The surprising effect of Steam Sales on non-discounted games

Fishing Cactus currently has three games on Steam. Epistory was released in March 2016, PC only,  Algo Bot was released in February 2018, PC only and Shift Quantum was released on the 30th of May 2018, PC and all three current-gen consoles. Note that when I talk about sales, I only talk about sales on Steam.

Epistory is enjoying a pretty good long tail. Algo Bot has not made many sales but has a good conversion rate and had a decent amount of exposure in the press and on Youtube. Shift Quantum didn’t have the exposure we expected yet. No major website coverage, no major youtuber played it: and so it gets very few sales. Upon entering the Steam Summer Sale of 2018, Shift Quantum didn’t qualify to participate (too recent for a sale), the two others did. My initial expectation was that the sales count would be even lower than before the Steam Sale.

I didn't expect Steam sales to benefit the non-discounted Shift Quantum - but I was wrong!

To give a bit of perspective, these bumps at the end of the graph represent about a 4x sales increase. To explain this increase, the first thing to do is check Youtube and Google, see if someone with a big following talked about you or your game. But there was none of that: only the usual background noise. So I fell back to the only tool left to explain why a non-discounted game could have increased sales during a Steam Sale: the traffic breakdown page.

Shift Quantum visits over its lifetime (~1 month). Normalized (0-100). Look at that peak! Higher than release day.

The Steam interface for traffic metrics is divided into two main stats: the impressions and the visits. An impression is when the game is displayed as a banner/title somewhere in the store. A visit is when someone lands on the game’s store page (either from Steam itself or from an external website).

Steam’s traffic stats confirm that the increase in sales was due to an increase of visibility - but without the intervention of a major media. What’s going on? Let’s compare with our other games.

Algo Bot visits over its lifetime (~4 months). Normalized based on Shift (0-125).

There’s a visit peak! But it’s so small compared to the one for Shift, roughly 10 times smaller. But even though the impact on Algo Bot is less dramatic, it’s still there. Let’s continue with Epistory.

Epistory visits over its lifetime (2+ years). Normalized based on Shift (0-2150).

Let’s start this comparison by stating that Epistory has benefited from a lot of exposure over the years and has a massive wishlist backlog (as in more wishlist than units sold through Steam). The huge spikes are made possible by tapping into those potential buyers.

At a glance, nothing is out of the ordinary. Massive spikes appear for each “special deal” (weeklong & daily), and we see modest increases on each “major” sale. When I say “major”, I mean a sale with some sort of gamification, mostly from the Trading Cards that users can receive in a variety of ways. Usually, this means either a winter or a summer sale.

Now, there’s one last thing to talk about for all the pieces to fall into place. On each of those traffic graphs, the most influential portion is the “Other product pages” traffic, shown in green. It’s an aggregation of several sources, like the “other games like this” found in all games pages, but it’s also where the “discovery queue” lies. Across all games, the green line is ~95% discovery queue clicks.

Once put all together I think it’s clear. During the major sales, players are incentivized to go through their discovery queue to earn more Trading Cards. That brings an influx of people directly to the game’s page. Having a Steam Sale brings roughly ten times more people to your store page, no matter your discount. The increase in visits is driven by the queue, while the deal spikes are driven by the homepage (everything visible on the main page & in the various lists).

All in all, I think one of the best ways to exploit this is knowing what kind of customer is coming to your store page. If it’s someone who knows your game from the media, they’re fairly likely to buy your game or add it to their wishlist anyway.

When it’s someone who just stumbles on the page, without even clicking a banner beforehand, they won’t have a lot of patience for a poor trailer or first image. This means the store page should be even more polished than the game. Have a hook three seconds in the trailer, include badass screenshots and make sure you have a killer pitch. Imagine having an elevator pitch for a publisher - and divide the time to convince by 10. Gamers are a way tougher sell than publishers.