Your payers are converting fast (and you shouldn’t be looking at average time to convert)

I’m a firm believer that 1) only thinking about monetization from a rationalist/functionalist point of view is a mistake. Many users will make a purchase before having a clear understanding of how your game works. And spending in a game is not (always) a matter of making a rational decision with limited, partial or erroneous information – it’s not a case of bounded rationality. In other words, it’s not that (some) users are making an imperfect rational decision – they are making a decision that is simply not rational in nature. Not only are many users spending before they understand how your game mechanics work or the value of what they are buying. Many users are also willing to purchase things that will not help them perform better and/or progress further and faster (that don’t have an impact at the level of game function and systems) – and this is not just a Fortnite thing. And as a consequence of that (and other supporting data) 2) you should be designing conversion points and offers from the moment your users install your game.

A lot of data – which I’ll be discussing further in my next few posts – backs up these 2 assertions. Most of these have to do with the time it takes users to convert, and how patterns such as redeposit, LTV and churn are associated with the time it takes a user to convert. I want to begin this series of posts here by a general discussion on the time it takes to convert, and common patterns that occur across all games. Specifically, users convert quickly – it’s not just a matter of days since install; it’s even within hours and minutes. In fact, day 0 is the day users convert the most. That might seem contrary to an accepted narrative (in some circles) that argues it’s counterproductive to try to get users to convert quickly because the average time to convert is very high. I’ll conclude this post by showing how having a high average time to convert (or average time to do anything, for that matter) is only a reflection of the time your game has been available in the app store – not of user behavior patterns.

 

Look at time to convert

Looking at when your payers convert – i.e make their first IAP transaction – is a crucial exercise. That exercise can be interesting in of itself – to gain a better understanding of your mobile title. More importantly, it’s an interesting exercise inasmuch as it can help rethink the purchase triggers – and more generally who is paying and why – in mobile gaming entertainment.

The thing to keep in mind is: conversions (the first time a user will make an IAP transaction) will probably not account for a large portion of your revenue. Your most engaged users – those who are making repeated IAP transactions – are the ones contributing to your bottom line the most. In the grand scheme of things, the first transactions in your game will account for maybe 15-20% of all your IAP transactions and 10% or less of your lifetime revenue (another way to frame that is 90% or more of your revenue is probably coming from repeat purchases). And that’s despite the fact that 35-40% of your payers (or maybe even more) will only ever spend once in your game. Conversions might not be a great revenue contributor, but they refer to a critically important behavior – on which your title’s viability depends. You need to spend a first time to spend again. Furthermore, conversion refers to a significantly different pattern and triggers than redeposits – what makes a nonpayer spend for the first time is not the same thing that makes a paying user spend again. So, what will be valid for conversion triggers will not always work to get engaged payers to spend again. That’s why it’s crucial to observe the specificities of conversion patterns.

When looking at conversion, it’s particularly interesting to consider things from the perspective of the time since install it takes users to convert. It’s interesting for two reasons. First, it provides a common standard to compare games that might not otherwise have many comparable features (or audiences). Second, very consistently and regardless of the game or genre, similar patterns emerge when looking at the time it takes users to convert in mobile games, and how this correlates with payer quality. The fact that these patterns occur regardless of the game is an indication time to convert reflects something that is characteristic of monetization in mobile games in general – it has more to do with what characterizes a paying user than the game s/he is paying in. There is an inherent predisposition for paying users to pay soon after install.

 

Users convert more day 0 – in absolute terms

If you consider conversion from the perspective of the number of days following install it takes users to convert, very consistently – regardless of the title or the genre – the day users convert the most is day 0: install day. Depending on your game – and how long your title has been live for – you can expect between 20 and 30% (or even more) of your lifetime payers to have converted on their install day. By extension it’s not uncommon to have 50% of all your payers having converted within 3 days of install. If you were to distribute your payers according to the number of days it took them to convert, then you will probably see something like in the graph below (datediff is your friend).

days_to_convert

 

Your day 0 payers compose an important portion of your payers – and they also usually end up being your best payers. Those day 0 installs are not spread out evenly throughout that 24-hour period. Conversion occurs within hours – minutes even – after install. Even in a game that has been in the store for a long period of time and seen a lot of installs and paying users, you can probably expect to see close to 10% (or more) of your lifetime payers have converted the same hour they installed. This is the clear indication if we ever needed one that spending in mobile games is a process that occurs fast, soon after install.

hours_to_convert

 

There is even a sizeable chunk of your payers that first spend within minutes of installing and launching your game for the first time. If you’ve never looked at it before, I suggest you run the equivalent for your game of the query below (you can switch minute for hour or day – there is a long list of options).

min_since_install

 

Users convert more day 0 – in relative terms

Now, on one hand, you can expect 100% of your installs to be active on day 0 (even if nothing is actually ever 100%). So that means there are more users who have the material possibility of converting on day 0. But this is not the reason why most payers have converted on day 0. If you look at the relative conversion rate  – the percent of active nonpayers who is making an IAP purchase for the first time – it still appears that relatively speaking more users convert early on. In other words, the percent of non-payers that make an IAP purchase for the first time is higher soon after install. See below a sample query and output

active_nonpayer_query

conversion_active_nonpayers

 

If you simply look at the % of (unqualified) active nonpayers converting, chances are you will see the conversion rate to be higher on day 1 than day 0. But this is not really the best way to look at this, because by looking at things this way, you’re comparing apples and oranges for one key reason. Namely, every user active 1 day after install has played your game more than 1 distinct day (that user has played on install day and on day 1). That’s not necessarily the case for your users active on day 0. Depending on what your D1 retention is, you might be looking at 50% of your installs who are active on day 0, but never play your game another day in their life. Not only that, some users leave the game before they are able to spend in your game. Very often, users need to progress enough in the tutorial to even have the material possibility of spending and a sizeable portion of your users will never make it to that point (often the store is not functional from the get go. Users can only start spending after they are out of the tutorial). So being active on install day and being active 1 day after install refer to qualitatively different engagement patterns. If you try to control for this factor – by considering only active users who have played more than one unique calendar day – then it clearly appears that install day is the single highest day for conversion of your active nonpayers. And this pretty much consistently across all games.

conversion_active_nonpayers_2p

 

So, the fact that most of your payers convert on install day is not the result of higher traffic day 0 – i.e. that more users are active on install day. Relative to activity, day 0 is the day which is most conducive to conversion. In addition, this relative conversion rate is higher the closer to install day. It will gradually go down as time passes, and plateau at very low levels. The more time passes, the less likely you are to have your active nonpayers convert. This is a key factor you should never lose sight of, and that should guide your monetization strategy. There is a higher predisposition to convert early – therefore you are more likely to be succesful at getting your users to convert earlier rather than later.

 

Your title age impacts the average time to convert – and you shouldn’t be looking at average times in general

The fact that such a high percentage of payers convert on install day might appear to be in stark contrast with some claims that on average users convert on their second or third week. But it’s not. The longer your title has been in the app store, the more time users have had to convert. If your game has only been live for a week, then 100% of your payers will have converted within 7 days of install (and your average can’t be above 7 days). If your game has been live for 3 years, then some users had 1000 days+ to convert. And those users converting later are driving the average up. And average times are misleading because if you are thinking averages, then your main point of focus is the passing of time: how is the total sum of time having passed distributed among payers. If you are thinking median, then your point of focus is your userbase: what is the time it takes the majority of users to have performed action x. Time is not a commodity you need to keep track of. You need to get a sense of how your game and its tuning is making your population of users act as a whole. Average time would only make sense if time were the finite commodity you were selling and wanted to keep track of it – which you’re not, even if you say you are monetizing impatience. When you say you are monetizing impatience, that’s exactly what you are doing: time is not what you are selling; it’s the lever you are using to make what you are actually selling (gacha, character, extra moves, etc.) appear more valuable and desirable (or to make your users more impatient).

So, when you consider the average time to convert is 17 days, that might be the case (it’s lower for your high spenders than your low spenders). But that indicator is not providing you insights into how your userbase is behaving or your game performing. And you definitely should not be designing your monetization strategy to start getting your users to convert 2+ weeks after install. Your average time to convert might be 17 days, but 75 or 80% of your payers are converting before 17 days. When looking at the passing of time the median (or the percentile you will deem most relevant) will always be the way to go. A change in the median time to convert will reflect an actual change in conversion patterns. A change in average time to conversion will most likely only ever reflect the aging of your title.

If we return to the time to convert this is exactly what you will see in your game if you consider conversion. Chances are high that very consistently the median time to convert will be in the range of 2-3 days after install. It can be a bit more, it can be a bit less. But it won’t be weeks (it won’t even be one week). Furthermore, you shouldn’t be seeing big variations in your median time to convert if you don’t make big changes. The average time to convert, on the other hand, will increase as your title ages. I provide steps to reproduce that in your game here.

avg_and_median

 

There is one thing to keep in mind: the longer your title has been live, the more weight your previous converters will have – that means the longer your game has been live, the harder it will be to see a change in overall median time. If this is a concern to you, you always have the option to segment your conversions per install week, month or quarter (depending on what makes the most sense in your specific context). If you were to do the same exercise while looking at average times to convert, then you would see that the average will be higher for users who installed the longest time ago – that’s the “average effect” at play again.

avg_and_median_install_mt

 

So looking at percentiles – and focusing on the time frame that is relevant for you – will highlight 2 of the most important things to keep in mind for your monetization strategy. 1) Very consistently, converting is something that occurs early – as a matter of a few days, hours even – after install. And 2) chances are very high this trend will be stable throughout your title’s lifecycle. So over time as your title ages, the percent of payers who converted day 0 specifically might go from 30% year 1 to 25% year 3. But the median time to convert is likely to remain quite stable – because only your outliers will be converting on day 1000+. The main point remains: the vast majority of users who will spend in your game will not wait weeks and months to start spending. And to set yourself up for success you need to design your monetization strategy to kick in as soon as users install and launch your game.

And as a final note, looking at time to convert per LTV bracket – whether it’s an absolute value or a percentile (percentile is better!) shows that users with a higher LTV tend to convert faster than users with a lower LTV.

time_to_convert_per_ltv

 

So it’s not that everybody converts early but your better payers convert later. The majority of your payers are converting early – and the higher paying users tend to have converted even faster. You should not be shy, and try to get your users to convert the day (hour, even) they install your app. All the data indicates paying users have a propensity to convert fastyour users who pay. By capitalizing on this and desiging accordindly – whether by creating a “conversion moment” or offering appealing offers – you will be setting yourself up for success.

15 comments

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s