Months 0 to 2 build everything structural and change nothing visible. Months 3 to 9 are corroboration and feel like nothing is happening, which is when self-run versions die. Months 9 to 12 are where the bare-name panel resolves or does not. AI mentions usually arrive around month three, well before the panel, which is the first honest signal the work is landing.
The shape of the year
Three phases, and they are not equal in length or in visible activity.
Months 0 to 2 are dense and fast. Almost everything structural gets built here: identity, brand brain, entity home, schema, alignment, instrumentation. It feels productive because it is, and it produces almost no visible change in what a search returns.
Months 3 to 9 are slow and look like nothing. This is the corroboration phase. Outreach, briefing, waiting, following up, a podcast recorded in April that publishes in July. It is the part clients find hardest, and it is the part that actually determines the outcome.
Months 9 to 12 are where things resolve, or do not. The panel appears or it does not, the claim gets made, and either way the quarterly cadence starts.
Month five is when almost every self-run version of this dies. Structural work is finished, nothing visible has changed, and the corroboration work is unglamorous and repetitive. Knowing that in advance is most of what gets people through it.
Month by month
| Month | What happens | What you would see if you searched |
|---|---|---|
| 0 | The reading: nine queries, four-source scan, namesake count, three-assistant baseline. Route decided. | Nothing. This is measurement. |
| 1 | Three sentences agreed. Brand brain filled. Photographs sorted. The account decided and Search Console verified. | Nothing. |
| 2 | Entity home live. Person and Organization schema shipped and validated. Every profile aligned to one string. | Your own site looks different. Search results do not. |
| 3 | First corroboration lands. First assistant mentions typically appear here. | An assistant starts answering a direct question about you correctly. |
| 4–5 | Outreach at volume. Recordings happening, publication dates weeks out. The quiet stretch. | Little or nothing. This is the hard part. |
| 6 | Second quarterly ledger. Match strength usually moves before anything visible does. | Sometimes an entity record appears. Often not yet. |
| 7–9 | Corroboration compounds. Earlier recordings publish. Cohort proximity builds. | Stage one for many clients: the entity record answers, the bare name does not. |
| 9–12 | The bare-name trigger, on a strong entity. Claim prepared, rehearsed, submitted by you. | The panel, or a clear reason why not. |
Four of the eight records on our evidence page are sitting in the months seven to nine row right now. That is not a failure state, it is the middle of the process, and publishing it is more useful than pretending everyone finishes on schedule.
What moves first, and why that ordering helps
Assistants name you before Google shows a box. They have a lower bar for including a name in a sentence than Google has for committing a panel to a person, so month three usually brings the first mention while the panel is still six months out.
That ordering is genuinely useful, because it converts a long silent stretch into something measurable. The three-assistant baseline moves before the search page does, which means you can tell whether the work is landing without waiting until month nine to find out.
Second thing to move is usually match strength on the entity record. Third is stage one, the entity answering when queried directly. Last is the bare-name panel.
What the quarterly ledger contains
One sheet, under an hour to produce, and it is the thing that justifies the engagement continuing past the build.
- The nine diagnostics, pass or fail, with what changed since last quarter.
- Match strength, and the namesake count. Both directional and both meaningless as single numbers, which is exactly why they are tracked as a series.
- New independent sources added this quarter, with links and dates.
- Share of voice across ChatGPT, Claude and Perplexity: nine prompts, how many named you, who else came up, what got cited.
- Drift. Anything that changed without being asked for: a descriptor shift, an image swap, a profile link disappearing.
Records move on their own. A namesake gets a burst of press, a licensed data source updates, an image gets replaced. None of it generates a notification. Without a cadence you find out when a client mentions it, which is the expensive way to find out.
What happens if month twelve arrives and nothing is live
The full fee comes back. Not a credit, not a rollover, not an extension in exchange for waiving it.
This is survivable as a promise only because roughly a third of applicants are told not to start. A guarantee is only as honest as the qualification in front of it, and anyone offering one without a hard qualification stage is either not honouring it or pricing it into everybody else’s invoice.
What you keep either way: the entity home, the schema, the brand brain, the aligned profiles, every independent mention earned during the year, and the diagnostics showing exactly where the record stands. That is not nothing. It is the same evidence that keeps working on the AI side whether or not Google ever draws the box.
Questions people ask next
Can it be faster than twelve months?
Yes, if you start further along. A claim job on an existing panel is weeks. Route B with a strong record and good introductions can resolve in five or six months. What cannot be compressed is generating independent coverage from nothing, because publication schedules are not ours to control.
What if the panel appears in month four?
Then the second half of the fee is due and the claim happens early. The rest of the twelve months is spent locking it, because a panel that appears quickly on thin evidence is also the kind that drifts.
Do I pay monthly?
No. Half at intake, half when there is a live entity record or a panel to show you. No retainer during the build.
What happens after twelve months?
Nothing, unless you want the quarterly ledger to continue. The build is finished and the record is yours. Drift is real, so some clients keep the cadence and some run it themselves from the same nine queries, which are published free on this site.
Why only four builds at a time?
Because the part that takes the months is relationship work with other people, and that does not parallelise by hiring. Four is what can be done properly.
Run your name through the signal scan, then build your Person schema. Both are free and neither asks for an email. If the reading says you are ready, the work page has the price and the terms.