Most SharePoint complaints — “nobody can find anything,” “there are four copies of every document,” “we just search and pray” — get blamed on the software and are actually about structure: the platform faithfully preserved whatever organization (or lack of it) was poured in, usually a decade-old file server’s folder maze migrated wholesale in a hurry. Protecting SharePoint covers permissions and recovery; this is the other discipline — information architecture: designing the structure so the everyday question “where does this live?” has one obvious answer. It’s less technical than it sounds, and it’s the difference between a shared brain and a landfill with a search bar.

The design principle: sites are rooms, not folders

The mental model that fixes most SharePoint structure: a site is a room with a purpose, a membership, and an owner — and content gets a room per how it’s used and who uses it, not one giant room with a deep folder maze. Practical translation: a site per department for its working documents · a site (or Team) per significant project or client engagement · dedicated sites for the special-handling content — the Restricted-tier material whose room-level membership is its security, and the company-wide reference (policies, templates) in a read-mostly room everyone can find. The payoff of room-thinking compounds: “who can see this?” becomes answerable at the room level, sharing rules apply per room, and navigation becomes a directory of purposes instead of an archaeology dig.

Flat beats deep: the folder-habit intervention

The file-server reflex is nesting — folders in folders, eight levels down, paths that eventually break sync — and SharePoint rewards the opposite: more rooms, shallower structure, with metadata and search doing the work folders used to fake. The working rules: if a folder tree needs a map, it should probably be two libraries or a new site · names carry meaning (the same convention discipline as Teams — CLIENT - Project, dates ISO-style, no “FINAL_v3_REAL,” because version history is the version system) · and each room’s structure gets a one-line convention its owner enforces (“client sites contain: Contracts / Deliverables / Working”), so the fourth project site looks like the first three. Consistency is the entire findability trick — people locate things by pattern, and patterns require repetition.

Every room gets an owner, every structure gets a lifecycle

Architecture decays without stewardship, so two standing assignments: a named owner per site — two for anything that matters — accountable for membership, the naming convention holding, and the eventual archive decision; ownerless rooms are where sprawl and stale access breed · and the lifecycle from the governance playbook: creation through the light-touch request path (which is where the “does a room for this already exist?” check happens — the cheapest anti-duplication control there is), and expiration-or-archive for the dormant, so the directory describes the living organization rather than its history.

Migrating a mess (the realistic path)

Businesses rarely design SharePoint fresh; they inherit the maze. The sequence that works: design the target rooms first — an afternoon with department leads drawing the site map · move actively-used content into the new structure, team by team, with the communication that every migration deserves — where things went, said twice · archive the old maze read-only rather than sorting a decade of history (the 80/20 of migration sanity: current work gets structure; the archive gets retention decisions and search) · and hold the line at the door — the new structure survives only if the creation path and naming conventions apply from day one, because entropy reapplies for its old job immediately. Expect the honest timeline of a quarter of gradual moves, and expect the payoff to be audible: the “where does this live?” questions simply stop.

The bottom line

SharePoint findability is designed, not searched into existence: rooms with purposes and owners instead of one deep maze, shallow consistent structures with conventions that repeat, a creation path that prevents duplicates, and a lifecycle that retires the dormant. It’s a whiteboard afternoon and a quarter of tidy migration — and it converts the platform’s reputation from “where documents go to hide” into what it was always capable of being: the place where everyone knows where things live, because there’s one obvious answer.