Experiment Results: Changing Keywords to Increase Downloads

Earlier this year I wrote about changing the keywords for Reminder+ (App Store link) – an experiment to see if I could increase downloads via App Store search.

This is what I wrote back in January:

Whilst I’m confident I can make an improvement, I don’t know how big the improvement will be. I’m not expecting anything dramatic, but I do think double digits is a possibility.

It’s taken a while but I now have some results…

Before and after download numbers

Before:

  • Period: 05/01/2016 – 17/03/2016, 72 days
  • App version: 2.0
  • App price: free
  • Downloads: 467
  • Downloads / day: 6.5
  • In-App Purchase: Send via Messages
  • IAP price: $1.99 (Tier 2)
  • IAPs: 37
  • IAPs / day: 0.51
  • IAP conversion rate: 7.9%
  • App Store views: 1411
  • Views per day: 19.6
  • Conversion rate: 33.1%

After:

  • Period: 01/04/2016 – 03/05/2016, 33 days
  • App version: 2.0.1
  • App price: free
  • Downloads: 280
  • Downloads / day: 8.5
  • In-App Purchase: ”Upgrade Pack”
  • IAP price: $1.99 (Tier 2)
  • IAPs: 7
  • IAPs / day: 0.21
  • IAP conversion rate: 2.5%
  • App Store views: 720
  • Views per day: 21.8
  • Conversion rate: 39%

Not quite double figures, but I’d say the change from 6.5 downloads per day to 8.5 downloads per day, is a positive result. (And, though the data only spans a month, the trend is rising slowly.)

The download increase comes from a slightly higher number of views per day and a better conversion rate. The new keywords have higher traffic, but the improved conversion rate suggests that the new keywords are more relevant to the application. I’m pleased about that.

Whilst this experiment was based on an app that has (relatively) a tiny amount of downloads1, I still think the process of reviewing, researching, and iterating on keywords2 – is worth the time investment.

Negative results: In-App Purchases

You may have noticed in the numbers above, that the IAP conversion rate has all but dropped off a cliff, dropping from 7.9% to 2.5%.

To facilitate the keyword change for Reminder+ I made a small bug-fix update to the app so that I could submit v2.0.1…

Unfortunately, the app was rejected. The rejection required a couple of innocuous changes along with one that became bothersome. The IAP Send via Messages was deemed unsuitable as it provides access to built-in iOS capabilities3.

To cut a long story short, I had to change the name, presentation and content of the IAP. Send via Messages became Upgrade Pack.

I anticipated that a less direct IAP would lead to a lower conversion rate – and that’s exactly what happened 😉


For more download and IAP numbers for Reminder+, look at the Downloads and Sales page.


  1. This app has a tiny amount of downloads: will factor into the changes to App Store views and ultimately, downloads.
  2. Researching keywords: if you’re curious about how I go about changing keywords, look at Researching (iOS) App Store Keywords with Sensor Tower.
  3. Rule 11.8 – Apps that use IAP to purchase access to built-in capabilities provided by iOS, watchOS, and tvOS, such as the camera or the gyroscope, or Apple-branded peripherals, such as Apple Pencil or Apple Keyboard, or Apple services, such as Apple Music access or iCloud storage, will be rejected.

Reviewing App Store Page Conversion Rates

I was reminded recently (thanks, Daniel Alm) of how important it is to review your app’s App Store page conversion rate…

What is the App Store page conversion rate?

It is: the percentage of customers that view your app’s “page”, in the store, and then go on to download it.

Note: that means actually selecting to view your app’s individual page; showing up in the search results list doesn’t count as a view of your app.

At a first glance this topic might seem simple: you want a high conversion rate and, by extension, if your conversion rate is too low you’ve got a problem.

Whilst that’s true, I think there are aspects surrounding this conversion rate that are worth consideration…

This is a sales process. The purpose of your app’s page is to convert a potential customer into a download. Whilst we are restricted with what we can do in the App Store, as compared to say, a regular web page, it’s still useful to think about approach it in this way.

Whilst I’m no expert the Internet is full of information on this subject. That said, here are some of the basic things to think about:

  • What will potential customers see first? Visually, the icon, app name and 1st screen shot/preview video attract attention – do they communicate a reason to download your app?
  • Is that reason clear? Are you communicating benefits or features? People don’t buy features…
  • How does that look from a search results page?
  • What do you know about the search terms your app ranks for in the store? Does your page communicate solutions that match these problems?

Monitor the conversion rate between versions. This is more obvious if you were actively trying to improve your conversion rate by changing the App Store page, but what if you weren’t?

Did you add a preview video or change your screen shots? A slight change to the name shown in the store? Or the icon?

Be sure to line up App Analytics queries with changing product versions to check if anything has changed.

How might we go about improving the conversion rate? Whilst we can’t do A/B testing in the App Store, as mentioned above, we can still monitor changes between versions.

Additionally, we can do regular usability testing on our App Store assets before making a change. For example, if you’re thinking about changing the first two screen shots along with the title, mock up two or three different versions and then test them on people.

Show them a version of the assets and then ask: what problem does this app solve? Or, why should you download this app? This type of testing can be very revealing.

Is this high priority work? It’s good to know whether or not you have a conversion rate problem. Addressing a conversion rate problem could be the highest priority work for your app, with respect to improved downloads/sales.

What’s more, the conversion rate is especially important if the majority of your traffic comes from search, as potential users are considering your app for the first time and so you only have the App Store page to convert them.

Conversion rate numbers for Reminder+

The origin of this post was a reminder to look at the conversion rate for Reminder+ (App Store link).

The following numbers are for the lifetime of v2.0 of the app (26th August 2015 – 17 March 2016):

  • Version 2.0
  • App Store views: 3735
  • Downloads: 1300
  • Conversion rate = 34.8%

I’ve heard that between 20% and 50% is a reasonable guide of acceptability for free apps, so 34.8% is not too bad…

However, consider the numbers for the previous version, between 1st April 2015 (when App Analytics started) and 25th August 2015:

  • Version 1.2.2
  • App Store views: 2192
  • Downloads: 1042
  • Conversion rate: 47.5%

That’s a significant drop in performance, worth consideration.


Note: I’ve updated the Reminder+ numbers page to include the App Store conversion data.

Researching (iOS) App Store Keywords with Sensor Tower

I’ve mentioned “researching keywords” a few times in this blog. I thought it might be useful to describe what I mean/do when I talk about researching keywords…

Keywords, the goal

Here’s what I’m trying to achieve. I want the best set of keyword phrases I can find for my app. In this case, it’s Reminder+. And by best, I mean:

  • keyword phrases that have high traffic; lots of people searching for that phrase and viewing the resulting list of apps.
  • keyword phrases that I can perform well with; will my app rank highly enough in that list to be considered, if I have those keywords?
  • keyword phrases that are relevant to my app; if I do rank highly enough to be considered, will my app match what the person was searching for?1.

The “researching” process

When I say researching, I’m using that word to encompass a few different steps.

1. Idea generation

I start by generating a list of possible keywords (and phrases). Generally, I approach this as follows:

  • My app offers a solution to problem x
  • What words and phrases might someone use when they’re searching for a solution to this problem?2
  • In addition to the obvious terms, can I think of any less common words that might also be used?

Those bullet points may seem obvious, but they’re important, especially the last one. A lot of the initial keyword phrases that come to mind are probably oversubscribed, and your chance of ranking highly for them, is slim. Given how much competition exists in the store, a key part of this process, for me, involves thinking outside the box.

2. Investigate those phrases in the App Store

Having a list of ideas is fine, but you can only use so many keywords, so you need to know which of them are the best for your app within in the context of the App Store:

  • How much traffic does a keyword/phrase get?
  • How much competition is there for that keyword/phrase?

This is where Sensor Tower comes in. Sensor Tower is a service that, among other things, can help with App Store optimisation. I use one of their free accounts3, which has several restrictions, and I (currently) only use one of their tools/pages: Keyword Rankings4.

Here’s an example showing some of the keyword phrases that I target with Reminder+:

I find this tool useful for the research phase as it gives me an indication of both the Traffic for a keyword phrase and the difficulty associated with ranking for that phrase.

It’s also possible to select one of the keyword phrases in question and look at the apps that currently rank for it, giving you a quick look at the competition.

One of the restriction associate with a free account is a limit on the number of keyword phrases that can be entered (tracked). Whilst this can slow you down and limits what you can realistically track for an existing app, it doesn’t prevent useful investigation.

3. Iterate and review

This process is iterative.

  • Invariably, some keywords phrases will turn out to be unsuitable, either because they don’t have enough traffic or they have too much competition. So they come off the list and I look at alternatives.
  • The process of reviewing the list and trying things in Sensor Tower often generates new ideas, which are then tested.
  • The two places we can put keywords are the keyword field (100 characters) and the App Store name (255 characters); as I get closer to the set of keywords that I want to use, if some of them are going to feature in the App Store name, this also needs to be reviewed.

  1. Relevant keywords: for example, it’s no good ranking first when people search for “fly fishing” if your app is for making VPN connections…
  2. What words and phrases might someone use… – you could ask people, use a thesaurus, look at keywords similar apps are using…
  3. Free account with Sensor Tower: my current app revenue wouldn’t cover the subscription price…
  4. Keyword Rankings tool: is accessed via App Intelligence > App Store Optimisation > Keyword Rankings