I've long since lost count of the exact number, but I'm pretty sure that my current 'City of Spires' campaign has now run for almost as many sessions as the 'Team Tsathogga' campaign that preceded it. This has prompted me to think a bit about the different shapes that the two campaigns have taken. 'Team Tsathogga' was, from beginning to end, an extremely freewheeling, even anarchic campaign, with the PCs roaming randomly around the map getting involved in whatever seemed most interesting at the time. In 'City of Spires', on the other hand, the PCs took over their city twenty-odd sessions ago, and everything since then has dealt with their ongoing attempts to cement their positions as regional power players.This has been a new experience for me as a GM, as I've never had to deal with this form of domain-level play before. The PCs in my long-ago AD&D games sometimes rose to become high priests and archmagi and whatnot, but their non-adventuring duties always remained firmly in the background. The Team Tsathogga crew regularly took over entire communities by accident, but they never stuck around long enough to actually rule them: they'd just appoint some viceroys and wander off. In City of Spires, by contrast, we now sometimes have whole sessions that are basically just 'upkeep', with the PCs checking in on all their various civic projects and trying to deal with whatever barriers they may have encountered. (One recent example saw them plotting how best to scam an ancient subway AI into lending them a digging robot through the use of rigged customer satisfaction surveys.) They've built bridges. They've set up trade routes. They've negotiated diplomatic marriages. They've organised the planting of stands of date palms and the digging of irrigation canals. I keep worrying that they'll get bored by all this SimCity stuff, but they insist they're really enjoying it. Mysterious wildernesses on the edge of the map remain resolutely unexplored in favour of yet more civil engineering.
Over the course of these sessions, I've developed a set of rough-and-ready principles for running games which devote a lot of time to domain management. I don't claim that this is the best or only way to handle such situations, but this is what's worked for me, at least so far...
1: Keep the focus on problem-solving.
I think that one reason people often shy away from domain management is because they worry it will make their games dissolve into a morass of tedious accountancy and logistics. In reality, of course, such logistics are absolutely crucial to running a successful polity. But they're also crucial to running a successful military unit or long-distance wilderness expedition, and we never let that stop us when running normal D&D!
If you were running a wilderness trek, you probably wouldn't keep the focus on the exact logistics of pack animals, trail rations, and so on. Instead, you'd focus on the moments of crisis: how are the party going to get themselves and their supplies over this raging river? How will they sneak them through this hostile territory? Running a domain is just the same. Of course, you need a general sense of what kind of resources the community does and does not possess, to keep decision-making grounded in some kind of shared imagined reality. But for the most part I've found it helpful to assume that the day-to-day stuff just proceeds at its own pace in the background until it hits a specific problem, at which point the PCs have the option of stepping in. They're not project managers - or maybe they are, but that's a role they play off-screen. When they're onscreen, it's because they're acting as troubleshooters.
2: Keep the problems OSR-style
The general principles of OSR encounter design - 'an encounter that can be solved by simply crossing some resources off your character sheet is a bad encounter' - apply here, too. The problems the PCs encounter should never be the kind that can be solved by just throwing more resources at them. Instead, they need to be qualitative problems, the kind of things that act as potential bottlenecks for the whole project. 'You thought you'd need X amount of lumber, but then a fire destroys some of it, so now you need Y amount of lumber instead' may be a serious problem, but it's not an interesting problem. You can assume all these sorts of things are already 'priced in', and are being dealt with by the local management: that the mine foreman, for example, knows perfectly well how to deal with normal problems and setbacks involved in running a mine. It's only when his crews accidentally mine their way into a haunted subterranean city, or when a tribe of goblins cuts off the roads that the ore is carried down, or when flooding cuts off production just before a time-critical deadline, or whatever, that he'll come running back to the PCs to beg for help, because he knows that they're the kind of people who can be relied upon to come up with creative solutions to otherwise-intractable difficulties.
(Important addendum: this doesn't imply that every time the PCs attempt something, no matter how routine, you should throw some kind of intractable difficulty in their way. If it's the sort of thing they have the resources to straightforwardly accomplish, then just let them have it. But if it's something more ambitious, then they should have to overcome obstacles to achieve it - and the more those obstacles are qualitative rather than quantitative, the more rewarding the resulting play is likely to be.)
3: Make sure the PCs have access to lots of highly specific assets
4: What were once threats are now resources
The PCs are now the masters of their domain. The dungeons that they once fought their way through fearfully, one room at a time, have now been mapped and cleared out. And that means their resources are now available for the taking.
If your PCs are anything like mine, then by the time they acquire a domain of their own they'll already have done plenty of more traditional wilderness exploration and dungeon-crawling, encountering all sorts of weird and dangerous nonsense in ancient ruins and accursed tombs. But while, from the perspective of an adventurer, an enchanted lake of acid is a dungeoneering hazard, from the perspective of a ruler it's a resource. Just think of what you could accomplish with all that acid!
If you're like me, you'll probably feel an instinctive resistance to the idea of PCs taking things that were once expressions of the Mythic Underworld and turning them into military-industrial assets. Resist that instinct. The PCs earned their access to these things, access that they paid for in time and hit points and dead characters: it's only fair to let them enjoy the fruits of their exploits. Let them turn the Heat Metal room into a power plant. ('Hey, free energy!') Let them weaponise the monsters and traps and curses they've long since learned to evade. ('What if we kite the zombies all the way to the frontier?') Let them redirect that river of screaming ghosts out of the dungeon and into the moat around their castle. ('This'll keep the barbarians out!') In this way, all the weirdness they've encountered in their adventuring career so far becomes the kind of highly specific resources that allow them to come up with creative solutions to their civic problems, solutions that would never have occurred to anyone other than a D&D PC.
5: Simplify factions
When my PCs took over their city, the first thing they did was organise a grand council of all the other local power players to help them run the place. I tried to run a full meeting of this council, with all the dozens of NPCs involved in it, exactly once. Never again.
When PCs are on the outside of a power structure, it makes sense to play out each of their interactions with it individually. But once they are the power structure, and everyone else has to come to them, trying to play it all out would be madness. Logically, they'll need to talk to every single local stakeholder about each new development: but for sanity's sake it's much easier to simplify all these groups into a few main coalitions, and reduce what would actually be a long series of interactions into a handful of conversations with their spokesmen. All those weird and idiosyncratic bandit chiefs they had to negotiate with back when they were ruin-crawlers can now merge into the Bandit Coalition, with one representative who speaks on behalf of all of them. Otherwise you'll never get anything done.
(PROTIP: The representative for each coalition should be whomever the PCs have the most history with, even if they're not actually the coalition's most senior member. This both makes in-world sense - the person appointed to talk to them will be the person who knows them best - and makes interactions more meaningful, because of all the shared history they have behind them. I have loved seeing relationships that began with cutthroat encounters in the ruins end up taking on institutional significance. 'Hey, remember when you tried to kill us with hellfire? Good times. How's the literacy project coming along?')
6: Don't overestimate the powers of the state
Compared to modern states, most pre-modern polities are ramshackle as fuck. Remind your players early and often that just because they have 'a government' doesn't mean they have anything resembling a modern bureaucracy, with a police force and a civil service and so on. They probably have a stronghold, an army, a treasury, a bunch of advisers, spies, and informers, a network of local 'big men' who can be expected to semi-reliably enforce their edicts as long as they are kept in line with threats and bribes, and not a whole lot else. Their writ may run along the roads and the rivers and the major agrarian areas - but in between, in the woods and the swamps and the deserts and the mountains, there are going to be all kinds of places where state power barely functions, and where adventures can consequently continue to flourish. There are still lots of situations in which having an army isn't actually all that useful, and where it might consequently still make sense to get the old adventuring team back together for another journey into the unknown.
7: Let the PCs enjoy the fruits of their success
If your players have gone to the trouble of building a real powerbase, it's probably because they're interested in actually having and using power. So let them. It's totally OK if the rise of the PCs to power mean that a lot of things that were previously threats for them can now be trivially dealt with. The level of power they wield in the world has just increased by an order of magnitude: 3d6 goblins in a cave just isn't going to cut it any more.
D&D PCs tend to be powerful, highly competent, individualistic, and more than a little crazy, so if a bunch of them have just seized hold of a domain, then that domain is probably going to be in for some interesting times. Let your PCs make changes. Let them make big changes. This doesn't mean that everything they attempt should succeed, but everything they attempt should have consequences. If their domain has been changed beyond recognition within a few years of them taking over, then that's a good thing.
You know all those crazy lords and wizards in the backstories to D&D scenarios, the ones who are always building weird strongholds and meddling with arcane forces and making pacts with inhuman beings and bringing about lost golden ages and magical cataclysms and so on? Well, now your PCs have the chance to be those people. Let them make the most of it.
After all, just think of the dungeons they'll leave behind them!