Survivorship Bias Is Quietly Wrecking Your Amazon Decisions.
This is a practical follow-up to a thesis I've written about before — Amazon Isn't Lying to You. It's Just Not Telling You the Whole Story — so if you want the deeper why behind this, start there.
Here's a mistake so common it has a name, and once you know the name you'll start catching yourself making it every week.
Think about how people study success. Someone reads five biographies of billionaire founders, notices they all dropped out of university, and concludes that dropping out is the path to riches. It's rubbish reasoning, and the reason it's rubbish is invisible: for every dropout billionaire, there are hundreds of thousands of dropouts who didn't make it — and nobody writes their biographies. The study only ever looks at the ones who succeeded, and draws confident lessons from a sample that has quietly deleted every failure. The winners are visible; the far larger pile of identical-looking losers is not.
That's survivorship bias — drawing conclusions only from the things that made it through, while the failures, invisible and uncounted, are exactly where the real lesson lives. And it quietly wrecks Amazon decisions every single day.
Your dashboard only shows you the survivors
Think about what your Amazon data actually is. It's a record of what happened. The searches you showed up in. The clicks you won. The keywords that converted. The campaigns that worked. Every number on your dashboard is a survivor — a thing that made it all the way to being measured.
And just like the person studying only successful founders, the natural instinct is to study the survivors and reinforce them. This keyword converted well, so pour more budget in. This campaign worked, so scale it. This product sells, so double down. It feels rigorous — you're "following the data." But you're only ever learning from the winners, because the winners are the only things your data contains.
What you can't see is everything that never survived to become a data point. The searches where you never appeared, so there's no impression to count. The customers who wanted something like your product but were shown a competitor and never knew you existed. The demand in an adjacent niche you've never entered, which shows up nowhere in your reports because you were never in it. These are your dropouts who never got a biography — the failures the data silently deleted — and they're where your growth almost always is.
Why "follow the data" is half an instruction
I want to be careful here, because "follow the data, look at the funnel, see where people drop out and fix it" is advice I give constantly, and it's correct. You absolutely do need to do that. Watching your funnel and fixing the leaks is real, necessary work.
But it's only half the job, and survivorship bias is what reveals the other half. Following the funnel optimises the survivors — it makes your existing, visible flow better. It does nothing about the demand you're not capturing, because that demand never entered your funnel to be analysed. So a seller who only follows the visible data can optimise themselves into a beautifully efficient version of a business that's quietly missing most of its opportunity. Everything on the dashboard looks great. The growth that isn't on the dashboard goes to someone else.
The complete instruction is: follow the data and keep asking what the data isn't showing you. Where might there be demand I'm not in? What would I expect to see if there were an opportunity I'm currently blind to? Which of my "the data says this is fine" conclusions are really just "the survivors look fine"?
Training yourself to see what isn't there
You can build this into a habit. A few questions that force you off the survivors and toward the failures the data deleted:
"What's not here that should be?" When you look at your winning keywords, ask what adjacent terms, use-cases, or customer types are absent from your data — not underperforming, but simply not present. Absence is invisible unless you go looking for it deliberately.
"Who bought the alternative?" For every customer who found and chose you, others in that search wanted the category and went elsewhere. What did they want that you didn't offer, or didn't show them? The competitor's success is a map of demand you're not serving.
"Am I scaling this because it's the biggest opportunity, or because it's the one I can see?" The winning campaign is visible and comfortable to scale. The unentered market is invisible and uncomfortable to test. Survivorship bias makes the visible-but-smaller opportunity feel safer than the invisible-but-larger one. Notice when you're choosing comfort over size.
The operators who consistently outgrow their competitors aren't the ones with better dashboards. They're the ones who remember that the dashboard only shows the survivors, and who spend real effort thinking about the demand that never made it into the data — because that's where the opportunity has been hiding the whole time, precisely because it's the part nobody can see.
FAQ
What is survivorship bias, and how does it show up in Amazon selling?
It's drawing conclusions only from what succeeded, while ignoring the much bigger pile of failures nobody counted. On Amazon, your dashboard only shows the searches you won, the clicks that converted, the campaigns that worked — every number on it is a survivor. It says nothing about the customers who wanted something like your product and got shown a competitor instead, because that never became a data point in the first place.
Why does "just follow the data" still leave growth on the table?
Because following the data only optimizes what's already visible. Watching your funnel and fixing the leaks is real work and worth doing, but it only ever improves the flow you can already see — it does nothing for the demand that never entered your funnel to begin with. You can end up with a beautifully efficient dashboard while a competitor quietly picks up the growth that was never on it.
How do you actually catch survivorship bias before it costs you?
Ask what's missing, not just what's underperforming. Look at your winning keywords and ask which adjacent terms or customer types simply aren't there. Look at who bought from a competitor instead of you and ask what they wanted that you didn't show them. And before you scale a winning campaign, ask honestly whether you're scaling it because it's the biggest opportunity, or just because it's the one you happened to see.
About the author
Zamir Cajee is co-founder of This Way Up, a UK business specialising in Amazon marketplace strategy, and co-host of The Upside Podcast, where he and the team break down how Amazon actually works — and how it lies to you. Zamir has built multi-million dollar businesses from scratch and has been selling into the EU since 2016.
Ready to take your Amazon Store to the next level? Start by exploring the tools available to create a standout Storefront or make the process even easier by reaching out to us for tailored guidance.
Subscribe to the This Way Up newsletter for impactful strategies and insights, or tune in to The Upside podcast for expert advice on successfully managing your e-commerce efforts.