Days 43–63: operator workflow
You've done the foundational work for six weeks. The platform now has enough context (proposals, citations, press and audit results) to prepare more useful recommendations. Your job is to review, approve, and execute with control.
Day 43: switch to operator-first mode
Until now you've been the picker: choosing threads, drafting pitches, running scans. From Day 43, Navigator is the picker. Your job becomes:
- Approve good proposals
- Reject wrong ones
- Execute on the work the agent stages
Open /dashboard/proposals first thing each morning. Most days the queue is two or three proposals across the brand's ecosystem (and across other clients if you're an agency).
Days 44–49: the daily rhythm
Each weekday:
- Morning (5 min): Work the proposals queue. Approve, reject, queue work.
- Midday (10 min): Execute one approved tactic (one citation batch, one press follow-up, one directory fix). One tactic per day, done well, beats five half-finished.
- End of day (2 min): Review customer-visible Activity and Logs for incomplete work. Contact support about persistent errors.
That is the core rhythm. Time per client will vary with the recommendation and the amount of work your team approves.
Day 50: add two more monitor prompts
Re-run keyword research with a different seed (a competitor's name, a specific buying scenario, a vertical-adjacent term). Promote one or two new clusters to monitor prompts.
The point isn't to add scope: it's to keep the SoM measurement aligned with how real buyers are actually asking AI models.
Days 51–56: record what went live
Open the Work Queue and go through the work you approved. For each item, record what actually went live and add the evidence (a URL, a screenshot, a confirmation). Approval and dispatch are not proof that work was implemented, and nothing here measures whether a tactic caused a change.
Any tactic history available to your role is observational background. It has no comparison group, and new outcomes are not being measured. Restricted platform administration is not a customer workflow.
Days 57–63: Press & authority push
Navigator typically doesn't auto-execute press distribution: that work lands with you. So this final stretch is when you push the press cadence:
- 2–3 thoughtful pitches per week
- One thought-leadership angle (founder-led commentary, data study, contrarian take)
- Aggressive follow-up with anyone showing interest
- Coverage logged manually in Press so the audit catches it
Press is usually one of the slower pillars to move. Track completed outreach and earned coverage, then use the next audit to measure whether the signal changed.
Mid-window check-in
Around Day 56, run a manual audit. Compare:
- Composite score against the baseline and its confidence range
- Which off-page signals changed after approved work
- AI Presence by prompt and AI experience
- Press movement only where coverage actually landed
- Any entity inconsistencies that remain unresolved