Invisible Asymptotes in Growth Curves

How Amazon found an early growth ceiling and solved it with Prime

As you all know I’m fascinated with patterns. So much so that I wrote a book on applying software engineering design patterns to SQL. What fascinates me even more are data analysis patterns because they apply directly to solving business problems.

Someone recently posted on LinkedIn that there’s only a handful of analytics patterns (e.g. aggregation, regression, etc) I find that incredibly reductive. It’s like saying that finance is nothing but algebra. Sure that’s how you do the work but that’s not where the magic lies. I now realize the post was probably ragebait, designed to tick people off and generate engagement, but it did lead to this post so maybe useful?

One such pattern is called “invisible asymptotes” and it led to the creation of Amazon Prime. An asymptote happens when the growth of a company slows down and flattens. It’s like driving uphill for a while then hitting the top of the hill where things become flat again. While that’s desirable when driving, for companies what you want is continuous growth.

This post is a summary a really long article with the same title by Eugene Wei.

Every successful business goes through the famous S-curve, and most companies, and their investors, spend a lot of time looking for that inflection point towards hockey-stick growth. But just as important, and perhaps less well studied, is that unhappy point later in the S-curve, when you hit a shoulder and experience a flattening of growth.

The invisible asymptote is a forecasted growth slowdown at some point in the future assuming nothing changes. It’s like a sign in the road telling you that you’re about to climb up a hill. It’s incredibly useful because you can do something about it before it happens.

Eugene Wei’s first job at Amazon was as an analyst in strategic planning. This is usually a function of financial planning and analysis (FP&A) in most companies. He built and maintained forecasts for different time horizons so they could be used to provide guidance on quarterly investor calls. He was specifically focused on finding these growth asymptotes.

For me, in strategic planning, the question in building my forecast was to flush out what I call the invisible asymptote: a ceiling that our growth curve would bump its head against if we continued down our current path.

For a publicly traded company like Amazon the stock price is heavily dependent on forecasted growth so discovering potential bumps in the road is crucial. The power of this pattern is that once you discover a potential growth slowdown you can do something about it ahead of time. As Eugene writes, many companies find out way too late, if they find out at all.

For so many startups and even larger tech incumbents, the point at which they hit the shoulder in the S-curve is a mystery, and I suspect the failure to see it occurs much earlier. The good thing is that identifying the enemy sooner allows you to address it.

Early on Amazon sold only books. And as an e-commerce business one of the key factors of growth (aside from acquiring new shoppers) is repeat shoppers. When you get the same customers to make multiple purchases you can keep growing without spending additional marketing dollars on acquisition. Amazon’s problem was that people were just not ordering as frequently, and as Eugene forecasted, that would lead to growth tapering off. So they decided to do something about it.

We had two ways we were able to flush out this enemy. For people who did shop with us, we had, for some time, a pop-up survey that would appear right after you'd placed your order, at the end of the shopping cart process. It was a single question, asking why you didn't purchase more often from Amazon. For people who'd never shopped with Amazon, we had a third party firm conduct a market research survey where we'd ask those people why they did not shop from Amazon.

Eventually they found it! A single factor that if not addressed would definitely cause a growth asymptote.

Both converged, without any ambiguity, on one factor. You don't even need to rewind to that time to remember what that factor is because I suspect it's the same asymptote governing e-commerce and many other related businesses today.

Shipping fees.

People hate paying for shipping. They despise it. It may sound banal, even self-evident, but understanding that was, I'm convinced, so critical to much of how we unlocked growth at Amazon over the years.

It took Amazon years and a few tries to finally figure out the solution but the important part was that it was discovered early. They created Amazon Prime (which I’ve been a member of for years) and exploded past that initial growth barrier. I’m sure they have found others since then.

That brings us to Amazon Prime. This is a good time to point out that shipping physical goods isn't free. Again, self-evident, but it meant that modeling Amazon Prime could lead to widely diverging financial outcomes depending on what you thought it would do to the demand curve and average order composition. To his credit, Jeff [Bezos] decided to forego testing and just go for it.

Finding growth asymptotes

What I like about this pattern is the idea of an early warning signal, the discovery of a critical growth impediment in a company’s journey. It’s not your typical engineering problem, it’s not a pipeline, it’s not about Python or SQL. It’s not a fancy data model or a dashboard. It’s an actually useful analytics pattern that almost nobody talks about, which if taken seriously has tremendous impact.

So how do you find these growth asymptotes? Where do you look?

One of the first tasks the analytics team should do is produce a replica of the business model with metrics. I call this the “metrics tree” and have written a lot about it. Once you have it, you can now use it to find potential growth levers and forecast growth asymptotes.

I was talking to a friend recently who runs the FP&A team at a tech company. He had built a simple metrics tree that broke down revenue into several components and had found a key factor in growth as “frequency of transactions.”

What struck me about this discovery was that frequency of transactions is a direct influencer of growth slowdowns. Assuming the company can serve their full TAM, at some point the growth will slow down, but if you increase the frequency of transactions per user, growth will keep going. Figuring out how to influence this factor is going to take more digging but at least now you know what to focus on.

That’s it for this week. If you enjoyed this post, please like it and share it. If you’re brave you can also comment publicly or reply to me privately.

Here’s the link to the post in case you want to read the whole thing.

Until next time.