A multi-unit franchise owner is running the same business five times in different rooms with different people. The unit economics are nominally identical and operationally never are. Location 2's manager is excellent at training but light on inventory discipline. Location 4 has the best margins and the worst foot traffic. Location 5 is six weeks from a brand audit and you can't remember whether the corrective actions from the last one ever got closed.
The cost of running multiple locations isn't visiting each one. It's holding each location's specific operating context cleanly enough that the next conversation with the manager — or the brand auditor, or the lender — is grounded in the actual record instead of "I think we fixed that?"
A working notes setup doesn't replace the time on the floor. It holds the operating record per location and lets the agent read across the portfolio so the patterns become visible. The same per-unit shape carries over to agency owners scaling without hiring and fractional executives managing multiple engagements — different "unit," same vault structure.
A vault per location, plus a portfolio page
The shape that holds up is one top-level page per location — Location 1, Pearland, Location 2, Sugar Land, Location 3, Katy — with sub-pages for staffing, inventory, vendor relationships, brand-audit findings, manager check-ins, and the operating numbers. A separate top-level portfolio page sits above them for cross-location patterns and decisions.
Capy supports unlimited page nesting, so a location with a complex remodel project or a regulatory issue can fan out into per-project pages without forcing you to flatten anything. Plain markdown matters because the agent can read across every location in a single query. When the area developer asks "which of your locations are out of compliance with the Q2 packaging update," you ask Capy to read the brand-audit pages and tell you. The answer takes the time it takes to read it.
Manager check-ins that capture the texture
The weekly call with each location manager is the highest-information surface in the operator's week and routinely the worst-recorded. Manager 2 mentions a turnover problem. Manager 4 talks through a vendor pricing change. Manager 5 raises a customer complaint pattern. By the time you finish the fifth call of the day, the texture of who said what has dissolved.
Record each call in Capy. The transcript comes back with speaker diarization — labels like Speaker 1: … so you can tell what the manager raised versus what you committed to. Park the recording on the right manager-check-in sub-page. Ask Capy to pull out:
- Action items, with owner (you or the manager)
- Open issues raised on the call
- Anything that contradicts what was raised in a prior week
- Items worth comparing across locations
The fourth bullet is where the agent earns its keep. If three of five managers raised inventory concerns this week, that's a portfolio-level signal — not a per-location problem. The cross-location pattern only becomes visible when somebody can read the five calls together.
Brand-audit prep that actually closes the loop
Brand audits run on a 6–12 month cadence and the corrective actions decay between them. The auditor surfaces twelve findings, the manager confirms they'll address them, six months later you're in another audit and somehow the same three findings have shown up again because the close-out never got recorded.
A working setup is a brand-audit findings inline database per location, with rows for date, finding, severity, owner, status, and the close-out evidence. The database lives directly inside the markdown page via the :::database::: directive — alongside the prose context, not in a separate tracker. After each audit, drop the audit report PDF onto the page; Capy auto-converts it to markdown via docstrange so the findings are searchable text, and you ask the agent to propose database entries for each finding. You confirm and edit before they land. The same audit-trail loop also runs deeper in AI notes for compliance and audit preparation.
When the next audit shows up, the prep is two minutes: ask Capy to give you the open findings from the last audit by location, with current status. Findings that closed properly stay closed. Recurring findings get the operator's attention before the auditor does.
Vendor tracking that survives location handoffs
Each location has its own vendor stack — produce, paper goods, equipment service, pest control, payment terminal. The relationships are mostly identical and the specifics are mostly different (different reps, different contract dates, different pricing). When a manager rolls off, the vendor knowledge usually rolls with them.
A vendors inline database per location captures the basics: vendor, contact, contract end date, pricing terms, last issue and how it was resolved. Drop vendor contracts and pricing emails on the page; PDFs auto-convert via docstrange. When the manager hands the location to a new manager, the new person inherits the vendor record instead of having to rebuild it call by call.
Across the portfolio, ask Capy to find vendors used at multiple locations and surface where the pricing or service terms diverge — sometimes the divergence is justified by local market conditions, and sometimes it's a renegotiation opportunity hiding in plain sight.
Operating numbers in markdown alongside the context
The standard franchise operator tool is a spreadsheet of weekly numbers per location. It works as a numerical record and fails as a context record — the spreadsheet doesn't capture why labor costs jumped at Location 3 in week 14 (a manager turnover, a one-time training event, a holiday shift pattern).
Park the weekly numbers in a weekly operating page per location with the spreadsheet exported as a markdown table — and write the context underneath each notable number. When labor jumped at Location 3 in week 14, write the two-sentence explanation right there. Six months later, when the area developer asks about the labor pattern, the answer isn't "let me check" — it's a paragraph already on the page.
Ask Capy to read across the weekly-operating pages and surface the patterns that don't fit your operating model: locations consistently above the labor target, weeks where revenue dropped without a clear explanation, the seasonal patterns that recur every year. The agent does the cross-location reading you don't have time for; the patterns become a discussion item at the next operator meeting.
A portfolio decision log that doesn't decay
The decisions that affect more than one location — pricing changes, menu rollouts, bonus-program changes, hiring profile updates — are the ones that should be best-recorded and usually aren't. They get made in a sprint, get communicated to the managers, and the rationale gets lost.
A portfolio decisions inline database with rows for date, decision, scope (which locations), rationale, and the metrics we expected to move. After each portfolio-level decision, ask Capy to draft the entry; you edit and confirm. Six months later, when somebody asks "remind me why we moved every location to the same opening time last fall," the answer isn't a guess.
This becomes especially valuable when you're considering whether to adjust a portfolio-level decision. The original rationale is on the page; the actual outcomes are in the operating numbers. The before/after comparison is a one-paragraph synthesis, not a half-day research project.
Onboarding the next location without restarting
The expensive part of opening or acquiring a new location is the onboarding — every operating practice, every vendor relationship, every brand standard has to be transferred. Most operators do this informally, which means the new location takes 6–12 months to operate at parity with the existing portfolio.
A playbook page in the portfolio vault holds the operating practices that should be true at every location: open and close procedures, weekly cadence with the manager, monthly reporting expectations, audit prep checklist, vendor stack template. When a new location opens, the playbook is the starting point. The new manager reads the playbook, you walk them through it, and the location starts in the same shape as the rest. (The same playbook-as-living-doc pattern is the spine of AI notes for standard operating procedures.)
Across the year, when an existing location surfaces a practice that's working better than the playbook version, ask Capy to draft a playbook update. The playbook stays alive instead of decaying into a once-written training doc that nobody opens after week one.
What this isn't
Capy isn't a POS, isn't accounting software, isn't a labor scheduling tool, and isn't a brand-compliance platform. The structured-data side of the franchise still lives in the tools that handle structured data. The vault holds the unstructured side — the calls, the audit-finding context, the vendor relationships, the rationale behind portfolio decisions, the patterns across locations — which is the part currently spread across email, Slack, and a notebook.
It's also single-user by design. The franchisee owns the portfolio vault. Outputs ship to managers, the area developer, the lender, and the brand as briefs and reports through the channels you already use.
A small first test
The cheapest way to see whether this fits your portfolio is to take the last month of one location's manager check-ins and rebuild them in a Capy page. Drop the recordings, let Capy transcribe with speaker labels, and ask the agent to summarize the recurring issues across the four calls. If the synthesis surfaces a pattern you'd missed in the live calls — or surfaces it sooner than the next operator meeting would — you've got a sense of what running the portfolio in this shape would do.
Try Docapybara free. Load one location's month of check-ins and see what the agent does with the operating record.