in that aren’t that. For two years, I worked with a tech blog called The Wirecutter. Since DraftRevise’s inception, I’ve worked with e-commerce sites a lot. And I found that e-commerce gets more traffic and more sales volume, especially like lower involvement ones that have really high sales transactions. you need at least 500 discrete transactions per month in order to justify having any tests on your website at all. And that is a high number for especially SaaS businesses. A lot of places don’t get that kind of volume, and it doesn’t actually work well for them. So what ends up happening is I just have a lot of like qualifications in my application process that are like, how many sales do you make? How much traffic do you make? Okay, great. You want a one-off teardown. And I can still do design and optimize your funnel a little bit by saying, okay, well. this one browser actually is doing horribly, maybe you should fix it. Bugs are a huge enemy of conversion, and you don’t do enough browser testing, I guarantee you. there’s probably something on mobile that’s failing horribly for you also. So I can outline that and do pretty well by it. But most of the time I’m working with like higher volume SaaS and e-commerce businesses. And they know they need to be testing. They know they’re actively losing money by not doing this. And they don’t know where to start. Hi. I’m here to help you. The number one reason people quit actually and end the engagement is because I’ve taught them everything they need to know, and they just want to do testing in-house. And when that happens, I’m thrilled. But then three months later, I email you, I’m like, How’s your testing practice going? And they’re like, it’s a disaster.

No, um, I started designing. I did like wire Frames and like UX strategy and that sort of thing. It was a typical design consultancy. It felt like a Agency, I would do one-off projects, and I actually created DraftRevise because I was wondering what the Venn diagram overlap was of UX work and retainer-based engagements. And I came up with A-B testing. Seemed like a good idea. Launched it on a whim. It blew up. I was not expecting that. And then I hiked the price eight times. Well done. Charger. And now people are taken aback by the price. But I tell you, man, I closed two in the past couple of weeks. Congratulations. Thank you. Thank you. So I feel really good about where we’re at right now and other ways it’s evolved. The role of research has become not just prominent, but essential. And it used to be not really present. It was something that we would do as kind of a hunch. But I was stabbing in the dark too much, and now I’m not doing that. Another thing that happens, I contractually require that revenue-generating variants have to be rolled out. You can’t tell me no. And that’s fun. Another thing that happens is actually, you can’t tell your customers no. That’s really what it is. You can’t tell your business no. Another thing I do, I have put in more and more controls on what it qualifies a good client. So you have to have at least actually one thousand revenue generating conversions every month. And if you have less than a thousand, under every circumstance but one, I will like aggressively nope you and send you off to get a one-off teardown, which is called Revise Express. That one circumstance is because I use your product and love it and know your CEO and have drank him under the table. Then I’ll get you on a call and tell you no and why. But you have to be making more than that to even get me on the phone. Point because I’m too busy working with other stuff and talking with Kai Davis and having a great time and loving my job. The design Has never changed. The role of design has never changed. It has only become more prominent. I’ve only beaten my chest more about the role of design in A-B testing. And not enough people do it. They’re like, let’s change your headline. It’s like, why? Let’s change your value prop. It’s like, why? The only people I know that do A-B tests and don’t have a capital D design process that I respect tremendously are probably Joanna Weeby at Cubby Hackers. Because she is a copywriter, right? And a bit by trade at least. And then she vets her copy using A-B tests, which is fine. But that is also a design decision. And I don’t think she recognizes well, I think she recognizes it’s a design decision, but I don’t think she she says that it’s a design decis Publicly I think she would readily recognize it’s a design decision. But yeah, it’s I I make design decisions by measuring twice and cutting once. That’s a beautiful way to put it. Yeah. Thank you. I’m not a move fast and break things type person.

Yeah, you’re going to work out at one point. You ended up closing that deal. Too, right? I did. That’s dope. Hell yeah.

I’m like a sloth.

I you know. They persist. It’s high. It’s high.

Uh that is n as in nick, u as in the letter u. Usability. Usability Unctuousness. Now, draft. neu.

That’s if you want to learn everything that I know about AB testing, go to draft. nu slash revise/slash weekly. And every week you will get a 1,000-word plus actionable lesson about A-B testing. Kai is a paid subscriber. What are your thoughts on Revised Weekly, Kai Davis? Give me a testimonial right now.

Forward a couple of my lessons too. I mean, just plain text. They’re eminently piratable. You’re probably on the pirate bay or something.

Go to yahoo. com for that. If you want to download all of Revised Weekly for free, go to thepiratebay. org. If you want to actually pay me money for it, go to draft. mu slash revise slash weekly, and I promise you I will give you a high five in person if we ever cross paths.

Look at the elbow. That’s how you have a good high five. I know it sounds counterintuitive. Try it.

I have a book. About interaction design called Cadence and Slang. It’s widely considered the best in my field. If you go to cadence. cc. If you want to actually think you’re a good fit for draft revise, take a look at draft. nu/slash revise. It is a fantastic way to begin the conversation. If you just throw your email address at the bottom there and fill out the application, I would love to hear from you if you’re a business that thinks you may benefit from A-B testing. and the kind of research-driven design work that I do. And yeah, that’s it.