Month one: what should you expect?
Month one with a Search Visibility FTE is about establishing the machine rather than moving numbers. It connects to the site and Search Console, reads the existing pages, proposes targets, and begins issuing worth-it verdicts on candidate topics.
Expect to reject some of those verdicts and to send back early drafts. The rejections are how the brand rules and quality standard transfer; a month with no send-backs usually means nobody is reading closely enough.
Which mechanisms need tuning first?
The two mechanisms that need tuning first are the topic exclusions and the tone. Topics the business should never publish on must be listed explicitly, because the FTE will otherwise treat a high-opportunity topic as worth writing regardless of whether the business wants to be associated with it.
Tone transfers through examples and specific send-backs. Two pieces you consider on-brand and one you do not, each with a sentence on why, does more than any adjective-laden brief.
Month two: when does the rhythm settle?
Month two is when publishing becomes a rhythm rather than an event. Send-backs on the same reasons fall as the remit tightens, approval time per draft drops, and the fact gate stops flagging things the manager has to explain.
This is usually when the first narrow autonomy promotion makes sense: refreshes of existing pages that have been approved repeatedly without amendment. New topics keep their gate.
- Month 1: connect, propose targets, tune exclusions and tone through send-backs
- Month 2: publishing rhythm, falling send-backs, first refresh promotion
- Month 3: first measured movement, refresh proposals, quarterly target review
Month three: what does measurement show?
Month three is when scheduled measurement starts to show movement, and when the first refresh proposals appear for pages that were published early and have begun to shift. Rankings on new pages typically take weeks to settle, so early months show direction rather than destination.
AI assistant citations often move differently from Google rankings, and the FTE reports both separately. A page can rank well and be uncited, or be cited while still climbing; treating them as one number hides the picture.
What targets are realistic in a first quarter?
Realistic first-quarter targets are about coverage and process rather than outcome: a defined set of buyer questions with a published, fact-checked answer for each, and a measurement baseline for rankings and citations recorded for those pages.
Outcome targets such as share of answers belong in the second quarter, once the baseline exists and the first pages have had time to settle. Setting outcome targets in month one produces a quarter of arguing about noise.
What goes wrong, and how do you catch it?
The two things that most commonly go wrong are an unread approvals queue and topic drift, and both are visible in the workspace before they show up in results. A queue that is not cleared daily means nothing publishes; a run of verdicts on topics that feel off-brand means the exclusion list needs work.
Both are management fixes rather than system fixes. The role produces what its remit permits, and the remit is yours to sharpen.
What should you refuse to do in the first 90 days?
Refuse to promote publishing on the autonomy ladder. Public content is the class most organizations keep gated indefinitely, and in the first quarter there is no record long enough to justify otherwise; the FTE's value is in the proposals, the fact gate, and the measurement, not in publishing unattended. Refuse to judge results before the second cycle, because the first month is setup, memory correction, and remit tuning, and its numbers are not representative.
Refuse to let the prompt set drift. The citation samples are only comparable if the same prompts are run the same way on the same schedule, so revise the set quarterly rather than weekly. Refuse to accept a page that fails the fact gate by adding a source after the fact; send it back and let the FTE revise. And refuse to estimate what the performance view marks as unmeasured, because a number nobody can defend is worse than a gap honestly stated. The first 90 days are for building a record; the refusals are what keep the record clean.