Slitherine: Steam refunds not harming sales

At the beginning of June last year Steam introduced its new refund policy. Valve would, upon request, issue a refund for any reason, if the request was made within 14 days of purchase, and the title has been played for less than two hours.

Since the beginning we had mixed feelings about the validity of this new system as we were torn between its positive benefits and the misuse that some users could make of it. Nevertheless, we kept an interested and somewhat aloof eye on the system: we really wanted to understand if the system was good and how our titles were doing.

We have a total of 117 SKUs on Steam, so we have a pretty good overall picture of how our types of games would probably fare within the Steam returns policy. In our list of games, we have full games, expansions, DLC, bundles and packages, as well as a few extras like soundtracks and e-books that are usually sold in conjunction or in addition to a full game.

The overall return rate for all of the SKUs has been 1.99% on number of units sold and 4.16% on value. The difference in percentage here is due to a number of reasons, but mainly to the fact that units (or Steam keys) redeemed from external sites are non-refundable. So places like Humble Bundle, Bundle Stars or even our own site, where we give a Steam key to redeem the product on Steam on top of the game's serial number, have an impact on the overall numbers.

Full games are, of course, the ones that have a bigger impact on the general numbers, especially in value, but also in terms of number of units. Returns on units sold for full, stand-alone games, were 2.28%, while in value the returns were 5.07%. For DLC and expansions, in units sold returns were 0.32% and in value 0.46%. For packages and bundles, in units sold returns were 2.81% and in value 2.3%.

The full game with the highest level of returns in units had a total of 9.83% rate, while in value the highest was 12.01%. Every other game was well under 10% and outside of the first 10 games, everything else was under 5% return rate overall, both in units and in value. Bigger franchises like Panzer Corps and Close Combat had return rates of less than 2%, both in units and value.

These are the dry numbers and there is a lot of food for thought here.

First of all, it has to be noted that the impact of returns has been minimal -very much like in a normal retail store and even lower than on Amazon boxes, for instance. The rules for returning products are pretty clear, well explained and pretty prominent on the platform, so it can't really be the case that players are not aware of the system. It is evident that, at least for our type of games, that the purchases are made knowing what the product is all about and with a certain degree of consciousness. In terms of percentages, it is not even true that the more a game sells, the more returns there are (like after a big sale, for instance). Actually, the top 10 games, despite weighing around 50% of the total value, have a return rate of 4.37%, which is almost a percentage point below the overall average.

Rules are slightly different for DLC and bundles, but in these categories the number of returns is minimal to non-existent. Established games, where players know what they are getting, are less subject to returns, but this is still a good indicator that in the long term our audience does not really abuse the system at all. Bundles and packages tell their own story. On one hand these are big revenue spinners; they sell a proposition more than anything else, as they tend to contain a full franchise experience and they represent a big percentage of overall sales. They also provide a much smaller number of returns, so in general they make a very compelling proposition for players, but also a very good and risk-free revenue opportunity.

Percentages, however, do increase slightly for purchases made 90 days after release or later. It is interesting to see that titles suffer very little returns at launch, or just after launch. This is mainly due to the fact that sales made on Day One are specifically addressing a well-informed audience, waiting with anticipation on the release. What happens after sales, or big promotional periods is also very interesting: in general, returns don't go up, but they do for certain titles. There is no indication that the higher the discount, the higher the returns, but there is a pattern of higher returns for a limited number of games immediately after they have been on a sale.

On such a large number of titles it is normal that the average percentage is a good indicator of trends, especially for the genre of games involved. Of course, Slitherine doesn't have any causal game in the catalogue and the niche it covers has very specific needs and interests. It's important to note that the numbers above may not be valid across the industry as they are an average and don't take into account special cases. For instance, return percentages go up significantly when games are released with notable technical flaws or large bugs. This happened on a couple of occasions and it shows in the average percentages. The type of game plays another significant part.

Strategy games (and even more so war games) are very peculiar beasts. These percentages may be completely different for driving games, for adventures or RPGs. They may also be very different on big blockbusters or on games listed as Early Access. The Slitherine catalogue only features two games in Early Access, so numbers are not very significant or reliable enough to make a case.

In terms of countries, there is not a big difference in returns between specific countries. It's only fair to say that the US, UK and Germany play such a big part of the overall revenue (around 50%), that their role is dominant in these average percentages. Other countries do play a part, but don't manage to influence the final numbers. In general, some countries do show slightly higher return percentages, but nothing that screams out to be noted.

From our experience, after a full year, the Steam returns policy has proved to have a very minimal impact on the overall business. Now, we don't really know if it actually helped to improve sales, but looking at the numbers it certainly doesn't seem to be a tool for players to try a game and decide if they want to confirm their purchase at a later stage. It is indeed a safety net for gamers, but it's not perceived as some type of time-limited demo they can make use of - certainly not for our type of games.

We will continue to look at the numbers with a critical eye and keep our minds open to any changes and innovations. It is indeed important that companies with such a large user base like Steam keep trying to improve the marketplace. Some things just work better than others, but that's how industry standards are generated.