Marketing sounds good. However, it isn't going to happen if nobody does it. That isn't going to be me either, as I'm terrible at it and not personally interested in it as a skill. The right people talking to the right streamers could be fruitful. Eg someone led Day9 through some TA 1v1s (no matchmaking required) a year or so ago. Why not ZK?
Learn from the BAR people. They seem to be quite good at hype. One of their most active developers is a professional web designer. A big modern animated landing page backed up by fancy articles and unit pages surely helps new people take BAR seriously. Something like that could help ZK, although possibly just the landing page. Integrating it with all the dynamic stuff in the background of the ZK site seems tricky.
Whatever Sanctuary is doing is probably effective, even though it comes across as a bit annoying to me.
BAR also has people semi-officially talking to other discords, twitter, youtube stuff, people pre-rendering scenes. This is the sort of "grassroots" stuff anyone could try their hand at. But note that ZK is entirely grassroots. 64_Bit_Dragon
has recently restarted the ZK twitter and has been making videos for a while, while Kingstad
has done some communication between discord.
On graphics, there are multiple levels on which people could contribute.
The top levels involve making quality models or learning/knowing enough openGL 4 to write shaders etc. I'm not that interested in most parts of graphics programming (it strikes me as a lot of global scopes, magic numbers, and boilerplate - although I enjoy it once that it set up and I'm writing something an algorithm to run on pixels), and there is a lot of other stuff to maintain so I don't expect to dive deep any time soon.
A simpler level would be to look at what BAR does and adapt it to ZK. They have their own art, yes, but they are happy to share graphics code.
Even simpler levels involve barely more than text file editing, eg GFX can be touched up just by tweaking numbers. Again, possibly with inspiration from BAR. We are on exactly the same engine as BAR so all the same technology is there.
The simplest level is ridiculously simple. A lot of maps have pretty poor lighting an water settings. Improving the lighting of the units and terrain can have a massive impact on perceived quality. There is a widget to tweak all these things and output it to a config file, which is read locally, but can also be put in the game for everyone.
I don't think graphics (and certainly not the rest of the game) is the low hanging fruit for growth though.
I don't think ZK has reached everyone that might be interested. It may well have reached a sizeable portion of the most active indie-RTS online space, at least on Discord, but that feels like a tiny bubble. This will be a slim fraction of RTS players, and a slimmer fraction of people who play games in general. People have said ZK is their first RTS.
Here are some server-side stats. Note that in the following a player is someone who played a game on the server (such that it ended up generating a replay). This is not directly related to singleplayer.
Don't worry about retention. It always looks like that near the end of the period, with a surprisingly wide end. It drops off, essentially, because nobody who started playing 20 days ago has been playing for 30 days.
What these graphs tell me is that ZK has been fairly steady, with bits of growth, over about four years with next-to-no marketing. So surely marketing would do something.
On new players:
The server definition of "player" says we have about 30 new players a day.
The steam definition says we have 100 "licenses granted" per day, which I assume means "Add to Library".
But only 65 downloads per day. Which could include people reinstalling.
On the flipside, with ZK going steady, this could mean we are losing 65 players per day. Alternately we could be losing nobody, but each existing player could play progressively less often so as to cancel out the gain. Reality will be somewhere in between.
Steam won't give me monthly active users, but it says daily users averages about 950, peaking around 1,200 on an average weekend. The casual ladder has 1729 people on it at the moment and I recall it being set to expire after two weeks, so that should be a good estimate of the two-week playerbase of hosted competitive battlerooms. These numbers are very hard to compare. Coop seems quite popular and singleplayer even more popular than that. We had some stats a few years ago that said something around 3500 weekly active users. But the wide down ramp on the multiplayer retention graph implies that people may play a bit then return weeks later. So it is hard to guess monthly players. Somewhere between 5,000 and 10,000 doesn't seem far off. But then, assuming the 65 downloads are new, we're losing 2,000 per month. This doesn't seem implausible if we assume that the majority of the lost players are new downloads that bounce off the game (which is fine imo if RTS just isn't for them).
Retention isn't necessarily a good end goal. How long "should" people play ZK for? Steam says that the average player has played for 21 hours, out of 190,100 players. That seems pretty good. The median is 71 minutes (first standard deviation of 7 minutes to 17 hours). Sticking with a free game for 71 minutes, to give it a go, seems fine to me. Most people will just play a bit of coop with friends or even just the campaign. We could try sucking these people back in with things like daily challenges and rewards.... but why? I like doing the releases at about the rate that they happen, and events are good. More types of events are possible even (eg unit tweaks allow for "special balance" weekends). But I wouldn't want to ramp up retention via the modern habit-forming systems. I think we need to put ZK in front of more people who would get a good 1000+ competitive MP experience out of it, rather than try to draw out the time spent by the 20 hour campaign people.