If it’s going to be a paid game, I might suggest waiting for the whole game to be done, then releasing the full game (paid) along with a demo on the page (free).
If it’s a free game, I might suggest releasing a demo first.
Of course, there’s multiple schools of thought. And it’s not really the case where one way is guaranteed to bring more traffic or interest than another.
Couple of reasons, really:
The game might still change after the first demo release, creating a little confusion. Confusion which generally would be more likely to be dismissed if a free game (in my opinion).
While you might build hype initially with a first demo release, the hype will tend to go down over time. With the potential loss in hype, it might take good communication (and a bit of extra work) to keep people who would want to buy the game interested until the full release.
Put yourself in other's shoes. If you hear for a game and like to play/buy/read about it, would demo be helpful for you?
To me, Demo (limited in levels not in time) is great asset to spread the word of your game and build a fanbase, but is also the best way to get something much more valuable than that. First impression is key of your success. If the first 1-2 minutes prolongs to 2-3 minutes attention, you have a good chance. It can be used to gather data, and player feedback/critics/suggestions.
But the most important is what type of game you make. Not all games are suitable for demo.. Good luck!
For Star Story, Dyrt and The Fairy Tale I did release demos showing the first parts of the game while the rest was still in development, in order to have a bit of pre-promotion indeed. Sometimes a little feedback can then also help, as some fixes or features can be easy to implement in an early stage of development, but be a downright disaster when the project is almost finished.
Now these were all games on which I had a very clear view on the entire story line and such before even starting on that. In Luna's Father I went a bit more "on the fly" and as such too much uncertain factors to go on. As soon as my revamp on Dyrt is finished, I should see to finish it though. I already implemented the final boss fight, but I was always in disagreement with myself about the game's ending.
In my opinion, this game did it right. https://tarkovsky.itch.io/fumes Released a free demo, advertised a release date on Steam, released a cute parallel game called Rust Mites, an hopefully will be keeping the demo free after the release. Because of the long demo stage, it has a small and dedicated community who are huge fans and will likely buy the full game. Parallell games are a neat idea. Make a minigame with the same graphics style and roughly the same content, and release it seperately. That way you don't have to ask "how much free is too much?" Longtime demo players will get a completely new experience once the game releases, while giving the community something to play and wishlist if they like it.
In point of my view, I guess it's up to market. Mobile games, PC games, Free-To-Play, Premium and lots of other parameters.
based on my experiences, if you have F2P game, you must launch a mvp version and track your user behavior and polish your games, mechanics and systems based on your data. In Premium games, its a bit different, though. You should create a demo and launch it free to see how much user like your game. it like be a proof of concept for game creator and after that if green light shows up, you should focus on full version. one of my example is SuperHot game. at first it was a web-based minimum game that launch at a game jam as Demo. another example is Celeste. it was a mechanics that developed in another game jam.
But after all these, in game industry there are lots of parameter to success a game in market. not just gameplay. story, another rivals, marketing budge, game story and team (sometimes the most).
Context matters.
How popular is the developer. What does the game's audience expect. Where will the game be released. And so on.
On Steam you can get away without a demo version, because Steam has no questions asked refunds and a entry bar ensuring a little bit of quality standard.
On Itch the preferred way of release is pay what you want. So a demo version is not even a thing. Also, those projects get published while in development. If you appear out of nowhere and release a game, you need a lot of marketing. It takes time to build an audience, if you are on a budget.
