On this page, I’m going to show you how traditional split testing is costing you tens of thousands of dollars in lost profit every year, and what you can do instead to optimize your sales without driving yourself crazy.
The method I'll show you will allow you to get better results than traditional A/B split testing, with less effort and 80% less wasted traffic.
At some point, you’ve probably been sucked into running A/B tests, with the promise that it’s an essential part of optimizing your store’s sales.
It’s not surprising - the premise makes sense on the face of it…
They tell you: “Don’t try to guess what will have the best conversions, let your customers tell you by watching their actions”
The experts tell you to test EVERYTHING!
“The images / descriptions / prices of your products can all be tested, and you can see what does the best over time!”
“If you’ve got unlimited energy and focus, you can keep running split tests to see if you can beat your control, and eek out an extra dollar or two!”
It pretty quickly turns into a crazy stressful guessing game, where you worry that you’re throwing away sales in the name of testing, and always have to be coming up with new variants to test. The whole process is maddening, and it doesn’t help that you’re only ever confident about the results at a single point in time.
Think about it this way - If you’ve run 10 tests to optimize your main product image, and ran each for a week, you’re over 2 months out from the results of your first test.
Your “winner” image won 10 weeks ago, but is that still true today? Changes in the season, current style, and your incoming traffic sources (among many other factors) could well have rendered those findings obsolete. Maybe the image that lost 2 months ago, would win today!
More annoying, if you’ve actually split your traffic (ie 50% to each option), then you’ve wasted half your traffic 9 times over.
If you have 1,000 visits/week, and you’ve “split” those for 10 weeks, then you’ve wasted 5,000 visits on suboptimal displays, costing you thousands of dollars.
Yes, you know that your tests picked a winner, but how long are the results of those tests valid for? Do you have to constantly be re-testing, or constantly be adding new images / descriptions to test?
It’s an especially tough problem to solve when you’re selling fewer than 25 products.
Sites that have hundreds, or thousands of products can get away without optimizing the images/descriptions on every one - they’re operating on a model of quantity and novelty.
Your store, though, makes its money by having a handful of high quality products, and so it’s critically important that every single product carries its weight.
This is why it can feel overwhelming sometimes - how can you balance the need to ensure each of your products is pulling its weight, without wasting a ton of traffic with losing tests?
Here’s what we’ve learned, working with dozens of Shopify stores doing between $10k - $500k/month in revenue: There’s a simple workaround to this problem, that gives you the best of both worlds.
You can still run tests & ensure that you always know the "best" variant of your images and product descriptions, and you’ll know that at any point in time, the vast majority of your visitors are seeing these highest converting options.
Best of all - you can ‘preload’ in a bunch of options up front, and the system will automatically find the best performers, send traffic to those, and adapt in real time.
With less than 5 hours/month of work, you can have world-class continuous testing running on your store, fully optimizing each of your products, around the clock.
Let's use an example of a store selling coffee - you've got a lot of options for the header image for each product.
Here, let's imagine we have 6 possible images:
Which image gets the most clicks? Which ones are duds?
Instead of split testing them one by one, let's throw them into the ring together, only showing the image with the highest conversion percentage at the moment a new visitor comes to the site. This way, you'll never be wasting traffic - everybody is seeing the highest converting images. If an image starts to underperform, then it will be replaced quickly.
Press "Start Simulation" to see how each image performs over the first 300 visits, and watch the highest converting image rise to the top (and get the lions share of traffic).
Best of all, notice how little traffic the worst performers get - no need to waste a single visit on "split" testing!
We all know a picture is worth a thousand words, and for shopify stores, the right picture is worth thousands of extra dollars in sales.
After all, the gateway to your lovingly crafted product descriptions is the gallery image - if you can get more clicks there, you're that much closer to a sale.
You’ve got a lot of pictures for each of your products - so picking the right picture is vitally important.
Our app will run the initial tests to determine the highest performing picture to use as your main image, and ensure that each product has the “best” picture showing in your galleries
You know it’s not enough to slap together a quick description of the product & hope people buy it.
So, with our app, load anywhere between 3-10+ descriptions for your product into the system, and we’ll quickly determine the one that has the highest sale rate.
Feel free to periodically re-vamp the descriptions to see if you can find new winners - since the system automatically uses the highest converting descriptions, it will quickly either rise to the top or sink to the bottom.
That means you can be as agile as you like, without it driving you nuts. New descriptions that convert will rise to the top. The ones that don’t shine will quickly sink. The whole process is nearly effortless.
This is the most important piece of all.
As we mentioned earlier, one of the biggest drawbacks of A/B testing is that the results really only tell you what was true at one point in time, and it wastes too much of your precious traffic showing non-converting tests.
The better solution is to quickly identify the highest performing assets, and then to use those for 90% of your new visitors.
If the success rate of your best option drops below the success rate of option #2, then they change places, and your new “best performing” option will receive 90% of the new traffic that comes in.
This simple principle ensures that for ever new visitor, they are seeing the highest converting images / descriptions at that moment in time.
This is as close to magic as you can get.
The remaining 10% of traffic is distributed among options 3 and beyond, to ensure that they have a “fair shot” of climbing the ranks.
Here's a concrete example to make it easy to understand: If your best performing asset has a 40% win rate, and your next best performing asset has a 39.9% win rate at the conclusion of the first test, it may well be the case that the 2nd one will perform better over time.
With traditional A/B testing, you’d never know - The first one would be the “winner”, and the 2nd would be forgotten about.
But with Ecomm-AI’s TestAI, you don’t have to guess whether or not your first option is always the best choice. The second it drops below the conversion percentage of the next best option, that next best option will “take over” as the recipient of most traffic.
This will effectively put it in first place, and it will get 90% of traffic.
If it really does better than your previous #1 option (in this case, if it gets 4 more sales over the next 10 visitors), then it will cement its place at the top. Options 2 & 3 will still get some traffic, in case some outside factors change to make them the best options again - in which case they’ll take the top spot.
In essence - each of your testable options has the opportunity to rise without you thinking about it, and you can sleep well knowing that at any given time, the images & descriptions that are getting the best results RIGHT NOW are ones being shown.
Even better? When you add in a new option, it will immediately be initialized to get some traffic, and will either out-perform your #1 choice, or settle into the ranks within the first 50 visitors.
If it immediately outperforms your current pick, it’ll stay on top. If additional traffic shows that early performance was just a fluke, then it’ll fall in the ranks. Either way, it will very quickly find its place without you having to lift a finger.
This allows you to test new image / description combinations effortlessly.
As much or as little as you like, you can add in new options for your images, and/or for your descriptions, and we’ll make sure your customers are always seeing the ones that are converting the best.
Behind the scenes, we’ll be sending just enough traffic to your lower performing options to ensure that they’re not a better choice.
What this all means is simple - you can stop throwing away traffic, and stop driving yourself crazy with traditional A/B testing, and instead “set and forget” your testing, fully assured that you’re not wasting a single visit.
So are you ready to give up traditional, time consuming, AB Testing for a proven method being utilized by some of the world’s top E-Commerce stores? Now’s the time.
Just visit the Shopify App store to install. It's free to get started.