7 Ways To A/B Test Your App Page Creatives

When you might enhance your app’s conversion price by 66% by altering one graphic, would you do it? After all you’ll! By now, you know the way essential A/B testing is on your app. For each 1% enhance in conversion you obtain, you’ll be making 1% more cash; it’s that straightforward. The issue for most individuals is: how do you really do it?

Table of Contents

When you’re on Google Play, utilizing Google Experiments is the plain reply. However what in case your app is simply obtainable for iOS? What if you wish to run a take a look at earlier than launching?

On this publish, I’ll present you a handful of how to A/B take a look at so that you gained’t have any excuse to not!

WHAT SHOULD YOU BE TESTING?

When deciding what to check, search for elements of your app web page that are clearly uncared for. For instance, when you have a ravishing icon however your screenshots are weak, take a look at out some new enhanced screenshots with captions and see what occurs.

In line with StoreMaven, these are the typical will increase in conversion for every ingredient listed so as of significance:

The place do you test your present conversion price?

In case your app is within the Google Play Retailer, you possibly can click on beneath the “Acquisition Reviews” tab to see the conversion information on natural downloads. Google Play can also be type sufficient to point out benchmark information of how your app performs towards different apps in the identical class.

As you possibly can see, In case your app converts under the typical app in your class, Google will tease you about what number of downloads you’re lacking.

To test your conversion price within the iOS retailer, open up the “Analytics” web page and click on “Conversion Charge”. As soon as that is chosen, the chart will present your app’s conversion price each day.

To be sure to’re solely getting information from natural customers within the iOS dashboard, setup a filter for both “App Retailer Browse” or “App Retailer Search” like so:

It will make sure that you’re solely amassing information for customers organically looking the shop.

1. BEFORE AND AFTER TESTING

Onerous to name this one A/B testing however it may well nonetheless be helpful to offer insights into which creatives carry out higher.

When you’re critical about bettering your conversion (and you have to be), I might advise towards utilizing solely this technique.

Professionals:

Tremendous straightforward to do
Simply add the brand new variations, wait just a few weeks, then examine the conversion price earlier than and after.

It’s free

Cons:

Very inaccurate
Check out your conversion charges proper now and also you’ll see that they bounce round loads.

Even once you’re not making any modifications, your conversion price can fluctuate by as a lot as 50% each month which might result in incorrect assumptions.

No Pre-launch testing
You’ll must have the app within the retailer to make use of this technique.

Takes a very long time
You’ll must publish the app and provides it time to collect sufficient information earlier than making any conclusions.

When do you have to use this technique?

It is best to think about this technique if:

Your app is simply obtainable on iOS
You don’t have the finances for paid adverts
You don’t have a touchdown web page for the web site you possibly can run assessments on
You’re completely positive the brand new variations will enhance your conversion price

2. RUN TESTS WITH GOOGLE PLAY STORE LISTING EXPERIMENTS

In case your app is within the Google Play retailer, this must be your go to technique.

With Google experiments you possibly can A/B take a look at on issues like your icon, description, characteristic graphics, screenshots, and promo video.

Establishing a Google Experiment is straightforward.

There are two kind of assessments you possibly can run:

World– World assessments will present to all customers. These kind of assessments are good for belongings that don’t contain textual content (i.e. icons) because it is not going to isolate anybody who doesn’t communicate the language.
Localized– Localized assessments will solely present to customers within the focused language. With these, you possibly can take a look at issues like description, movies, or screenshots by ensuring the individuals who see these belongings can really communicate the language.

When you’ve determined what kind of take a look at to run, select what you need to take a look at and let it rip!

Professionals:

Correct
Because you’re operating actual A/B assessments straight within the Google Play retailer, your outcomes are going to be rather more correct. Granted, the statistical significance and “90% confidence interval” leaves room to some errors.

Simple to make use of
All you want to get began is an app within the retailer and a few variations to check.

It’s Free

Cons:

Just for Google Play
Hopefully quickly Apple will launch their very own A/B testing software, however for now, all we will do is dream.

In case you have an app in each Google Play and iOS shops, you possibly can run a take a look at with Google Play Retailer Itemizing Experiments then attempt to apply one of the best ones to the iOS retailer (however be sure to consider the influence).

No Pre-launch testing
To run an experiment, you’ll want the app printed already. If you wish to take a look at your belongings earlier than launch, you’ll want to make use of one other technique.

When do you have to use this technique?

It is best to think about this technique if:

Your app is on the market for Google Play

Google Experiments must be your go to testing technique in case your app is on Google Play.

Case research

Take a look at these nice case research on Google Experiment success tales:

EverythingMe Google Experiment case research
Dashlane Google Experiment case research
Aviary Google Experiment case research
Kongregate, Dots & extra Google Experiment case research

3. USE THIRD PARTY TESTING SOFTWARE

When you’re app is simply obtainable for iOS and you’ve got the finances for it, this must be your go to technique of A/B testing.

With this technique, you’ll use a 3rd occasion service like SplitMetrics (my private favourite), Storemaven, or Take a look at Nest to arrange mock app retailer pages. In my expertise, they stunning a lot do the identical factor with minor variations within the analytics reporting.

As soon as the web page is setup, you then create a number of variations of it with completely different modifications to check towards one another.

For instance, you possibly can create two mock app pages, every with a special icon.

From there, you’ll drive site visitors to the variations with Fb Advertisements or any site visitors supply of your selecting to see which one converts one of the best.

Right here’s an instance of a marketing campaign we ran with Phantasm Labs to check 3 completely different icon variations for in the course of the delicate launch of their new recreation BaconEscape utilizing SplitMetrics.

As you possibly can see, the model displaying the pig on the practice tracks transformed one of the best which is what they ended up utilizing.

To run the assessments, Apptamin already has 2 depraved guides for this technique, test them out:

The way to A/B take a look at App Preview movies vs. Screenshots
Double Your App Conversion Charge with App A/B Testing

Professionals:

Permits for pre-launch testing
When you’re seeking to take a look at your graphical belongings earlier than launching, utilizing third occasion software program is an ideal option to do it. As soon as they click on “set up” on the mock app web page, you possibly can direct them to the touchdown web page the place you possibly can accumulate their e-mail.

The draw back of that is it might upset some customers who’re excited to attempt it out since they’ll have to attend till launch.

Permits testing on extra variables directly.
For instance, you possibly can take a look at your screenshots in a single take a look at with paid adverts, then icons in one other marketing campaign.

You should utilize attribution hyperlinks to trace person behaviour
When you’re utilizing a service like AppsFlyer, you should use their hyperlinks to trace how customers behave as soon as they obtain the app.

That is essential as a result of completely different variations could appeal to customers who behave in another way and spend more cash.

For instance, maybe variation A converts 10% higher than variation B however AppsFlyer exhibits that customers who noticed variation B spend 20% more cash. Which variation would you select? Variation B after all!

Cons:

Costly
Not solely can paid adverts value loads, the softwares wanted to run the assessments begin their pricing at $149/month (TestNest) and shortly goes up from there.

Not all clicks on third occasion retailer web page will convert within the app retailer
When a person clicks “set up” on the mock app web page, they are going to be taken to the actual app web page the place they must click on it once more. Throughout this course of, a handful of customers will drop off, that means that you just’ll find yourself paying extra for adverts operating to a mock app web page than you’ll usually.

When do you have to use this technique?

It is best to think about this technique if:

Your app is simply obtainable on iOS
You need to do pre-launch A/B testing
You’ll be able to afford to purchase the testing softwares and run a paid adverts marketing campaign alongside

4. A/B TEST DIRECTLY IN PAID AD CREATIVES

With this technique, you’ll be utilizing paid adverts like Fb Advertisements or any advert community of your selecting to A/B take a look at your creatives.

The way it works for Fb is you run Fb Advertisements displaying off your completely different creatives.

This works loads like utilizing third occasion software program talked about above besides as a substitute of driving site visitors to a mock app web page, you’ll be utilizing Fb Advertisements straight to check your creatives.

It seems one thing like this:

For different advert networks apart from Fb, you possibly can take a look at your icon, description or screenshots inside the adverts.

For instance, in the event you had been to make use of Chartboost, you can take a look at these:

Right here is a further hyperlink from Chartboost: Enhance Your Cell Promoting Conversion by as much as 15% in One Click on with Social Metadata

Professionals:

Cheaper than with third occasion software program
Because you gained’t want to incorporate any of the third occasion software program, all you’ll must pay for is the adverts.

Can pre-launch take a look at completely different belongings
With a Fb marketing campaign, you possibly can take a look at completely different inventive earlier than launching the app.

Can take a look at a number of variables directly
For instance, you possibly can take a look at screenshots in a single marketing campaign and icons in one other all on the similar time.

Cons:

May give skewed outcomes
Though this technique gives you a good suggestion of what converts one of the best, the information is not going to be as correct in comparison with utilizing an actual or mock app web page. It’s because the adverts will no look the identical as an app retailer web page so customers gained’t see the graphics in the identical context as they might on the app retailer.

Nonetheless may be costly
Though it will likely be cheaper than utilizing a mock app web page, it may well nonetheless value a good bit to drive sufficient clicks on the advert.

When do you have to use this technique?

It is best to think about this technique if:

Your app is simply obtainable on iOS
You need to do pre-launch A/B testing
You’ll be able to afford to run a paid adverts marketing campaign however not the testing software program resembling Cut up Metrics

5. A/B TEST DIRECTLY ON YOUR LANDING PAGE

In case you have a separate touchdown web page on your app (which you must have in the event you adopted my 14 app launch ideas infographic), you should use it to A/B take a look at completely different graphics.

For instance, you should use a service like Optimizely or Google Optimize to run assessments on icons, screenshots, descriptions, title, and even your pricing straight in your touchdown web page.

Supply: Turnplay

Supply: oggl (Hipstamatic)

Shoutout to Sujan Pate for this one.

Take a look at this Kissmetrics Information for extra particulars on how one can pull this off.

Professionals

May be achieved in pre-launch
All you actually need is straightforward e-mail assortment touchdown web page to check variations.

May be low cost and straightforward
In case you have a web site already arrange that’s getting first rate site visitors, all you want to do is instal an A/B testing software and run the assessments.

Cons

Not as correct as app retailer web page
Because the context in your web site will probably be completely different than on the precise app retailer web page, the outcomes could also be skewed.

May be expensive and time consuming
When you don’t have a web site already made, it may well take a while to get one arrange. Additionally, you’ll sufficient site visitors to make the assessments occur so you could must run paid adverts to the location.

When do you have to use this technique?

It is best to think about this technique if:

Your app is simply obtainable on iOS
You need to do pre-launch A/B testing
You’ve got a touchdown web page already obtainable with sufficient site visitors you possibly can run A/B assessments on

6. IN-APP OFFER WALLS

This one will solely work when you have different apps you should use to cross promote with.

The way it works is you’ll arrange an app supply wall in one other app you personal to cross promote the app you need to take a look at on. From there, you’ll A/B take a look at this icon utilizing a service like Apptimize or Optimizely.

Right here’s what the supply wall you’ll A/B take a look at icons on would appear like:

Professionals

May be low cost
When you get your supply wall setup with completely different icons, all the present customers of your recreation will probably be your A/B tester free of charge!

Cons

Not as correct as app retailer web page
Because the context in your web site will probably be completely different than on the precise app retailer web page, the outcomes could also be skewed.

May be time consuming
It will possibly take some time to get the supply wall arrange together with the A/B assessments.

You want no less than one different app that’s already profitable
Because you’ll be utilizing one other app’s present person base to A/B take a look at, you’ll want to really have one other app with an present person base.

When do you have to use this technique?

It is best to think about this technique if:

You’ve got a big base of comparable apps you possibly can cross promote between

7. RUN AN A/B TESTING POLL

With this method, you’ll share your graphics with a closed group of individuals and ask them to vote for his or her favourite one.

Right here’s what it seems like in a Fb group (shoutout to Steve Younger):

You’ll be able to run assessments like these in locations like Fb teams, Subreddits, and even by displaying variations to family and friends. Listed here are some hyperlinks:

Fb – App Entrepreneurs and Entrepreneurs Group
Reddit – suggestions request instance

One other nice paid possibility is to run a ballot with a service like Pickfu evaluating two variations. You will get 50 responses in your ballot for less than $20

Professionals

Low cost and straightforward
All it’s a must to do is publish your ballot to a bunch of individuals and ask their opinion. Even with a paid ballot you will get outcomes pretty cheaply.

You will get direct suggestions
Because you’ll be interacting with folks extra straight, you possibly can have them give suggestions on what they like and didn’t like.

Permits for pre-launch testing

Cons

Inaccurate
Having folks choose which one they like higher is way completely different than having folks about to obtain the app determine. Usually instances, they could simply choose probably the most fairly or flashy icon, not the one that truly converts one of the best.

When do you have to use this technique?

It is best to think about this technique if:

Your app is simply obtainable on iOS
You want to get some actual qualitative suggestions with the assessments which you should use to tweak the designs

CONCLUSION

So there you will have it. You now don’t have any cause to not be A/B testing it doesn’t matter what platform your on.

One closing piece of recommendation is that you may begin your testing with a qualitative technique (like operating a ballot) to level you in the suitable path.

For instance, let’s say your app will not be but launched and also you’re deciding on what colours to make use of in your icon and screenshots. What you are able to do is run a ballot utilizing a number of completely different colour palettes, accumulate suggestions on each, then use that information going ahead to make higher designs.