Session Folders: The Workspace Organisation Model That Keeps 50-Process Portfolios Usable

Folders are the small comfort a consultant juggling several clients comes to appreciate, and the organising layer a small-and-mid-sized business uses when it pilots a transformation in multiple waves (Q1 candidates, Q2 candidates, already shipped). Not glamorous, but it is the thing that keeps a 30-process workspace usable instead of turning it into a scroll-and-search chore.

For the first five to ten processes in a workspace, a flat dashboard with most-recent sorting is enough. You know each process by name, you remember what each one is for, and finding the right one takes seconds. Past fifteen processes, the flat view starts to fail: you scroll, you search by keyword, you end up re-mapping a process you already mapped because you could not find the original. Past thirty processes, the flat workspace is genuinely unusable and the team starts complaining that the platform is 'hard to navigate'.

The feature that fixes this is session folders: a hierarchical organisation model with specific design choices that avoid the two most common failure modes of enterprise folder systems. The first failure mode is forcing strict one-folder-per-item, which creates agonising categorisation decisions and breaks down when a process genuinely belongs to multiple categories. The second is unlimited-depth hierarchy, which produces folder trees so deep that navigating to a specific item takes as many clicks as a flat search would. LucidFlow's model addresses both.

The folder hierarchy is capped at three levels. You can have a top-level folder, a subfolder inside it, and a sub-subfolder inside that, and then no deeper. The cap is enforced at the API level, not just suggested in the UI, so clever users who try to nest deeper find they cannot. This is a deliberate design choice, not a technical limitation.

The reason for the cap: enterprise folder systems that allow unlimited depth consistently end up with trees that are six or seven levels deep because the people creating folders are not the people navigating them. A seven-level folder tree is actively worse than a flat workspace because the user has to remember the exact navigation path, and the cognitive cost of navigating outweighs the organisation benefit. Three levels is the empirical sweet spot: enough to separate 'Business Unit', 'Function', and 'Process Family' for a typical mid-market company, not so much that the tree becomes a maze.

The second design decision is that a single process can belong to multiple folders simultaneously. This is implemented via a junction table rather than a foreign-key relationship from the session to a single folder, which is the standard database shape for tag-like systems. The practical consequence: a process like 'New Client Onboarding' can live in 'Sales Processes', 'Customer-Facing Processes', and 'Active Transformation Targets' all at once, and removing it from one folder does not remove it from the others.

This choice is what prevents the classic enterprise folder failure mode: the team spends two weeks debating whether the onboarding process 'belongs to' Sales or to Operations, when the honest answer is that it belongs to both and the attempt to pick one is artificial. With tag-like membership, the team can file the process in both folders and move on. The junction model also makes folder-based reporting more accurate: 'all processes in Sales' includes onboarding even though it also belongs to Operations, which is usually what the reporter wanted.

Two smaller features round out the folder model. Each session tracks a last-accessed timestamp that is updated whenever the process is opened or edited. The dashboard's default sort is by most-recently-accessed, which surfaces the process you worked on yesterday at the top without requiring you to remember which folder it is in. For teams with multiple active processes, this sort is often more useful than the folder view, you navigate by 'what did I touch recently' rather than by 'where did I file it'.

Each folder also carries a colour (default indigo, customisable to any hex value via the folder settings). The colour shows up in the folder navigation bar and on the process cards, which is a small-but-real usability gain when the workspace has 10+ folders. Humans are faster at scanning colours than at scanning text, and the colour coding means 'where is the Finance folder' becomes an immediate visual answer rather than a scan of text labels. The default palette offers enough distinct colours for 15 folders before you run out; beyond that, custom hex codes are supported.

Folder hygiene is an enabler of the transformation workflow, not a feature in its own right. For a consultant running several client engagements in parallel or an SMB sequencing its transformation in waves, keeping processes filed cleanly is what lets Portfolio Analytics and the AI transformation plan stay legible when the portfolio grows. Boring, but the plan is easier to defend when the supporting artefacts are where you expected to find them.