App Marketing

Measuring K-Factor: What are the effects of paid UA on organic installs?

James Haslam
Content Manager

It’s a common belief in mobile that there’s a definite relationship between paid and organic installs. That is, that the more paid traffic you have, the better your organic traffic will be. The idea even has its own name, ‘K-Factor’.

K-Factor was initially used as a term to define measurement of social sharing - namely the relationship of shares to install. But the phrase has gradually changed over time, and now has closer associations between many types of relationships within marketing. For our purposes, we’ll be talking about how many organic installs result from paid.

And while it is common, it’s a relative unknown, and there’s yet to be a full-proof study showing whether the K-Factor exists (and, if it does, how big a factor it really is.)

So, we set out to prove just that, and analyzed thousands of apps to satisfy two questions: Do paid installs affect organics? And, if so, how strong a relationship is it in reality?

What is K-Factor?

K-Factor is the basic assessment of whether organic installs are effected by paid installs. So, the theory goes, if I pay for a lot of traffic, I should also see a natural rise in organic users, based on a number of factors such as impressions, app store rankings, and virality.

Here, we’re going to test just that, and find out if the K-Factor really exists, and (if it does) how much influence it has.

The math behind it

One of Adjust’s Data Scientist, Roman, started out by rewriting our initial question as an equation:

Y = F(X)
Y = organic installs, X = paid installs, and F(X) = is some function of X.

Then, we assumed a linear dependency between X and Y (i.e. the relationship between organic and paid installs is constant over the paid installs volume) so we could test a more developed model. This assumption is that for a given app, if they grow or fall in size (installs) their K-factor will be constant. Therefore our formula becomes:

Y = b*X
b = the coefficient we want to estimate

Since we’re analysing weekly time series data, we face a few limitations: two time series Y and X need to be stationary (constant mean, variation and autocorrelation) to be included in the model.

Our X and Y are not stationary, so we take the first differences of these time series. The advantage of this approach is that we don’t lose the interpretability of the results, so we still can interpret absolute changes.

After that the model becomes:

Y* = b*X*
Y* = first difference of organic installs (absolute growth)
X* = first difference of weekly paid installs (absolute growth)
b = K-Factor itself.

If, in the current week, we change our paid installs relatively to the previous week by one unit, then we can expect a change of organic installs relatively to the previous week by b.

Our sample

Roman took the sample from a year of installs, looking at weekly data of each individual app. He ran the test between 21/11/2016 - 26/11/2017. Apps in our sample had to satisfy three criteria:

  1. Apps had to be consistently installed from both paid and organic sources for more than 50 weeks (during the dates above.)
  2. Apps had to receive more than 500 average installs per week.
  3. First differences (absolute growth) of paid installs and organic installs had to keep stationary during this time. This means they aren’t affected by seasonality or other trends.

According to the sample criteria we analyzed a total of 1345 apps combined. Splitting by OS, we looked at 711 Android apps, and 634 on iOS.

What did we find?

What we ultimately found is that while K-Factor definitely exists, it doesn’t apply to the majority of apps on the marketplace.

30 percent of our sample had K-factor. For this 30 percent our data team established that the median K-factor was .45. This means that (looking at apps performing in the middle of our sample) for every 100 paid installs, you would receive an additional 45 organic ones. And that’s just the median. In our sample, we also found some apps received hundreds, if not thousands of additional installs.

Who has the K-Factor?

To figure out whether your paid installs affect your organics, take a look at your paid install volume. It shouldn’t be much lower than overall organic installs. Comparing the two, paid installs have to amount to as much as 65% of your organic total in order for your app to be affected by K-Factor. So, if you receive 100 organic installs, you’d need to receive at least 65 paid installs for K-Factor to happen.

One of the possible explanations for this is that when you have a high level of organic installs, it becomes much harder to reach new organic users.

That said, we saw a trend for apps that have lower downloads overall. These don’t tend to see any affect from paid to organic. In fact, the apps with the largest userbase also had the largest K-Factor, and gained a big influence in paid advertisement which boosted organic installs.

There’s no real difference from what we’ve seen between platforms: iOS and Android behave practically the same.

We also had a deeper look at how K-factor differed by app vertical. We started by looking at games, and (of our sample), we found that K-factor affected 22.5 percent of them (470 apps), while for non-Games (871 apps) it’s 33.6 percent. So, here, gaming apps are less affected by K-factor. In fact, it’s ecommerce apps (158 of them) that have a higher percentage of K-factor existence: 38.6 percent, while non-shopping apps have a 28.5 percent likelihood of K-factor by comparison.

Finally, there is no noticeable dependency between an app’s lifetime (from the release date) and the probability of having K-factor, or K-factor strength. This means that no matter how old an app might be, there’s always the chance of K-factor.

A couple of takeaways

So, we found K-Factor, but what does it mean for most marketers?

Identifying the influential aspects that go into K-Factor, and improving and optimizing them as much as possible, are crucial to boosting the sheer number of installs from organic that come from the force of paid. Virality, ASO, and creating apps designed to be shared, are key to cracking K-Factor, so let’s cover off a little about how they play into an overall improved K-Factor.


What’s more important to organic downloads than good ASO? Not much, but, then again, one of the largest ranking factors of ASO is download volume. So, a large install base will help to boost rankings, regardless of whether users are paid, or organic.  When app stores see consistent traffic, and an ever-increasing aggregation of active users, ranks increase, and solidify, creating a healthy circle of organic influencing paid, which in turn leads to a bigger K-Factor.


It’s the apps that make a splash that reap the rewards of K-Factor. Apps that go viral benefit from all sorts of coverage, and an uptick in organic traffic and paid engagement. If users have heard about it they’ll tap ‘install’ without hesitation. This, in future, creates a much lower barrier to activities like retargeting, and re-engagement. K-Factor isn’t the only beneficiary to creating a big hit.


Simply put, apps made to share benefit K-Factor immensely. Whether it be a multiplayer game, or a travel app that allows users to split the cost, encouraging your users to share the app with friends can turn one paid install into three or four more organic users, or more.

That’s all for K-Factor - keep an eye on the Mobile Marketing blog for top tips, and more.

Want more data stories?

Sign up to our mailing list below.

You're a member of our mailing list!

Keep your eyes to your inbox for more.