An Analysis of SteamGifts.com
By 
A little over three weeks ago, I discovered a user on Reddit who was giving away invites to a site called SteamGifts. I had never heard of it, but the idea of a site that focused on giving away games appealed to me. I remembered that I had a few extra game copies on my Steam account, and figured it'd be a good way to make sure they were put to use. After all, sending a gift using Steam is quite simple, since it exists purely as a digital credit of sorts - no files have to be sent around. With a use for the site, I requested an invitation code on Reddit, and received it shortly thereafter.
Once logged in, I began to poke around in an attempt to figure out exactly how the service worked. Creating a giveaway was pretty straight-forward - I simply had to choose the game, select the number of copies, and define how long it would be open. When a giveaway closes, one winner is selected from all the people that entered, and it's up to the person who created the giveaway to send the game manually through Steam. As for entering giveaways, there's a little extra mechanism that needs to be understood: in an effort to prevent people from entering every single giveaway, the site grants points over time. Reaching a maximum of 300, these points can be used towards entering giveaways - the more expensive the game, the more points needed to enter. Inviting friends to the site also uses a chunk of points, so it helps control how quickly the site will become populated.
Overall, I'd say the service is decently made, but currently suffers from some flaws - which I assume will get resolved in the long term:
- There is no search feature - so if you're looking for a particular game, good luck. I simply found myself paging through the public giveaways, hoping to find any game that was interesting to me, rather than the game I was looking for.
- The site is often slow to respond to requests - extending the time it takes to find what you want.
- No e-mails notifications are sent out when your giveaway is complete. In my case, I one day realized that one of my giveaways was likely complete - but had forgotten about it. When I logged in, I saw that it had, in fact, been closed 1 day ago. Not a big deal, but the winner had not received anything yet, and I didn't want him to worry, and others to question the legitimacy of my offer.
Now, the flaws I mentioned are all in the functionality, or lack thereof - but they're forgiveable given that the site is a free service, especially one which currently does not yet feature numerous advertisements. I expect these things to be resolved as time progresses; however, I am concerned about how this site approaches giveaways. To better illustrate my concerns, I'd like to start by examining the experience for users that enter existing giveaways. As an example, here's one of the posts I created:
It was configured to end in 5 days, and during that time, it managed to receive 55 comments, and 492 entries. Let me just clarify that last part - 492 people entered the contest, and one person won.
Is it just me, or is that a lot of disappointed people? Sure, they knew what they were getting into, but I can't help but think that maybe there is a better way. Out of curiosity, I glanced at the winner's profile - and here's what I saw:
A total of 3 gifts won, despite having 340 entries. That's roughly 113 entries per game. Considering the flaws I mentioned earlier, entering all of those contests could have been a rough experience for the user. Is the time investment worth it, considering the low chance of winning, and the flaws in the website? Keep in mind, also, that the chance of winning will decrease as the site gets more popular - as there will be more users, and thus, more competition. To help offset the low chances of winning, users seem to be encouraged to enter multiple contests, just in the hope that they will receive something free. I originally found myself doing this, as it simply felt logical - my time is valuable, so why would I spend it struggling to find only the games that I really wanted? It was simpler to enter any contests that were the slightest bit interesting, rather than making calculated decisions. Of course, I quickly realized that this wasn't the most ethical approach, so it lead to some self-reflection.
First of all, I wondered: was it even necessary to have more than one entry per giveaway? What if you could create a giveaway that closes immediately after one person enters it? If that could happen, only two people would be involved: the person giving the game away, and the recipient of that game. The giveaway wouldn't gather much attention from the public, but it will have served its purpose - giving away a game to a person that wanted it. As a bonus, there is a low probability of people being disappointed. Interestingly, SteamGifts isn't too far off from offering this functionality. It is currently possible to configure a giveaway so it closes in one hour, but no less than that.
While the above would help reduce the number of disappointed users, another key factor must be considered - exposure. I originally didn't feel like I needed recognition for giving the games away - I honestly did it because I realized they were gathering e-dust. Yet, I received 55 comments for giving away just one of game. After the contest had ended, I realized that it may have been beneficial to leave my post up for multiple days - for one primary reason: my Steam profile contained a link back to my blog. Now, as true as this was, I didn't feel like adding to the number of disappointed users, solely for a chance at additional traffic.
Ultimately, I decided it would be best for giveaways to strike a balance between getting exposure, and preventing entries from reaching extremes. I felt that one day was probably a good maximum time-frame, but of course, that maximum was closely bound to the popularity of the site. Early on, one day might bring in hundreds of entries - but in the future, maybe thousands. The more entries, the less chance of winning - which encourages people to enter any and all contests. What if you could have those statistics handy when creating a giveaway? By knowing how many entries are projected, would the decision be easier to make? I think so, but still feel like there's something missing to make the service work properly. As it is stands, I feel that users are too likely to entry-spam - all to improve the chances of them winning something. It's a pretty unique challenge, and I hope they manage to find a solution. In the meantime, I find myself less inclined to use the service as a means of receiving gifts - sending gifts, on the other hand, can work well, especially with the above ideas considered.
If you'd like to try out SteamGifts, I'd be happy to provide invitations - simply contact me at mref...@gmail.com. Even though I had some issues with the site, it's still worth checking out. My interactions within the community have mostly all been positive - users are quick to thank the guys creating the giveaways, and are also instrumental in warning their peers about fake posts.
UPDATE: I announced this blog post on the SteamGifts forum, and interesting comments were posted. One particular commenter was an administrator of SteamGifts, and he pointed out that they were working on improvements, some of which were particularly tricky. The administrator (named "cg") also called me out on a logical error that was present in my post. He made a reference to particular part of my post:
"Keep in mind, also, that the chance of winning will decrease as the site gets more popular – as there will be more users, and thus, more competition."
And replied with the following:
"Not true. Entries will increase, but you're missing the fact that the number of giveaways will also increase. If new users are as generous as the last, odds of winning stay equal. Right now we have 25k members and 8k gifts. We've been floating around that 3:1 ratio for a long period of time. Those are the numbers you should be keeping an eye on, not the total number of entries."
He was, of course, entirely correct.