The last chapter showed that the timing by which you deliver your digital memes matters as much to your success as the content itself. Sequence amplifies substance, so to speak.

So let’s explore the power of sequence and timing a little further.

Some of the most important digital interfaces you’ll ever build for us are (a) the initial registration and download sequence for your app, and (b) the checkout sequence. Sitting in your garage with your starry-eyed developer friends, you probably won’t think of these at first, concentrating instead on the core part of your meme that fulfills our needs or is fun or interesting. But at some point, you’ll need to monetize, and that means building a UI flow to enable us to register (thus setting up an attentional channel) and/or pay you money (checkout). These moments of truth are called conversion flows, because they convert us from one state to another, as from an unregistered to a registered user, or from a free to a paying user. Here is an example of a conversion flow once seen when we were subscribing to The New York Timesonline (Figure 16-1). As you can see, conversion flows almost always have multiple steps—at minimum, one screen to convince us to join, and another to confirm our identity and perhaps take our payment information.

Figure 16-1. A conversion flow used to register visitors to The New York Times online content. Step (a) mentions free services, (b) lists the value propositions, (c) establishes identity, and (d) is a confirmation.

The sequence of these screens and the textual copy you write on them is crucial to ensuring we finish. Screw this up and you’ll turn us away right at the moment we’re opening our brains and wallets to you.

You can anticipate getting into a debate with your business partners, perhaps the first real debate of your fledgling startup: one of you will notice that people drop out with every step, something that is common to all conversion flows, so he or she will logically feel that the fewer steps the better. But another partner, usually with a marketing background, will argue you need to woo people a bit, prove the value, and retain some extra steps or screen space for that purpose. Ultimately, it is a good question worth debating: why don’t you just ask for our email address and credit card number on the first screen?

If the debate gets heated, don’t feel bad. The same argument rages at internet giants from Amazon to WhatsApp . They pour huge sums of money into “conversion rate optimization,” the professional discipline tasked with tweaking these screens and testing whether doing so increases the ratio of how many of us complete the flow divided by the number who start it. This is a well-paying job you might consider for yourself, since you will be a trained memetic engineer by the end of this book.

No doubt, this same debate took place in 2001 at Classmates.com, the first profitable social networking site dedicated to high school reconnection that we mentioned in Chapter 9.

Part of Classmates’ success, as MarketingSherpa learned in a 2001 interview with then-CEO Michael Schutzler, was that “Every single element of the Classmates site and landing pages that you can possibly imagine has been rigorously tested and retested, including colors, exact wording, typeface, graphics… number of registration steps … and even how the steps in the registration process are numbered (or not.)” ⁱ

Knowing that, it may be somewhat mysterious to you why the initial registration flow for Classmates.comhad a ghastly number of seemingly unnecessary steps (Figure 16-2). Let’s count them:

1. Choose your state where you graduated from high school.

2. Choose the first letter of the city where you went to high school.

3. Is your city one of these? Choose it.

4. Choose the first letter of your high school.

5. Is your high school one of these? Choose it.

6. Wow! X thousand people from your high school are registered on Classmates.com! Fill in this form to join them!

7. Click the link in your email to confirm.

8. You’re confirmed! Want to be a (paying) Gold member…here’s why you should! (The main value proposition for many years was to be able to email others via the site.) Choose your plan (a month, a year, two years).

9. Great! What’s your credit card info?

10. Confirmed! What are your email preferences?

11. Super! Fill out your profile!

Figure 16-2. Classmates.com circa 2010.

Choose the first letter of your high school? Why didn’t they just put in a search field? Better yet, with autocomplete?

Math quiz (with fake data): Let’s say only 5% dropped out at each step of this flow (which would actually be awesomely low), then out of every 100 who began it, how many would join? How many would pay? How many would make it all the way and complete a profile? Answer: 63 would make it to step eight, 56 to step ten, and 54 through step eleven. Needless to say, Classmates wanted more of us to get through.

And by 2003 or so, autocompleting search fields were becoming a reality. Around that time a coding technology called AJAX (asynchronous JavaScript and XML ) had emerged that let you directly call server-side objects like pages and web services. AJAX let you get rid of many submit buttons in a conversion flow, and also eliminate the long, blank page refreshes of traditional server calls. Classmates’ developers were by no means wrong to suspect these pauses were annoying to us in our task-positive effort to register, and that a certain percentage of us was churning out with each one.

So on firm logical grounds, and using AJAX to autocomplete our entries and remove several submit buttons, they ostensibly tried all sorts of experiments. Imagine one variation could compress the first five steps of the flow to one step, like this:

1. Enter the first few characters of your high school into this search box.

2. Wow! X thousand people from your high school are registered on Classmates.com! Fill in this form to join them!

3. Click the link in your email to confirm.

4. You’re confirmed! Want to be a (paying) Gold member…here’s why you should! (The main value proposition for many years was to be able to email others through the site.) Choose your plan (a month, a year, two years).

5. Great! What’s your credit card info?

6. Confirmed! What are your email preferences?

7. Super! Fill out your profile!

Recalculating with only 11 steps not 7 (assuming the constant 5% dropout rate), now out of every 100 who started, 81 should join, 74 should pay, and 70 should complete a profile. That would indeed have been much better.

But that’s exactly what didn’t happen. To their bewilderment, the number or us who finished the flow actually got worse when it got shorter. As MarketingSherpa put it, “Nothing’s beaten the control so far.” How do we know for sure? Go to http://classmates.com right now and look how their registration is designed (Figure 16-3). Despite testing “every single element of the Classmates site,” it has not changed from the 1998 design. By 2003, Schutzler must have given up testing alternatives when he said, “very few companies have ever been successful by going straight for the subscription.” ⁱⁱ

Figure 16-3. Classmates.com circa 2017.

The point of this story is not to teach you to fear failure, nor to discourage you from multivariate testing, which every venture should implement from day one in the garage. Optimizers only roll out their test variations to a fraction of their traffic, so you can easily turn off the test and go back to the old flow. No harm done. But what psychological process could possibly explain why a longer flow would be better than a shorter one?

The reason is escalating commitment. ⁱⁱⁱ This is a pattern of behavior, observed by psychologists at least back to the early 1960s, in which a series of several minor behavioral commitments and small reinforcements can produce behavioral outcomes that we would otherwise refuse to perform if we are asked all at once.

When we visited Classmates, the site at first asked us merely to click our state. Then the first letter of our city. We thought what could be the harm of that? These were minor commitments that resulted in small reinforcements: it displayed our hometown. This was itself rewarding; we essentially thought well, at least they have my town in their database. Many of us were already smiling. ^iv Then the flow asked for the letter of our high school, displaying that result. Oh, it has my high school. By this point, the site had only asked for three clicks, completely anonymously, and the results appealed squarely to our sense of self-processing(from Chapter 10). Then, learning how many of our schoolmates had joined was yet another reward, and also triggered our belongingnessneeds (from Chapter 13). By the point were we asked to provide an email address and trade anonymity and attentional demands for further content, we were ready to do so.

In most alternative flows that Classmates might have tested, there was likely to be less of a slow give-and-get leading up to that trade. The shortest logical flow would abruptly ask us to cough up an email address on screen one. But ultimately, even with the churn caused by so many clicks in the longer flows, far fewer of us would. Shorter flows may have been more efficient, but they were less effective. It was like going on a date where someone asked us to get married even before they opened the car door for us.

The point here is regardless of whether you are an intern or a C-suite executive at your venture, take screenshots of your conversion flows right now. Then study them. Don’t count how many steps there are for people to drop out. Count how many small commitments you ask and minor rewards you give, leading up to the moment of truth. If you don’t see two or three in that dance, add more screens rather than take any away. Then check with your optimizers to see if your business indicators get better.

Unfortunately, we must tell the darker rest of this story as a cautionary tale. Once we had completed Classmates’ conversion flow to the point of paying with a credit card for full privileges on the site, we were in a vulnerable state, willing to agree to things that we would normally refuse and failing to scrutinize information that we would normally resist. Escalating commitment had predisposed us to readily click any Submit button appearing late in these flows. (The same tendency leads retail stores to pack additional products near the lines to the cash registers for our so-called “impulse purchases”.)

With this temptation, Classmates joined with “outside marketing partners ,” according to a statement issued after a multi-million dollar settlement in 2015, “which offered enrollment in membership programs through ads displayed during or following certain online transactions”. ^v Although Classmates had already voluntarily ceased the practice and admitted no wrongdoing, the courts made them agree to stop misrepresenting the reason why they were requesting account information right after registration, and making it clear when they were transferring customers to a third-party site to buy something more.

So if you’re a meme-maker, don’t be evil (as goes the oft-bent unofficial tagline of Google). And if you’re a user, don’t be fooled. Business people need to sell only what they say, and we need to buy only what we came for. In those moments where a series of minor gives-and-gets convinced us to open our minds and hearts to a meme-to dilate our protective bottlenecks as it were-we must quickly constrict them again when we have obtained what we came for. In practice this means carefully reading all “confirmation” screens and especially scrutinizing any further Submit buttons that appear.

In the next chapter, we examine that final step in a conversion flow even closer, and look at what it reveals about the psychology of giving and getting, offering and receiving, approaching and avoiding.