TL;DR: Yes. submitted by
Also, sorry for taking so long with the post. I wanted to do it already a few weeks ago, but life (and Stellaris >_>) got in the way. Previous post can be found here (mostly about itch.io stats and preliminary Steam impact)
In this post I'll go into the whys and hows of having a demo for your game and what it means for your visibility, reception, customer feedback and, also, wishlists & sales.
As previously, this is going to be a long one, but I'll at least try and make this entertaining. The game in question is Death and Taxes
. I'm a developer on the game, and it released earlier this year, on February 20th. The project, as it stands, has been profitable and is a success, in more ways than just financially.
Let us continue from November 23rd, 2019 - January 20th, 2020 (a month before release)
Lets start with the data.
A VERY strong disclaimer here: the numbers that Steam gives you can be pretty vague. It's really hard to pinpoint where users are actually
coming from, meaning you don't know what they clicked and where. Google Analytics can help with that, but it also underreports its numbers. YMMV, but if you multiply your Google Analytics hits by a factor of 3 or 4, then you'd get the actual numbers, more or less. Steam visits from November 23rd, 2019 - January 20th, 2020
- Full disclosure: The first "brown peak" is the first postmortem post. That actually drove our visitorship up a fair bit within Steam as well, so, thank you, reddit!
- The second "brown peak", near new years eve, is from getting featured on Game Jolt, and the days before and after as well, till January 9th, where some other traffic started coming in from various sources.
- The first "purple peak" is from getting onto the "More Like This" section for "Papers, Please" (we think, but are pretty sure of this), the subsequent visits from Other Product Pages are from the generally increased exposure due to this
- The peak on the 20th (the last metric) is kind of a perfect storm, where we had lots of visits from other games' pages on Steam, a lot of visits from inside our demo (since we included a link to our store page in the demo) and minor amounts of visits from all of our social media sites and storefronts. This is where you want to be with your game at least 2-3 months before release. For us, it was a month, but that was okay.
So... Where do we actually get seen (according to Steam, at least)? Steam impressions from November 23rd, 2019 - January 20th, 2020
- Direct Search took off at December 1st. You can make a guess at what drove it at this point. Go ahead, I'll wait. I'll give you the answer later on.
- A lot of impressions came from people browsing Tags. Now, this may be a big, big common sense thing, but having proper, descriptive tags for your game on Steam is of utmost importance. Check out this article and this talk by Chris Zukowski (bonus talk about copywriting for your Steam page) to see how and why this is important. Aside from taking my word for it. And these stats. >:|
- After New Years, when Game Jolt's feature highlight took us off, Steam started pushing us more and more as well. We are getting more domestic traffic, driven mostly by "Coming Soon" pages and the "Tags" page. Why did this happen? Here's the simple explanation: Steam supplements your own generated external traffic (like we had if you scroll up a bit) with showing your game off more to its users. External traffic is extremely important, and this drives your visibility inside the store as well.
As a sidenote: Steam will give you a visibility boost about 1-2 weeks before your release date which is one of your main funnels if you're a tiny indie (like we are), in the form of a spot in the Popular Upcoming category. People actually do see that and they do use that. We haven't hit that part yet in this post, so stay tuned.
So, did you already guess what made the Direct Search take off? Boy, have I got the answer for you. The demo played a significant role in getting people talking about the game and searching for it. Demo impressions from November 7th, 2019 - January 20th, 2020
This might not look like much, but it actually does matter. We started getting more and more page visits, impressions, because we got shown off on the Free Demos hub constantly. This is just organic traffic. From the demo hub. Nobody was in particular looking for our game, but an educated guess looking at the data is that Steam was starting to show it to people who actually seemed to be interested in this game. The organic demo impressions stayed fairly stable throughout this period after the initial peaks.
And.. just compare the scale of impressions of the Demo vs the base game.
it's pretty fucked up how much more eyes are going to be on your game with the demo.
More than 10x that of your full game's page impressions, potentially. One thing to note about general marketing knowledge: repeat exposure is extremely important. If you see something once, you might not think much about it. If you see the same thing over and over again, you're going to get curious. The demo does exactly that. It sticks around.
So now, you might be wondering, what about the wishlists, then? Glad you asked. Wishlist data from November 23rd, 2019 - January 20th, 2020
Pretty direct correlation to our traffic, generally. This also correlates well to our demo users. Demo players from November 23rd, 2019 - January 20th, 2020
Installs don't really matter as much as actual players. This correlates well to the previous graphs, but is more slanted towards weekends or holidays (obviously). Oh, and you wanna know the best part? Almost 25% of the players clicked on our demo link, which had a call to action to wishlist the game.
Here's how it looks in-game: \"WISHLIST US ON STEAM PLEASE FOR THE LOVE OF GOD MY LIVELIHOOD DEPENDS ON IT\" can just be turned into a polite \"Wishlist us on Steam!\"
It's a non-intrusive way of giving an incentive to remind your players to wishlist the game. Just one click of the button. Then they of course have to click the Wishlist button in Steam as well, but from what we can see from the data is that the vast majority managed to do that. We have no way of confirming this due to lack of analytics from both Steam and Google Analytics suites, but looking at the correlations, this looks like the likely case.
In addition to that, we had a short 3-step questionnaire with yes/no answers whether or not people liked the demo or not, whether they wanted to wishlist or not (it would again take them to the store page) and whether or not they wanted to give us feedback (by e-mail).
- Around 80% of people marked that they liked the demo
- We got around 2 e-mails per 100 people playing the demo as feedback
- We got around 25% of the people that played the demo to wishlist the game
We actually got a whole lot of feedback about the game during our extensive time having demos out. The full game would not be the same without it. The feedback we got was invaluable and I'm very glad we did it. We ended up with a much better game in the game by having done this.
During this time period, we went from 300 wishlists to 1400. wow
Conclusions for November 23rd, 2019 - January 20th, 2020
- Generating external traffic is imperative to generate additional domestic traffic on Steam
- Getting featured on other storefronts has a very positive long-term impact
- Having a demo out is going to get way more eyes on your game, including repeated exposure
- Having a demo out is great for gathering feedback and seeing how your game performs
- Having a demo is good for gathering wishlists
- Having a demo is good... period? Lets see
January 21st - February 20th, 2020 (release day)
oh my god what is this bullshit
Lets look at January 21st - February 10th, 2020
instead -_- ahhh much better
Okay, lets just look at this first and then see what the (bullshit) peak is about later on.
Stable territory for a few weeks, and then we hit February. What gives? Well 2 main things: a Curator review and YouTube videos. This drove up our visitorship immensely, and of course, Steam gave us a shoulder by also increasing traffic on their end, which was a nice surprise. So lets take a quick look at the main page impressions and demo impressions, and then see what we can derive from the data. Also, demo pages don't have visits, they're all bundled into the same main page visits. Main page impressions from January 21st - February 10th, 2020 Demo impressions from January 21st - February 10th, 2020
So this is great. We're a month before launch, and we're finally getting some traction. We started our marketing campaign by sending a lot of people keys to early versions of the game and also more and more people started playing our demo. Don't believe me? Here you go: Demo players from January 21st - February 10th, 2020
And again, maybe we should check how it matters in terms of wishlists? If we act on the basis of previous assumptions, we could say that there should be a correlation with wishlist increase as well? Well, absolutely. Getting videos made out of your game that can only exist by having a demo out and also having the link to the page under the video where people can go in and play the demo themselves
really seems to work well.
At this point we are doing very well. Before getting the big blowout, we are still getting stable wishlist increase, a lot of it being driven by demo availability and solid conversion from demo player to wishlist. Then the real kickers started to come from bigger youtubers and streamers. Our goal was to get 10k wishlists before launch (the golden number right now for indies, more or less).
We were doing a combination of sending e-mails with Steam keys to the press, to influencers and using Keymailer to distribute keys to influencers as well. In addition to that, the entire time we were doing around 3 dedicated social media posts per week (1 devlog, 1 #screenshotsaturday and 1 #followfriday usually). Doing those things consistently definitely helped grow this kind of following that we had. We also maintained a Discord server, but that didn't take off until the game was released, but it was still pretty active.
Conclusions for January 21st - February 10th, 2020
- Having a playable demo was great for player retention after they found it on a youtube video/stream
- Having a demo helped us immensely with improving the game based on feedback
- It absolutely, positively helps if influencers, in their videos, have a call to action to visit the devs' steam page, to wishlist the game, to try out the demo, etc. The performance from these types of videos was super helpful. You most of the time have no control over this, so it's just hoping that they do it, and when they do, it helps a lot.
- Steam curators are a thing, and if you can get a hold of with some of them who have a good audience, it helps with general visibility, too. Try to find curators who play your type of game!
Okay, so this day was basically.. an anomaly. We had ridiculous amounts of visits, but practically no wishlists that day. We tracked it down to some curators but it didn't really make sense. I think that it was Steam trying to push the game to a lot of people, or it was a bug in the system. I'll not include this day in the analysis, because it pretty much ruins a lot of the data that we have if it's in reference.
February 11th, 2020 - February 29th, 2020
Just a little bit before launch... expectations were reasonable, our performance was good and we were getting solid coverage by influencers. We still continued reaching out to press and influencers and got positive results. A lot of influencers, especially big ones, found us on their own, too. Our release date was set to be the 20th of February, 19:00 CET/10:00 PT
. We hit the button a little bit after that, but it worked well enough. There were some problems the previous days (and the same day) with some people releasing games on Steam and the Buy button not appearing, which royally screwed them over. Luckily we had zero problems.
Lets check the data.
Visits from February 11th, 2020 - February 29th, 2020 (While I'm writing this part, I discovered that my laptop's charger broke down and I'm on battery power. So uh. Forgive me if this suddenly gets a lot more hectic.)
As expected, Steam starts to pump a little bit of visibility for you immediately during launch, as you're going to be shown off on the "upcoming" page if you have enough wishlists or if there's not too many competing games coming out. This means you're going to be on the front page. And as you can see, that helps out A LOT. Check out the brown line. We released on a Thursday, and we peaked on Friday, as expected, but our launch week performance was solid.
We got the long tail of people watching all the influencer videos come out, both big and small, and then of course having a good roll due to constant hammering of a release countdown on every platform we had.
And before I really
hammer in the point why, at least in our case, a demo was a great idea, lets look at the last stats.
As I said, being on the front page is a big deal. You get so many eyes on it. MILLIONS of impressions every day. This is for the full game.
So what happens when your game starts to get played by a lot of influencers? People start talking about it. And what happens then? "Friend is In-Game" notifications.
This is for the demo. Obviously now the full game's impressions completely dwarf the demo impressions, but this still converted very well to players.
So lets see the player counts next.
SO MANY PEOPLE TRIED THE DEMO. We had unprecedented numbers of people coming in from the demo's in-game link as well. Literal thousands.
We see the same correlations as in the previous datasets. There is co-mingling between demo player counts, visits, impressions, wishlists, etc. Same patterns.
And how did this affect wishlists? Pretty well. Mind you, that the wishlist activations are not the same as total sales. Nice.
We hit just over 20k wishlists the day before launch. This gave us the confidence that we're on the right track and validated a lot of our efforts and ideas. The sales target for the first week was:
- Worst case: 1k sales
- Best case: 10k sales
- What actually happened: 20k sales
Okay so at this point we're pretty much over the moon. Release party, I drink too much, throw up, find out that there's a critical bug (or 5) in the game, fix it the next day with a massive hangover. The usual.
I think I vomited in 3 different toilets during that night.
This is how much we sold in total in that period. More than we imagined!
Everything went well, and we sold higher than expected. This does not happen that often in game development, so we were elated to see the results.
And now for some takeaways from this period.
Conclusions for February 11th, 2020 - February 29th, 2020
- Getting to, and staying on the front page of Steam is SUPER EFFECTIVE.
- Having bigshot gamers play your game, even if it's a demo, and everybody's Steam friends seeing that game being played means that you get a lot of people, at the very least, acquainted with the name of your game
- I haven't written this before, but we had a launch discount of -25%. This was to give back to the amazing support we had from our early adopters and fans, and people who tried the demo. This was also a strategic choice that led the game below $10 for all currencies, and thus, we got traffic from Steam's homepage from the "Best under $10 section". It really gave us a looot of views (and yes, views, not impressions).
- Don't have any intoxicating substances near you when you launch :|
- Demo traffic is still significant. We got even more feedback, and the vast, vast majority of it was overwhelmingly positive.
And this is probably the moment you've all been waiting for. Is it worth releasing a demo for your game? I again say: Yes. Yes it is. Did having a demo affect our sales? (as this is the most controversial thing that people like to point out with literal zero f***ing evidence on this subreddit):
It's hard to say definitively as impulse buys are indeed a thing, but it certainly did not drag anything below expectations. Rather, I think it worked out better for the game in the end, than *not* having a demo.
And I will tell you why.
First of all: the most important thing we got from the demo was live feedback from a wide, wide variety of players. The demo we have currently on Steam is the 7th iteration of the game's demo, and each of those iterations has been different. We got feedback we would never have got if it weren't for the demo.
Secondly: it allows players to see what the game is about, and either confirm or contradict their expectations. I'm putting this in bold because it's important: We have an all-time 92% positive rating on Steam currently.
We have had feedback in the past, which said that the game was not what was expected, when played, and it either a) turned the player away - basically avoiding a negative review on top of a refund
or b) surprised the player and actually converted them to a sale
. I am more than certain that our review score would be less than it is today due to this being an option. The negative reviews we do have, a measurable amount of them speak of broken expectations. This is a hard game to market and explain, because it's not exactly like anything else out there.
I will have to elaborate this a bit more. Impulse buys are great, until your customer actually plays your game
. You have made money, but now you're at risk of a negative review. And you want to know what happens when a game falls into the dreaded "brown" review score category (Mixed
, I think it was called)? Their visibility on Steam tanks.
That's an actual thing, explained by Valve
and reported on by PCGamesN
. This is an irrefutable fact on the workings of Steam's algorithms.
The consequence of this is that if you don't validate your game idea enough and if your game is at risk of being perceived in a different way than intended (Death and Taxes is an experimental game for sure), you might just get buried. No amount of sales matter then, because people won't buy a game they don't know that it exists. Higher review score means more traffic and more sales
, it's as simple as that. Check Chris Zukowski's finding's as well about this.
Thirdly: Repeat exposure matters.
The amount of domestic traffic on Steam is staggering and it's STILL breaking its own records
. The more you pop up on that immense platform, the more chances you have to convert a browser into a buyer and most importantly: into a player. We have had anecdotal, but still significant feedback about our game popping up in various places, this includes the demo. It has been mentioned in reviews, it has been mentioned in direct feedback sent to us, it has been mentioned by people joining our Discord... It is a deciding factor for some people and we decided to use this in our favour.
Key takeaways from our performance with the demo on Steam
- It's amazing for exposure and connecting with your target market
- It's great for retention, especially if you have people coming in externally, too - use the demo as a method of transition for players to turn into fans
- It's an invaluable tool for feedback (updating the demo is a great idea)
- We probably dodged a lot of negative review bullets by having a demo act as a mediator when people are on the fence about it
- Demo traffic correlates with increases in impressions, visits and wishlists
Some thoughts about making a demo
- If you have a game that has its mechanics exhausted by the demo, you would need to properly focus on the experience of it to keep the player wanting more. Having a demo that exhausts the full game, may in fact end up not working well for you. YMMV.
- If you have a demo that you get vastly negative feedback from - you should probably focus on polishing the game, and make a new demo when time permits.
- Add ways for players to keep interacting with you or the game directly even after the demo has ended for them. Add a button that takes them to your Steam page and another one for your Discord server, for example. It does work.
- Don't spend too much time on it. If you see that making a demo for your game would be as much work as maintaining the full game, it might not be worth it depending on your scope. I elected to develop the game in a way that I could extract a demo version from the full version with the flick of a button in the Unity Editor. This made updates easy and kept everything cohesive.
- Ask yourself what you want to achieve by making a demo. Our focus was feedback first and wishlists second. We wanted to use the possibility of a demo to make the best game we could within our constraints.
And last, but not least.... If you ever are afraid that you will lose on sales if you release a demo - maybe ask yourself, did you ever deserve those sales in the first place
? Or do you just want everybody to buy your game regardless if they would like it or not? Impulse buys are not what drive meaningful interaction and positive relations between developer and player. If you think that impulse buys are a great way for people to accidentally stumble upon a hidden gem... then why not have them do that anyway? For free. With a demo. And if they do really like it - they will buy it.
If you have any questions, feel free to ask me here. I would love to get another lively discussion going like the previous post had! I hope this is an interesting and entertaining read and that the data and analysis I have provided is useful to you. If you want to see any more data, let me know as well. I'm happy to share anything I can. We don't have a publisher so we have very few limitations on what we can or cannot show.
Have a great weekend!