Motivation was never the constraint.
Four workflows walked through every moment of their journey maps, each breakdown classified as a motivation gap, an ability gap, or a prompt fault. The result was consistent enough to be the entire design brief in one line.
The journey, audited
Block → Interpret → Act → Store, applied to twenty-four moments across four workflows.
Zero motivation gaps across all four workflows and twenty-four moments audited.
Every breakdown but one is an ability gap — something Chloe cannot do without unreasonable effort, not something she does not want to do. The single exception is a prompt fault, and it was deliberately deferred.
That deferral is worth stating plainly, because it is the kind of thing a roadmap usually gets wrong. Nothing cues Chloe to open the tool before Slack's badge does — a real gap. But the earlier motivation audit found that every prompt authored by someone else drains her. A notification from this product, competing with Slack before any trust has been earned, would be a sixth drain rather than a fix. So friction is removed first, motivation second, prompts last. Never the other way round.
Four workflows, four root causes
- W1 — Morning Orientation anchor
- Nothing aggregates, marks completeness, or persists what she already learned. Five of six moments are ability gaps; the sixth is the deferred prompt fault.
- W2 — Delivery → Approval
- One interpretation-phase failure cascades through four later moments. Work ships with zero interpretive scaffold — no stated intent, no rationale — so client feedback defaults to taste. The transaction's only record is a verbal “approved!” with no timestamp and nothing retrievable, carrying real financial and legal weight on nothing.
- W3 — Stall → Re-entry
- The deepest pit in the study, and an absence rather than a failure — there is simply no signal to interpret. Two weeks of silence could mean “busy” or “gone”, and nothing distinguishes them, forcing a hold-the-slot-or-refill decision with real cost either way she guesses wrong.
- W4 — Proof of Work
- A pure storage failure: the decision record needed for a case study was never captured — not deleted, never assembled. This workflow has almost no independent root cause. It is where W1, W2 and W3's unfixed gaps come due at once, which is why its fix is sequenced after them rather than beside them.
Nobody names “I need to write case studies” as a standalone pain. The pain they actually feel is Morning Archaeology, Revision Hell and the Ghosting Sequence — arriving late.
What the market already tried
Five categories, five different architectures, one shared failure — arrived at independently of the interviews, which makes it a cross-check rather than a coincidence.
| Category | Representative tools | Why it does not close the gap |
|---|---|---|
| Creative proofing & approval | Filestage, PageProof, Approval Studio | No rationale travels with the deliverable, so approval is captured with nothing to attach it to and no decision record ever forms. Requires manual export and upload on top. |
| Client portals & ops workspaces | HoneyBook, Bonsai, Copilot, Moxie | Replaces existing channels instead of sitting above them. Demands migration, and does not touch Morning Archaeology — it is one more tool to check, not an aggregator of the ones already open. |
| Conversational AI | Limitless, Fathom | No design-context parsing. Cannot connect activity to a project's rationale or a specific phase. A general productivity layer, not a decision record. |
| Unified personal planners | Sunsama, Akiflow | Solves the aggregation half only, and only privately. No approval gate, nothing a client can ever be shown. |
| Visual canvases | Miro, Milanote | The user becomes the aggregation mechanism by hand — the exact inverse of automatic indexing. Abandoned under time pressure, which is precisely when it is needed. |
| Vantage | — | Reads what already exists, writes only to what it owns, zero migration. |
Every one of them either asks the user to do the integration work by hand, asks them to migrate away from tools they already trust, or aggregates without producing anything a client can be shown. None combine “reads what already exists” with “produces a retrievable, client-facing record.”
That combination is the whole thesis — and it was reached twice, from two unrelated evidence sources: the interview transcripts and this scan.
Ruled out before design started
Ethical exclusions fixed at the project frame, so they constrain every later decision rather than being negotiated per feature.
- No silent read-scope expansion
- Never index beyond what the user explicitly granted.
- No fabricated proof of work
- A case study is drawn only from what was actually captured — never invented or paraphrased on the user's behalf.
- No resolving ambiguity by inference
- Activity that cannot be confidently attributed, or a client message that merely sounds like approval, is held visibly rather than guessed at.
- No engagement-maximising prompts
- Any cue competing for Chloe's attention must feel self-authored and be earned by trust — never a vendor notification arriving first.