Archive for February, 2018

Predicting 2018 sales based on past data

Tuesday, February 27th, 2018

Cash flow is super-important. I need to know how much money I can expect to earn from game sales, tax breaks and whatever else may come my way in order to know how much runway (months I can afford to pay myself to work) I’ve got. Then I can plan what sort of projects I should work on, or if I should seek external funding etc.

In order to predict cash flow for 2018, I looked at my 2016 sales and worked out how much revenue has dropped by in 2017. Then I used that drop to predict 2018 sales (see below).

Why has revenue dropped?

Well the obvious reason is that, apart from Shadowhand which released in Dec 2017, I haven’t released any new games since Regency Solitaire in 2015.

Sales for all my old games naturally decline over time. Though another possibility is some of the distributors I use are in decline (such as the casual portals) or overcrowded (such as Steam) which would also affect my sales.

Direct Sales more stable than Steam

You can see from the chart at the top of this post that Direct Sales in 2017 were 62% of 2016, whereas Steam sales in 2017 were 47% of 2016.

Some distributors like iWin (73%) and Mac App Store (75%) appear to be even more stable. A big worry for me is Big Fish Games (BFG) at 47% because they were previously my biggest source of casual portal income.

Overall 2017 generated about 50% of the sales of 2016.

Per Game

I also worked out the drop in sales per game on the casual portals.

This is interesting because my oldest games (Oz and Holiday Bonus) haven’t dropped much at all. This is because the long tail has pretty much stabilised for those games.

Spring Bonus is a bit newer (2011), and Spooky Bonus (2013) and Regency Solitaire (2015) are newer still. So it seems reasonable that they may still be in a long slow decline rather than a stable long tail.

On Steam the values for Spooky Bonus and Regency Solitaire were very similar – around the 50% mark.

2018 projection

So anyway, it seems reasonable to assume that 2018 will only generate about 50% of the revenue of 2017. However, if the casual portals decline rapidly this year, or the overcrowding of Steam and the general race to the bottom of discounts/pricing affects sales adversely, the figure could be lower than 50%.

However, if my newer games (which generate the most revenue) begin to move into the stable long tail period, then it’s possible that revenue will be greater than 50%, though I doubt it would reach 75%.

I have sent some of my games to two new distributors this week, and I might even try out at some point. So it’s possible that could boost revenue a bit more. We’ll see.


The good news is that Shadowhand is on track to produce about as much net revenue per month as all my other games combined! That’s if we hit 3x week one sales in the first year as per this blog post.

But we also have a few extra things planned that may boost the revenue from Shadowhand later in the year. So fingers crossed that those things work out!

The Bad News

The bad news is that sales from my old games, Shadowhand, and tax relief (due about June) will only give us a 6 month runway with a livable salary, which isn’t very long. So we’ve got a couple of quick projects lined up that should start generating money in the second half of 2018.

Also we just applied to the UK Games Fund for a £25,000 grant for a cool game concept that I’d be collaborating on with another dev if it goes ahead. So mega fingers crossed for that!

If that fails I’ve got other things I could do that are short and involve reusing existing tech/IP. Ideally I’d get some investment in for those projects though to help with cash flow and paying artists. But I don’t need to cross that bridge just yet.

Steam Week 1 vs Year 1 Revenue

Monday, February 5th, 2018

Recently I asked a bunch of devs how their week 1 (gross) revenue on Steam compared to their year 1 revenue (including week 1). I wanted to know because we had just launched Shadowhand on Steam and I was interested to see if the first week could in any way be a decent predictor of longer term sales.

I got about 30 data points and you can see the resulting graph above.

A fairly tight range

The results are interesting in several ways. One obvious one is that most of the results fall in the 2x to 10x range and aren’t something like 50x or 100x. However, one dev who runs a large studio did have an outlier title that had a very large ratio, but I excluded it because it’s really not typical of your “average” indie game.

OK, sure the range isn’t that tight because if your game did 2x week one you might go bust and if it did 10x maybe you’d end up buying a Tesla (depending on the scale of your week 1 revenue). But I still think it’s interesting and can be useful. Let’s consider some more things…


So the Mean (average) is 5.1x and the median (middle value) = 4.5x

This is pretty useful data. If your game made $10K in the first week (not an easy feat btw), you could perhaps reasonably assume it would make about $50K in the first year.

Be Conservative

This is not a political statement, but more like a sensible approach to predicting the future based on past performance.

Basically, it might be more sensible to assume something like a more conservative 3x ratio and plan your cash flow based on that.

That’s what I’ve done for 2018 after I saw the first week of sales for Shadowhand. At 3x, my year will be a bit tricky cash flow-wise, and at 5x it’ll be “OK”. Any more and I’ll be very happy. So, for me, based on a 3x-5x range it was clear that I need to get on and earn some more money pronto!

What about the Steam “flood”?

“But what if some of your data points are old and some are new? Surely that makes a difference in today’s crowded market?”

No. I checked, in a cursory way, and the release date of the game didn’t seem to make any difference.

What about big games vs small games

You will have definitely heard of some of the games in my chart, they have made millions of dollars. Others you probably won’t have heard of. Also some of the games were large in scope and others were not.

Again, I could not detect an obvious correlation between game size/fame and ratio. Though I haven’t spend very long on this, so take it with a pinch of salt.

What factors can affect the ratio?

My own game, Regency Solitaire, had a 10x ratio. It has not made millions of dollars, but it does have a good Steam review score (97% positive). So it’s possible that is a factor.

Another of my games, Spooky Bonus, has a 98% positive score and a 6.3x ratio. So I can’t draw a firm conclusion from that, but it seems logical that “good” games have a better tail as people talk about them more, and maybe the Steam algorithms surface them more to potential customers, plus of course a good review score may sway a “floating purchaser”.

If your game had a poor launch with little marketing support but gained traction later in the year (with press/streamers), or you heavily discounted it later in the year, that could obviously affect the ratio too.

Updates/DLC, localisation, community, being in a bundle, more marketing etc. All of those things could affect the ratio and I’m sure you can think of more.

Ideally if you get everything right (and there’s a lot to get right), you’d hope to have a good launch and also go above the average 5x ratio.

Can I use this chart to predict year 1 revenue?

No. Obviously the above chart is useless until you know your week 1 sales, and predicting those is a black art because they can have huge variability in range.

But once you know your week 1 sales, you will have a better idea of if you need to rush out and find some contract work, or if you can chill a bit and make more another game. Some sensible people have a plan B, such as contract work, lined up for post-launch anyway.

However, let’s have a little bit of fun with some extrapolation…

Recently Mike Rose gave a talk about realistic PC sales figures in which he stated that the average “good” game on Steam will make $10K in its first month if priced around $10. Now we don’t know his criteria for “good” but let’s run with his numbers:

– $10K in month 1 from Mike’s data.
– Assume $5K in first week (it’s not 25% of $10K because the launch spike is most of the sales)
– Use a 5x average multiplier based on my chart above.
– This gives $25K in first year.

That’s for an average “good” game. Doesn’t sound like a lot does it? Now go and read You are spending too long making your game for further thoughts on how to make more profitable games.


This article is not advice, it’s just some personal musing that may or may not apply to you. So take it or leave it, and please feel free to correct me or expand upon my points in the comments. Thanks.

How did your game do?

Let me know your gross Steam revenue week 1 vs year 1 gross ratio in the comments. Probably best not to share actual $ values as that is no doubt disallowed in your Valve contract.

Also if you feel like it, please share your month 1 vs year 1 ratio because that could end up giving a tighter range.

If you enjoyed this article, please share it. Thanks!