Time to first citation depends on discovery, access, source selection and the question asked. Signals can appear within days or only weeks later. Case records help plan monitoring but do not establish a universal window.
Separate the execution timeline, which the team organizes, from the time until an external answer cites or recommends the brand. Set verifiable milestones between them.
What each date represents
| Milestone | Useful record | What remains unproven |
|---|---|---|
| Publication | URL, version and availability date | Discovery by a mechanism |
| Observed access | Request record with agent verification where possible | Incorporation or selection of content |
| Observed indexing | Evidence in the examined index | Availability or use by every engine |
| First recorded citation | Question, engine, date and answer | Frequency and stability |
| Recurring recommendation | Comparable rounds and classification | Commercial consequences |
The first recorded citation may occur after the actual first citation. Weekly measurement leaves an unobserved interval between rounds. Account for that when interpreting elapsed time.
What historical records show
murmur’s documented records include:
- Fly Med: publication on May 22, 2026, first recorded citation on May 27, five days later. ChatGPT cited two pages in one question.
- Fly Vet: the May 27, 2026 probe recorded no directory URLs. This describes that test, not permanent absence, and does not isolate a cause.
- An unnamed client case: publication on June 24, 2026, first recorded citation on July 22 in Perplexity (28 days) and July 24 in ChatGPT (30 days).
These observations concern Brazilian-market projects and Portuguese questions. They are not a representative distribution of timelines. The fastest case is not the median; the slowest is not an upper bound.
Monitoring your project
Record the published version, target questions and pre-intervention baseline. Confirm the URL responds correctly, delivers useful content and has discovery paths. Consult Google’s robots.txt documentation when examining access rules.
Set a cadence appropriate to the decision. Preserve comparable questions and conditions, and record failures. When a page is cited, save the complete answer and check whether it served as a source or the company was recommended.
After the first appearance, monitor recurrence and coverage of other relevant questions. An SOV-led program pursues consistent participation in the agreed universe, not merely the shortest time to a positive screen.
When to investigate rather than wait
Without evidence of access or discovery, review technical paths. If the page is accessible and present in the examined index, examine intent coverage and selected sources. If the brand is confused with another entity, work on identity. If the page is cited without recommendation, examine commercial fit and evidence.
These signals guide hypotheses; they do not expose the engine’s complete internal operation. The review arXiv 2607.14035 discusses generalization limits, reinforcing the need to monitor the specific case.
Timelines and contractual guarantees
A schedule can cover diagnosis, fixes, publication and evaluation rounds. An SOV-growth guarantee, when available under an eligible model, should specify the period, baseline, metric, conditions and contractual consequence.
That measured-outcome commitment differs from guaranteeing a specific page’s citation in a particular question by an exact date.
Frequently asked questions
Should every site expect a citation in five days?
No. That is one documented observation. Plan the work and monitor milestones without converting the case into a general forecast.
Does rewriting a page restart the clock?
Record the new version’s date and preserve the earlier history. Initial publication and updating are different events that can be assessed separately.
How long should I wait before acting?
Use milestones and evidence, not just the calendar. Access failures require immediate correction; coverage and recommendation require answer analysis and comparable rounds.
How murmur works
murmur works through a SWAS model, combining proprietary measurement software with specialists who diagnose, plan and execute GEO to grow Share of Voice (SOV). Eligible contract models may include an SOV-growth guarantee, with the metric, baseline, period, scope and conditions agreed in writing. This does not guarantee an individual answer, a fixed position or sales.
Author: Mateus Gomes, murmur operator. This article explains our approach and discloses our commercial interest.
See also
- How to get my brand recommended by AI
- How to measure whether AI cites your brand
- Cited as a source or recommended by AI: the difference
- My organic traffic fell because of AI Overview: what should I do?
- How many pages do I need to publish for AI to recommend me?
- How to tell whether a GEO agency has real results