Watch this person decide that they are going to buy this game

At the end of the day, she bought it.

The key things to notice in this video

How I watched people shop

Two months ago I ran a study where I watched regular Steam users shop. You can see all my findings and recommendations here. However, one issue with the past study was that I observed participants looking and adding games to their wishlist but never buying. Over and over I heard “Well I am waiting for a sale before I really buy anything.” I was missing a big part of the picture. the one that actually matters most: how people actually decide to buy games. So I committed to a follow-up study during a sale so that I could see how their behavior changed. 

The following is my study of 6 Steam users during the first two days of the 2019 Halloween sale which took place from October 28th  to November 2nd 2019. In general, I followed the same methodology as the previous study which you can read here. The one difference with this study was that I tested only during the sale. I wanted to make sure I got the first impressions from the participants so that I didn’t miss any key decision making by them. To do this I instructed participants not to open any “wishlist” notification emails, Steam app notifications, browse the store, or talk to friends about what is on sale until we were on the 1-1 screen share. 

I tested six different participants. Three of the testers were from the last test (Participant C, D, and F) and three were new (Participants H, I, and J.) During this test I watched users shop and ultimately buy 12 games. All participants were paid a $25 honorarium for 30 minutes of their time (which is standard practice for consumer research studies such as this one.)

My research found that during sales events participants enter a special qualification buying behavior where they start with their existing curated wishlist, then reassess each game based on the following:

How shoppers look at Steam

In general here is a step by step process of what I observed the participants do during a Steam sales. 

Receive notifications

Participants said they typically learn about sales because of the following

Check the Event Sales Page

This step and the “check their wishlist” step are interchangeable depending on the participant. Participants skimmed through the page to see what types of events are occurring or what games are on sale. I tested during a Halloween themed sale so horror and suspense games were on sale. Not all participants like these types of games so some were more excited about the offerings than others. 

Check their wishlist

(Sometimes this step was done before they went to the event sales page) I saw participants sort their list by the discount amount (not dollar price) and reacquaint themselves with it. If they liked the game, they added it to their cart

Cart review

At this point, the participant would just.... wait. They were aware that the sale would be running for a few days so they didn’t rush. They would use this time to message their friends who had played the game to see if they would recommend it, shop for more games they might have missed, and to debate whether they really wanted it or not. 

Here is a detailed breakdown of how they would manage these steps
 

How they browse the main sales page

Participants followed a pattern that was very similar to the browsing behavior I observed during the last study. They looked for interesting capsules, hovered, looked at screenshots and tags. I noticed that most participants gravitated to games that they were already familiar with. 

Analysis and recommendations

I do not have any recommendations that are unique to the sale because the same rules apply that I found in my last study. You can get all those recommendations here.

I did notice a couple of participants gave a second look to games that had special badges for the Halloween Sale. So, decorate your capsule. 

How they use their wishlist

Note that some users started with their wishlists and others with the Sale Page. 

I observed many users do what Participant D did in the following video: they would sort their wishlist based on discount. Note they don’t care what the lowest price is, they are just looking for the biggest percentage off. 

Other users like Participant C have such a big wishlist that they don’t even bother with the Steam interface. It just doesn’t provide the features needed to manage that long of a list. Instead, he uses the site GG Deals because it has something called historical low. Participant C liked to make sure that he is not missing out on any once-in-a-lifetime deals. If a game is regularly on sale for the same percentage he doesn’t bother buying it during the sale because he knows that that game will be on sale again later. This is "fear of missing out" in action.

I also observed a lot of participants totally surprised by what is on their wishlist. It was as if they were looking at the game for the first time. 
 

Analysis and recommendations

Don’t take wishlists for granted. If some wishlisted it, you might feel as if that is money in the bank because they will eventually buy it. But that is not the case! When your game goes on sale all those wishlisters are going to look at your store page and say “what was this game again?” 

Before big sales, update your store page. Think of better ways to describe what your game is and how it plays. Maybe you have been featured by some major streamers that did a good job showcasing your game. Make sure that video is easy to find for people visiting your page. Remember how Participant H went off to google to find more info about the game?. That indicates the page wasn't doing enough to describe the game. You want to make sure your page does a good job so they don’t leave Steam.

Also, if your game has been out for a while, consider doing minor updates or announcements before the next major sale. Clean out the cobwebs, put up a fresh coat of paint on your store page. Maybe put up some new screenshots or capsule art. Maybe even time an event to appear more alive.

Should you do frequent updates or not? There is no clear answer because it depends on your strategy and price. 

Frequent updates can keep your game front-of-mind for many players. Note how Participant D said he was familiar with Talos Principle because it is always on sale. Just be sure that you make your game look different and alive with every sale.

Steam DB is a site that tracks every price fluctuation of a game and charts them. This is the sales chart for GTA V. Expect that a portion of shoppers are looking at the price chart for your game.

The downside to frequent sales is that you lose the power of "fear of missing out." If customers know your game will have the same sale price again in 2 months, they might skip it during a sale. We saw this behavior with Participant C who filtered for games with “historical lows.” 

Historical lows for your game are very important. GG Deals dedicates a whole widget on their front page just to track them. 

So be very strategic with your “historical lows” because that is a non-renewable resource. If you go straight from 10% off and then 75% off the next sale you just forever forfeited a whole bunch of “historical lows.” Instead,  patiently discount by 10%, 15%, 20%, 25%... you then have many more chances to appear in user’s “historic low”  filters.

Also save your historic low discounts to be timed during the 2 biggest Steam sales: Winter and Summer. I heard many participants say that they really save up in anticipation of those sales. While I haven’t tested users during those big sales yet I suspect that there is an additional urgency around those two that will make your “historic low” sale price have more potency again because of "fear of missing out." 

How they use their cart

I observed participants quickly add good discounts they found on their wishlists or the sales page onto their cart. However, the cart didn’t mean that they are ready to buy. For the participants, their cart served as a secondary, temporary, wishlist. 

Many participants indicated that they step away from Steam and have a think about which games to actually purchase. For example, Participant D said he thinks about which one he wants to play, he consults with friends (see the section below for more on that) and then checks more reviews. 

There is a big risk here for us as developers because there is a real chance that people never come back before the end of the sale. Life gets in the way. Listen to Participant J as she mentioned that she forgets to go back. “I will probably end up forgetting when the sale ends and missing the sale completely.”

Although the Halloween sale has only one big discount round, Steam has trained users to wait to buy just in case additional items might go on sale. This is because the Summer and Winter sales have rolling discounts and additional unlockables that proceed throughout the week. Users don’t want to miss out on these so they might wait. 

Analysis and recommendations

In the wider world of online retail, there is a well-known concept called “open cart” or “abandon cart” which are sales lost because a customer leaves the site with unpurchased items left in their cart (about 75% of all customers will leave an item unpurchased). Many retailers fix this common phenomenon by sending a friendly reminder email a couple of days later. Some studies have found that as much of 10% of all lost customers can be brought back by these cart recovery emails. For examples of recovery emails retailers use, see this article.

Unfortunately, Steam does not have a follow-up email sequence. It would be so great if 12 to 24 hours before the sale ended customers with games that were left in their cart got an email alert that said “Sale ends in 24 hours! Look at these games you could be playing!”

Since Steam does not have any abandon cart system, we need to step in and do it ourselves. So in addition to your initial “we are having a sale” broadcast, also plan on doing a second one 24 hours before the sale ends and then do a 3rd one 12 or 8 hours before it ends. Steam has given us a new suite of tools to send alerts to people who have wishlisted or follow our games. Use them!

If you are uncomfortable sending 3 buy-buy-buy messages to your audience so soon after each other, you can alternatively make a fun event out of it. Consider doing timed reveals throughout the sale. Schedule a tournament that ends on the last day of the sale. You could even encode seasonal modes that unlock every day during a big sale and you send out alerts announcing what is happening in-game. These limited-time events encourage people who might have your game in their cart to get off the fence and buy because they don’t want to miss out on all the fun everyone else is having. Again, channel the fear of missing out. 

How word-of-mouth marketing makes or breaks your game 

I often hear indies say “well word-of-mouth marketing is still the most powerful thing we can do” and I would just nod in agreement but to be honest I never really understood what word-of-mouth actually was. What does it look like? When does it come into play in the buying process, and what can we do to encourage it? 

I think I actually witnessed it in this round of testing. 

My last study was done while there was no big sale on Steam so the participants were in a “window shopping” mindset. There was no pressure to buy so the participants were just wishlisting games with no intention of buying right away. In fact, in my first round of testing, not one of the nine participants added a game to their cart or purchased anything. 

However, I conducted this second study during a sale and I saw a big difference. Participants were now looking with a mind towards actually buying something. This lead to a new behavior that I didn’t see before: They consulted their friends at every turn. 

What I saw was many participants frequently consult the “Is this game relevant to you?” widget because it lists which friends had played the game they were looking at. 

In the previous study I found that participants didn’t stray from their favorite genre. They knew what they liked and weren’t interested in venturing far from it. This time I noticed that they also are keenly aware of what genres their friends like and that informs their judgement of the game. Rewatch the first video with Participant J. Notice how she said she and her friend have very similar taste and because her friend played it she would probably like it too. 

Watch how they were pretty quick to add a game to their cart but they “wanted to check with their friends” to see if it was actually good. Participant H showed me the discord message she sent her friend. This little interaction is happening millions of times across Steam and you better hope people who played your game are telling their friends “go for it”

For multiplayer games, several participants check with friends to see if they also wanted to play. 

Some friends also serve as ambassadors to games. If you rewatch the first video with Participant J notice how she said she has a friend who always recommends games to her and her other friend. 

Word-of-mouth (or lack of) can also work against you. People are well aware of the genres each of their friends like. Note how participant H said that she probably wouldn’t like the game Project Winter because her friend played it and he has pretty different tastes than her. Similarly, Participant H was intrigued by Chicago 1930 but when she realized that none of her friends had even played it she immediately moved on to the next game. Participant D mentioned he hasn’t purchased Swords of Ditto because his friends told him the gameplay wasn’t good despite the great graphics.

Analysis and recommendations

So if solving the discoverability problem isn’t hard enough, here is another hill we indies have to scale: the friend test. Expect your potential customers to ask their friends if it is a game they will like. 

So what to make of this? Realize that just because someone bought your game it doesn’t mean your interaction with them is over. If you show good, ongoing support, provide good customer service, and interact with your community, you will build goodwill with them over the long term. You want them to think back on their experience with your game, you, and your community in a positive light so that when their friends ask them if your game is worth buying they don’t have any reservations about recommending it. 

Consider surveying your audience and asking them on a scale of 0-10 if they would recommend the game to a friend and why or why not. This is called a Net Promoter Score and is a very powerful measurement for any product. Use it to drive future improvements to your game.

Now, what do you do if you are just starting out and your game isn’t very popular and just like Participant H that had none of her friends played it? It sounds like a chicken and egg problem. Faced with that I would start giving your game away as much as you can or as part of a contest or as part of bundles. Target players who would like your type of game. Even if a person redeems a free key, it still counts as them owning it and so your game will start appearing in their friend’s “is this game relevant” section.

What about Participant J and her friend who “had really good taste” and was “always recommending games?” How do you encourage this?

While you are building a community try to attract these taste-maker players and really spend extra effort to cater to their needs. I would give my games away for free to these types of people if I knew that they would later recommend them to their friends. In the restaurant industry, restaurateurs will often throw elaborate and exclusive parties for people who review a lot of restaurants on Yelp because they know if they have a good experience they will recommend them to others. I wrote a whole blog post about supertasters You should do everything in your power to keep them. 

Throughout my research, I observed that participants are very genre monogamous. I always wondered how anyone tries anything new. My hunch is that there are lots of people like participant J with a friend with “good taste” they are the ones who convince them to try a game outside of their comfort zone. 

Final purchases

As mentioned above, none of the test participants bought the games in their cart right away. Instead, they said they wanted to think about it and come back later. So I wrote down what was in their carts at the end of the test session (listed in "Games in cart" column). Then 24 hours after the sale was over, I messaged them again to ask them what they bought (listed in "Games actually purchased" column) and why (listed in "Reason given" column.) Note their actual words (SIC) are in quotes. My clarifying notes are not. 

Here are the results.

Summary

The typical Steam purchasing funnel looks like this

Discovery > Wishlist > Cart > Friends check > Final Cart Filter > Buying.

Think about how hard it is to sell a game. Even if your game gets wishlisted, even if it miraculously gets added to a person's cart, you are still 1 click away from being kicked out.

If anything this round of testing taught me was that Steam really is a social network that happens to sell games. Nearly all of the above participants had some form of interaction with a friend before they made a purchasing decision.

So much about a purchase depends on having friends who also play it. Studios would benefit by treating Steam not just as a store to place their game but as a place to interact with their players. I think this is also a contributing factor to why there are such strong backlashes when a game moves off the Steam platform in favor of an upcoming store. It is disrupting a player's social circle.

Additional help

If you found this study helpful, I publish a weekly newsletter covering the basics of indie game marketing. I talk about how to find and talk to your target audience and get them to fall in love with your next game. If you are interested, sign up you will get a free copy of my ebook about marketing basics.

Join my mailing list