I have always wanted to run urban adventures, but always had trouble doing so until I recently developed a system for organizing my preparatory work.
The first step was to brainstorm factions. I that each faction (or its prominent members) needed both public and secret goals, and that the goals of different factions should sometimes cause strife and other times encourage alliances. Because of the 9P setting I made sure the factions represented a variety of races and Powers.
After much experimentation I realized each faction only needed four descriptors: key person, summary overview, typical encounter, and connections.
As an example, within Industry District the Builders Guild is led by a Dweorg names Lachary the Solid. The guild espouses that Frosty Kostkey has indeed mended his ways and promotes selling utilitarian machinery to improve the district's reputation. Typical encounters could be testing or protecting new machinery for the city, or dealing with an over-zealous machinist who is creating dangerous machinery. The guild opposes people who harm the district's reputation by remaining loyal to Frosty Kostkey's historic plans of conquest and Winter bleakness.
The second step was to brainstorm a half-dozen wrinkles. This is my terminology for a new issue affecting Arlinac Town. The issue provokes responses by most factions. These responses need the same four descriptors.
As an example, poor people have been disappearing from the slums. A kobalt named Railey the Sweet wants to help, and bring public gratitude to the Builders Guild and the Industry District. She is a local expert at constructing sentry cameras, and has teamed up with a Bergtroll named Viczoria who crafts magical lenses to construct several cameras that will make time-stamps when they see Ogres or Undead. Railey needs clueless adventurers to discreetly hide a camera on a rooftop, stay somewhat nearby for several hours to guard it, and then bring it back to her so she can analyze the time-stamps it produced. But what else might the magical lenses be able to see or record? Unrelated to the missing poor people, the cameras might notice that a vampire pirate has started shipping ghouls into town. Also, Viczoria is a prominent member of a troublesome faction based in the Illuminated District that believes beautifully made things only count as "artistic" if they are magical and popular: her politics might cause trouble for adventurers working with her.
A wrinkle could be a policy change, holiday, notable visitor, or crisis that affects the city broadly. Because not all wrinkles are tied to a specific location in the city, the useful physical "map" if the city is the vague layout of wrinkles. But a worthy wrinkle does (eventually, if not initially) involve something tangible: a person, item, or location that is causing all the fuss. So I visually imagine this tangible layer as pins on a wall map. Each pin denotes the person, item, or location creating a center of gravity that pulls factions to action. Yet pins might need to be relocated on the wall map as events unfold.
In the above example, the PCs eventually discover a secret sect of Ogre Toxophilites has been having new recruits hunting poor people as part of their initiation tests. The wrinkle's "pin" is an Ogre named Phloe who is the current recruiter for that sect. The GM has no need to determine her initial location in Arlinac Town. Let the PCs can investigate however they want. Various wrinkles will all be unfolding simultaneously. When the PCs make appropriate decisions, the GM can improvise clues about Ogres and Phloe. The eventual encounter with the Ogre Toxophilites will be most meaningful if it develops organically.
The "true map" of the city is abstract and three-dimensional. Each faction has its own layer: a network of nodes floating above the more tangible pins. In the above examples I described the wrinkle "poor people have been disappearing" with the pin "Phloe the recruiter for the Ogre Toxophilites" and one of many nodes above it in the Builders Guild layer.
Not every faction will have a node above every layer. That is okay. The followers of Frosty Kostkey who still want to conquer in his name might not care about missing poor people or a small group of murderous Ogres. Plenty of connections are made as the different wrinkles interact, and the factions clash in the ways they are interested in the wrinkles.
I now have a general structure for urban adventures: wrinkles, with tangible pins, under a network of faction-layer nodes.
I owe a debt to Justin Alexander's writing about structural issues in fantasy role-playing games. He wrote about how a dungeon crawl has a standard location for player choices (a room) and default player choices (explore the room, move to the next room). Most fantasy role-playing games thus have game mechanics specifically designed to deal with these. There are rules the GM uses to randomly put enemies or puzzles in a room to make an encounter. There are rules the PCs use about combat and skills to fight the enemies or solve the puzzles. And there are rules about locked doors, secret doors, portcullises, and other complications to add variety about moving to the next room.
These structures and game mechanics do not apply in an urban adventure. Justin Alexander proposed his alternative. From that I took the insight of a three-dimensional network of faction nodes above a more tangible layer. But his alternative remained location-based without my distinction between wrinkles, pins, and locations.
My urban adventure structure has standard issues for player choices (the wrinkles) and default player choices (investigate a wrinkle, move to the next wrinkle). Stories now have a natural flow towards depth of problem-solving. Investigation and exploration lead from wrinkles to pins and nodes. More investigation leads from pins and nodes to locations. Entering these locations leads to encounters.
Unlike the classic dungeon crawl, there are no game mechanics to push the story along. There are no random tables for creating wrinkles or pins. There are no rules for moving to investigate the next wrinkle. There are no rules dictating when to finally encounter the wrinkle's tangible pin. It is even true that because players use creative methods of investigation it will be player choice more than game mechanics that determines the most obvious or appropriate skill used when investigating or exploring.
Now I have words for why I always wanted to run urban adventures. I like when the story itself, rather than game mechanics, pushes the story along. It feels right when the story itself takes over and handles its own flow.