Jarrah Technology logo

May 25, 2016

Category: News

Author: Charles

Steamgifts Twitter Facebook Share on Google+

As my indie game nears a full release, my interest in indie marketing increases. Recently I was on the Space Game Junkie podcast, and one of the hosts mentioned that the Steamgifts website had not worked well for their sales effort. At that point I was planning on including the site for marketing Concealed Intent’s full release. There are a number of game giveaways on Steamgifts with messages stating they are for marketing purposes, and a few people have contacted me asking for keys to giveaway on the website (all refused, I can do it myself). So I wondered, how useful is Steamgifts as a marketing channel? Not finding any information online, I decided to try a few amateur experiments just to see how it worked.

For those that don’t know, Steamgifts is a website where members can give away copies of Steam (and only Steam) games. There are hundreds of games given away each day. It is not designed as a site for game marketing, but giveaway creators can include a short message, and these are often explicitly advertising related (for a game, or group, or YouTube channel, etc). The winner is chosen at random among those members of the site who have chosen to enter. Members are restricted in the number of giveaways they can enter, and must have at least $100 worth of games on their Steam account (so it is hard to enter a giveaway multiple times). There are currently a little over 900,000 members, of which around 135,000 have given away at least one game, and some many more. If a giveaways is open to everyone for over a day it will usually have 1000-2000 entries, depending on the game. Some very large giveaways (50+ copies) or desirable games gather 10,000+ entries.

Here is an example of a giveaway created for this experiment. The marketing text in the message is the standard copy for my Steam Early Access game, Concealed Intent. A brief description of the game, followed by a link and call-to-action, “please wishlist the game” (I thought a “please buy” CTA might be overly ambitious), then some social media links. The only way to enter the competition is through this webpage. So theoretically everyone who enters the giveaway should be exposed to the message.

I do very little marketing for the game at the moment, because as a sole developer I chose to spend most of my time on development tasks. This will shift in the next few weeks as full release nears and I go nearly full-time on marketing. However, during the experiment, I did no other marketing so views on the Steam store page are at a low and baseline level. Above is a graph of the store page impressions according to Steam, the grey line is total visits. Can you see the giveaways’ effect? There are 8 giveaways through this period, each separated by at least a day. The two most successful giveaways ended on the 11th May & 17th May. On those days there is a noticeable jump in direct visits to the store page (the light blue line).

Let’s see the raw numbers for Concealed Intent giveaways. I made two giveaways, each for a few days, a week apart. Both giveaways spanned the same days of the week. The number of visitors from Steamgifts to the Steam store page is measured through Google Analytics (which I suspect under counts visits slightly), and the number of extra wishlist entries is taken from Steam after subtracting the baseline growth in the week before I started the giveaways (1.5 per day).

Entries Comments Steam Visits Wishlist CTR Visit CTR +Wishlist
890 6 70 11 7.9% 1.2%
1027 4 47 8 4.6% 0.8%

I believe this CTR compares very favourably with other advertising. However, it must be considered that the cost of each giveaway is a copy of Concealed Intent, with a retail price of US$15, making those clicks extremely expensive by traditional measures. Although, that is only the case if the winner was going to buy the game at full price (or at all), probably unlikely given current indie game sales. Both winners did immediately redeem the game and play, which makes me happy. There were no extra sales in during the giveaway period, but I did gain a couple of Twitter followers (of course I can’t be sure they came from Steamgifts).

It is interesting to note that rise in the number of entries for the second giveaway corresponds with a drop in CTR. Perhaps the second giveaway included many people already aware of the game from the first giveaway. In any case, there may be diminishing returns on repeated giveaways. Thus they didn’t need to pay much attention to the text or visit the store a second time. I suspect a third giveaway would be even lower.

Prior to giving away my own game, I also tried giving different games for which I had keys leftover from bundle purchases, all with the same Concealed Intent marketing message attached. The *’ed rows indicate multiple overlapping giveaways with aggregated statistics.

Entries Comments Steam Visits Wishlist CTR Visit CTR +Wishlist
1269 1 6 3 0.5% 0.2%
2655 15 8 5 0.3% 0.2%
671 4 3 4 0.4% 0.6%
4296 22 8 8 0.2% 0.2%
829 3 2 3 0.2% 0.4%
686 4 4 3 0.6% 0.4%

Here the CTR is much lower. So low in fact I think that errors in counting visits or random changes in non-Steamgifts wishlisting will greatly skew the figures. Basically, giving away other people’s bundled game does not have much effect (although of course using bundled games means there is not much expense either). I would suggest relatively few people read the message, and most of the clicks are from checking out the game actually being given away before entering (rather than for the game in the marketing copy - makes sense to me).

So giving away my game on Steamgifts netted me at least 117 views of my Steam Store page and onto 19 extra wishlists for a cost between 0 - US$30. Other marketing opportunities available to indiedevs are: self-promotion on social media; asking web magazines and YouTubers for coverage; or paying hard cash for advertising. Given this, I would consider using Steamgifts giveaways again - just to increase awareness of the game. Although in small numbers (1 at a time), spaced with larger time gaps between them (at least a week), and only until awareness of the game reaches some as yet undefined level.

It is left as an exercise for a reader to measure the marketing effectiveness of giving away 50+ copies at once. I am not brave enough to try it with my game (at least not yet).

I am very interested to hear other people’s experiences or comments.

Update: I linked this post in the Steamgifts forum and an interesting discussion ensued over the value of 50+ giveaways or restricting entry to more engaged members. Here it is, worth a read if you found this interesting.

Comments: I have removed the commenting system (for privacy concerns), below are the comments that were left on this post before doing so…

unknown @ 2016-07-08 - comment deleted

Charles @ 2016-07-09 - Hi Anton, I agree that using the full retail value of the game is not completely valid, although I disagree there is no cost at all. There is an opportunity cost, but it will be much smaller than the retail, as it would only be the case if the person winning the game would buy the game anyway. So the cost per giveaway will be somewhere between 0 & US$15. I just don’t know where. the full retail is a handy maximum. Other numbers would be US$11.04 (the average amount paid including discounts and foreign currencies), US$10 (the last sale price), about US$6 (what I get per average copy after Steam & tax). The likelihood (so far) of someone buying after going to the game’s store page is around 1 in 10000 (and about 1 in 1000 for wishlisting, so hopefully the sales will improve after full release). On that basis the value per view is less than a cent! Now I’ve worked that all out, it is a bit depressing. I just had another check of the Steam backend and I can’t find the number of follows the game has. Sales, wishlists and views are all very clear, but I can’t see follows. Doesn’t mean they are not there somewhere, I’ll keep looking. Sounds like you have some nice friends :)

unknown @ 2016-07-09 - comment deleted

Charles @ 2016-07-11 - It is a grey area :) Steam rules say I can’t give out the financials. I think I’m probably safe given that. To work out Steam’s cut from the above you would need to know my tax rate, refund rate and chargeback rate, plus the exact amount. So it is not obvious. Then again details of Steam’s cut is not hard to find - Cliffski (much bigger dev than me) put it on his blog just this weekend (http://positech.co.uk/cliff… ) and Garry Newman put up his sales & refund rates (http://www.gamasutra.com/vi… ). So if that is ok, I think my vague numbers are ok :)

unknown @ 2016-07-12 - comment deleted