How Netflix and Amazon Use Data for Growth (and What A/B Testing Can Do for Your Business)
Despite the ever-changing landscape of the internet, Netflix, Amazon, and YouTube are still going strong. But what’s the secret behind their consistent growth rate? Optimization.
These companies have found ways to ensure they are always relevant. Their audience is never bored because these bigger-than-life companies regularly come up with innovative tactics to keep themselves in front of internet users.
For instance, Amazon comes up with impressive product detail page (PDP) descriptions and suggests additional purchase ideas to sell more products. Netflix’s A/B testing allows them to make additional recommendations, which usually provides other movies and shows we actually want to watch. YouTube uses a similar system to keep users engaged.
They all nail it every single time, as if they have some kind of magical power. And in a way, they do: They use analytical data to optimize their sites, often utilizing A/B testing.
Companies such as these make sure that whatever they are coming up with has the potential to hold on to their audience or boost sales along with that hold.
This is what optimization is all about.
Take for instance the following graph, which, despite all their competition, shows that Netflix and YouTube are doing far better than any other streaming sites — and this is largely due to optimization.
Image source: emarketer
You’re probably reading this because you want to know how you can boost your conversions the way Netflix, YouTube, and Amazon do. But you know trying to copy their methods won’t work for you, because:
- Your business falls in a different category
- The optimization factors which apply to them may not apply to you
- Their optimization data are different than your data.
You’ve still observed the pattern: All these digital leaders with vast audiences take data-driven actions. They take criticism into consideration and try to keep their audience happy.
You can do this too, but on terms that apply to your business. And here’s how.
Netflix’s A/B Testing: Using Analytical Data to Its Fullest Extent
The motive of analytics is to provide companies knowledge about their customers. It helps you define your targeted audience.
Netflix’s recommendation engine suggests other shows and movies which “You may like.” They have an entire data system for each individual, storing what you’ve watched and making specific recommendations to keep you engaged with their platform.
They didn’t always use this model, but as their business grew, so did their strategy for increasing engagement. And for any business, any step this huge could be a risk, especially when customers may feel like their watch history is being stored and used to keep them engaged. But because the end result has been useful to customers, the system has been successful.
According to The Netflix Tech Blog, they succeed because of Netflix’s A/B testing. This involves dividing the audience into two halves and showing each group different versions to see which one works better to accomplish a set goal.
Image source: themarysue
A Similar Approach From YouTube
You may have noticed recent updates to YouTube as well. YouTube is working to optimize their platform using personalization.
They use data as a basis with which to engage you with videos they know you are likely interested in. Now, they feature recommended topics at the top of their home page. This makes their platform that much more user-friendly and therefore more likely to keep you engaged with it longer.
This is possible because of a recommendation engine, and you can have one on your site, too. For more information, see our related guide: Recommendation Engine: How to get it right with A/B testing.
So how can you build a similar recommendation engine for your site?
Amazon’s A/B Testing: Grab Consistently Convertible Ideas and Use Them Efficiently
Amazon’s A/B testing helps users receive optimized listings, which in turn helps Amazon sell more items.
Here’s an example, using a test we ran by accessing the same Amazon product page on two different computers. The following image is of a gaming product page on Amazon:
And this image is of the exact same product page at the exact same time, but on a different PC:
What Amazon did differently here is that we also see a suggestion for a gaming chair, a product placement accomplished with the help of A/B testing. Each of the two PCs being used may have fallen into two different divisions of Amazon’s audience while they were running an A/B test, which probably happens much more often than we realize.
According to techjury, 197 million people visit Amazon.com every month. If Amazon runs this experiment for over a month, they may have already sold many of these. If the suggestion works, we’ll start seeing more suggestions for related products in this way.
This system could be different for every single product page Amazon has, as there are systematic algorithms to accomplish this among various kinds of products.
Don’t think of this only as a way of upselling directly on the related product page. A direct upsell could be related items, which make an appearance on product pages or in that category only. But because Amazon is so good at following you around the internet, sneaky product placements could be anywhere via retargeting. You might start seeing suggestions to buy this product on blog pages, other websites, or other similar-category pages.
The question then becomes: How can you achieve the same effect as Amazon?
You don’t have to copy every single tactic Amazon uses. Instead, you can get a full conversion optimization test for your website or digital business. Know your needs, come up with ideas, and implement them in the best way possible. You can know exactly what those best ways are via A/B test results.
To understand the possibilities, here is an example:
- Assume that you are a company who sells music-related equipment online, but you’re suffering from poor conversion rates.
- To prevent your growth chart from going down, you contact your conversion rate optimization (CRO) agent or digital expert. They bring you a report of observations and related ideas to help increase sales.
- One of their ideas is to have extra product placements all over the site. You decide that’s a great idea, but you’re not sure how best to implement it.
- You hire Brillmark to investigate your best CRO options. We run an A/B test on your site, using pages with and without additional product placements. We then provide you with full development and implementation to accomplish your goals.
Based on the performance of the A/B tested pages, you’re still free to decide exactly what you want to do with your existing product pages. But with this additional information, you can implement profitable updates on your site with confidence. The goal of your business is to provide your customer with exactly what they want or need. A/B testing can help you serve up the right products at the right time to make sure customers are happier and you’re selling more items.
Replace Activities That Aren’t Profitable
Another way to improve your site with A/B testing is to ensure ease of navigation. If you’re not sure how to design the perfect user experience for your site’s visitors, A/B testing can take the guesswork out of the process and ensure any final changes provide a better experience for visitors, not a worse one.
This is important because customers no longer write complaint letters when they don’t like the way a website is configured. If they are trying to use your website but it’s confusing and hard to navigate, they’ll just leave — and shop somewhere else.
And that’s the last thing you want.
To prevent visitors from leaving because of a new navigation system, you can use A/B testing to try out possible replacements before implementing any of them permanently.