Framework

The Growth Loop: The Five Levers Every E-Commerce Business Runs On

A simple framework for where growth actually comes from, and why the brands that win treat it as a loop instead of a checklist.

By Oliver Silen. Founder of Sentopi, former Amazon Vendor Manager and Product Manager. Published June 2026. Updated July 2026.

Most e-commerce advice is a pile of tactics with no spine. Run more ads. Drop your price. Get more reviews. Each one is true on its own and nearly useless on its own, because growth in e-commerce doesn't come from a tactic. It comes from a system.

I spent years inside Amazon as a Vendor Manager and Product Manager, helping hundreds of brands scale. The ones that grew weren't the ones with the cleverest tactic. They were the ones running the same simple loop: five levers, each feeding the next, reviewed and refined on repeat. Pull them together and they compound. Neglect one and it quietly caps the other four.

Here is the entire model on one page.

The five levers, in plain terms:

1. 1

   Operations. Keep the product in stock and shipping fast. This is the floor everything else stands on. A listing that goes out of stock doesn't just lose today's sale; on Amazon it loses rank, and rank is expensive to win back. Out-of-stock is a silent tax you pay in sales you never see: across global retail, when shoppers hit an out-of-stock, about a third buy from a different store entirely rather than wait (Gruen, Corsten & Bharadwaj, 2002). Watch: in-stock rate.

2. 2

   Pricing. Deliver clear value at a competitive price, with promotions that have a reason behind them. Price isn't about being cheapest; it's about being obviously worth it next to the alternative on the same screen. Either you're priced competitively, or you carry a USP strong enough that you don't have to compete on price at all. Price is one of the most powerful demand levers there is: a meta-analysis of 1,851 estimates found an average price elasticity near −2.6, meaning demand tends to move more than proportionally as price changes (Bijmolt et al., 2005). Watch: your price against the comparable listing.

3. 3

   Assortment. Offer the product mix your specific customer actually wants: the size they were looking for, the bundle that solves the whole problem, the variant the category leader carries and you don't. Every gap in your assortment is demand you hand to a competitor. Watch: variant coverage against the category leader.

4. 4

   Visibility. Drive awareness through content, advertising, and merchandising. On a marketplace, most of it is won or lost in search and browse, and it comes down to one thing: relevancy at the ASIN level. I ran browse and search setup at Amazon, so here's the part most sellers miss. The algorithm doesn't rank your brand; it ranks each ASIN by how likely it is to satisfy the shopper who just typed that query. You earn that relevancy by feeding the other levers: the listing has to be in stock, available to ship quickly, and either priced competitively or carrying a clear USP. It also rewards demand you bring from off Amazon: sellers widely observe that external traffic which converts is one of the strongest signals you can send Amazon's ranking algorithm (what the seller community calls A10). Visibility you buy is rented and stops the day you stop paying. Visibility you earn through relevancy compounds. Watch: organic rank and glance views.

5. 5

   Ratings. Address what customers tell you is broken, and build the fixes into the product and the listing. Ratings aren't a grade you receive at the end. They're the loop's feedback signal, the one place customers tell you, in their own words, which lever is leaking. The financial stakes are well documented: a Harvard study of Yelp found that a one-star increase in rating drove a 5 to 9 percent increase in revenue (Luca, 2016), and online reviews move sales more when they're negative than when they're positive (Chevalier & Mayzlin, 2006). It's why we read every review rather than a sample; in our listing-accuracy study of 3,725 reviews across seven brands, the complaints that dragged ratings down were consistent and fixable. It's the lever most teams treat as a scoreboard instead of an input, and it's the one we built Sentopi to run. Watch: star rating and review velocity.

Here's what "good" looks like on each lever. Treat these as directional; they vary by category.

Why is this a loop instead of a one-time setup?

Because you never finish them. This isn't a setup you complete once and walk away from. It's an operating rhythm: you review the five levers on a regular cadence, ideally weekly, make small and specific improvements for your brand, then come back and do it again.

And more reviews lift your rating, which lifts conversion again. The loop closes, and each turn starts from a higher floor than the last.

Here's the part that sounds too simple to matter: the strategy isn't the secret sauce. Five levers is not a hard idea. The entire advantage is in the execution, the discipline to actually review these metrics every week and keep making small improvements to the foundational layer of the brand. Most teams set this up once and move on. The ones that win keep turning it.

That's also why the returns look the way they do. You'll often see large step-changes early, when you fix the obvious leaks. After that it behaves less like a hyperscaling stock and more like compound interest: quiet, unglamorous, and very hard for a competitor to catch once it has been running for a few years.

It's also why retention isn't a separate lever here; it's the output. A loop that's turning produces repeat customers and reviews, and those are exactly what re-feed Visibility and Ratings. Spin the loop well and retention takes care of itself, and retention is where the compounding really shows: cutting customer defections by just 5 percent can raise profits by 25 to 85 percent (Reichheld & Sasser, 1990). A weakness in any one lever, though, caps everything downstream of it. You can't out-advertise a 3.9-star rating, and you can't out-price an out-of-stock listing.

Where this fits with the frameworks you know

If you've read about growth, you'll recognize the DNA here. This borrows most from the Amazon flywheel, Jeff Bezos's model where a better customer experience drives traffic, which draws sellers and selection, which lowers cost and improves the experience again. Both are customer-focused and both compound. The difference is altitude: the flywheel describes the marketplace, while this describes the levers a single brand actually controls.

It also sits comfortably next to the models you already use. The classic marketing funnel and the Traffic × Conversion × AOV formula are still right; they just operate at one point in the loop. Use the funnel to optimize how a single listing turns visibility into a sale. Use the loop to understand why that listing converts in the first place, and what feeds it. They aren't competing: one is a lever, the other is the system the lever sits in.

Where this framework goes further than most is in naming the levers the popular ones abstract away: whether you're in stock, what you charge, what you actually sell, and what your customers rate you. Those are the levers that most decide whether a marketplace brand grows, and they're where this one starts.

Why did your sales drop? Find which lever broke

When sales fall off a cliff, resist the urge to start changing things at random. Run one diagnostic first. Pull your sessions and your conversion rate, and compare both to the same window a month and a year ago. The split tells you which lever to open:

• Sessions dropped, conversion held: the problem is upstream in Visibility. Your rank or your traffic slipped.
• Sessions held, conversion dropped: the problem is downstream in Pricing, Assortment, or Ratings. Something on the listing stopped persuading the same traffic.
• Both dropped: start with Operations. A stockout takes down both at once.

That one fork tells you which lever to open before you waste a week on the wrong one. And know that recovery isn't instant: organic rank typically takes two to four weeks to rebuild after a stockout, longer in competitive categories. The earlier you catch the week-over-week decline, the cheaper the fix. For the step-by-step version, with the five most common causes and what to check for each, see why are my Amazon sales down.

What this looks like in practice

To make it concrete, here is what the loop looks like when it's leaking. Take a composite brand, a mid-size Amazon seller doing roughly $2M a year. The numbers are illustrative, built from public benchmarks rather than a specific client, but the shape is real.

None of these leaks is a crisis on its own. Together, they're a brand running at a fraction of its own demand. Now turn the loop instead of pushing one lever. Recover the out-of-stock sales. Add the missing variant. Put one line on the listing that justifies the price. Read the reviews, find the three repeating complaints, and fix the two that are cheap to fix. Lift conversion from 3% to 5%, still short of the benchmark, and you've moved the run-rate by roughly $1.4M without buying a single extra click. You aren't adding 5% here and 5% there. You're multiplying. That's the difference between treating these as a checklist and treating them as a loop.

Where to start

Start anywhere. The loop doesn't care where you enter it, only that you keep it turning. Pick the lever that's most obviously leaking, fix it, measure, and move to the next. Then come back around, because the bar you cleared last quarter is the floor this quarter.

Finding the leak is the actual skill. The fastest way in is to watch each lever's metric for the biggest week-over-week and year-over-year declines. A sharp WoW or YoY drop is usually the leading indicator that a lever has started leaking, and it shows up there before it shows up in revenue, which is exactly when you want to catch it.

If you want one lever to start with, start with Ratings, because it's the one that tells you about all the others. Late shipping, confusing sizing, a feature that overpromised: your customers have already written down exactly which levers are leaking. Most brands just never read all of it. Reading every review and mapping each complaint back to its root cause is how you turn the scoreboard back into a steering wheel. That's the part we automate at Sentopi.

The brands that win in e-commerce aren't the ones that nail a single tactic. They're the ones that keep the loop turning.

Common questions

Sources

Figures in this article are illustrative and drawn from public benchmarks rather than a specific client. Conversion, out-of-stock, and revenue impacts are estimates intended to show the shape of the loop rather than a precise forecast for any one brand.