Amazon’s search engine, A9, quietly sets the fate of every business selling on Amazon. Here’s four keys to winning the game.

For most people, Google is search, and search is Google. It’s a common belief that the Silicon Valley giant powers nearly every aspect of the web, but in fact, it couldn’t be farther from the truth. Inside the internet’s universe there are a handful of galaxies that are powered by their own search engines. Apple has Spotlight; Microsoft has Bing; IBM has Watson. Dozens of website use Apache Solr technology. And Amazon has the fastest growing search engine of all — A9.

Never heard of it? You’re not alone. A9 is a hidden powerhouse in the search world despite impacting every interaction you have with Amazon. Unlike to Google, which crawls the entire web to feed its ever-shifting optimization algorithm, A9 pulls directly from Amazon’s coffer of webpages and looks at a distinctly different set of criteria to determine which pages rank highest. The good news is that optimizing for Amazon’s algorithm is, in many ways, far easier than optimizing for Google’s. When it comes to determining what you see first on Amazon, the company really only has one goal: get you to buy stuff.

If you’re a company that sells goods on Amazon, this single-minded goal makes your life a little easier. But don’t get too excited about the prospect of totally beating the system; optimizing A9 is still an imperfect science that takes a lot of know-how and a healthy dose of trial and error.

Search engines keep their algorithms opaque so that you can’t game them , but if you understand the levers to pull, you’ll find that you can take an aggressive stance on optimizing for A9. Here, we’ll discuss the four main components that drive Amazon search: conversion rate, relevancy, pricing, and reviews. And we’ll give you tips on how you can start bending them to get the search results you want.

The Right Way to Product Relevancy

All search algorithms are an attempt to tie visitor intent to some kind of end product. With A9, it’s about understanding what any given person might search for on Amazon and serving that to them as efficiently as possible. Unlike Google, where website creators have no boundaries or guidelines when it comes to what kind of content they make, Amazon removes some of the guesswork by providing shop owners with a template that outlines exactly what they need to have on their page. This includes: Product Title, Key Product Features, Product Description, Product Picture file name.

When someone searches on Amazon, the algorithm scrapes the pages looking for the most relevant content to the query. The trick to relevancy is understanding keywords. A good keyword in the product title can take a generic product and turn it into something that’s highly searchable. Amazon doesn’t have its own keyword research engine (there are plenty of external to choose from), but even without a specialized tool you can ask yourself questions like, what are some of the specific attributes around the product I’m selling? What is the color? In what context might someone use the product?

Think about the last search any consumer might do to get them to Amazon — those are your keywords.

For companies trying to sell their goods on Amazon, it’s important to remember that Amazon is the last step in a consumer’s long purchasing journey. Some people might search Amazon for a generic term like “running shoes,” but more often than not, people do their early research and browsing on Google. By the time they get to Amazon, they’re ready to click buy. Most people have a set hierarchy while searching. They start broad and narrow it down until they find what they’re looking for. Think about the last search any consumer might do to get them to Amazon — those are your keywords. Instead of titling your product “Adidas Running Shoes, Size 11,” try “Adidas Running Shoes, Size 11, Blue, Track.” The great thing about Amazon is that you have absolute control over what you enter into the product page fields. If keywords aren’t registering, change them up and try again.

Finding the Pricing Sweet Spot

Search Amazon for “running shoes” and scroll down. Notice something? The prices at the top are cheap, and all are within a $10–15 band of each other. This isn’t done by accident. Amazon knows you’re more likely to convert at these prices. As you scroll down the page, the options (for the most part), get systematically more expensive.

Knowing how to set prices is a key factor in pushing your product to the top of Amazon’s search ranking. But how do you know where to set your price? Let’s go back to our running shoes search. A price that falls between the highest and lowest on the first couple rows of products is your ballpark. Staying within this band is particularly important when it comes to commoditized, easily substituted products like razors or batteries. Customers aren’t willing to pay an extra few dollars on something they have no personal connection to.

Making it easy for people to leave reviews in the first place. This comes down to communicating with your customers.

Building a Community Through Reviews

You can do all the keyword research in the world, skyrocket your relevancy, and set your prices dirt cheap, but a bad review can still kill a product. Reviews are proof that people should invest in your product, and Amazon takes them seriously when determining search rankings. The best and simplest way to garner positive reviews is to produce a good product, but the savviest companies know that it’s really about making it easy for people to leave reviews in the first place. This comes down to communicating with your customers. Set an email strategy for reaching out after a purchase, and give your customers them ample incentive to take the time to review a product.

For items to which the consumer has a personal connection — beauty products, furniture, clothing — fostering an easy review process is crucial. The goal is to build what feels like an organic community around your product. If successful, reviews can not only increase your conversion rate, but they can provide companies valuable research insight into how they can position and market their product going forward.

Test and Learn

Keep in mind that all of the previous factors — relevancy, pricing, and reviews — are in service of increasing a store’s click-to-conversion rate, or the rate at which people buy stuff. Conversion rate is the holy grail of A9 since Amazon’s algorithms are designed to turn visitors into customers. Everything you do should be an effort to increase the number of sales, which will in part increase your page’s visibility. If that feels a little like a Catch-22, you’re not wrong. In order to increase your conversion rate, you have to get all of the other factors right. And because Amazon doesn’t allow A/B testing or duplicate posting, it can be challenging to understand what’s working and what’s not. The hard truth is that there’s no playbook for getting it exactly right. If there was, everyone would set up a shop, pull the right levers, and create a monopoly selling black running shoes.

Instead, selling effectively on Amazon requires constant attention to detail. You can’t write the product description and walk away. Someone needs to be there gauging how keywords are performing, how pricing is shifting, and fostering a sense of community. Optimizing for Amazon’s algorithms might — for the time being — be simple compared to doing the same for Google, but to truly maximize your reach, it’s something you’ll constantly have to invest in and nurture.

Magenta is a publication of Huge.