GEO typically produces measurable movement on specific, high-intent queries within eight to fourteen weeks, and on broad category queries within six to twelve months with the important caveat that in categories dominated by entrenched incumbents, the broad query may not be winnable at any realistic level of investment.
That last clause is the one most proposals omit, and it is the one that determines whether an engagement is judged a success or a failure.
The rest of this piece is the phase-by-phase version, what to expect at each point, and what should worry you?
Phase 1: Weeks 1–3 Diagnosis:
What happens? you establish where you stand and why?
- A fixed prompt set fifteen to twenty buyer-intent queries run across ChatGPT, Perplexity and Google AI Overviews
- Which competitors are cited, how often, and in what position
- Which sources the engines pull from in your category, using Perplexity’s citation display
- An audit of whether your existing content is structured to be extractable at all
- Technical checks: crawler access, rendering, indexability
What changes in the numbers?: Nothing. This phase produces information, not movement.
What should you have at the end? A baseline appearance rate, a baseline share of voice, and a specific diagnosis. Not “you need more content” something like “you are absent from all comparison queries because no comparative content about you exists anywhere, and the three cited competitors each appear in four or more third-party roundups you are missing from.”
Warning sign?: If the diagnosis could apply to any company in any category, nobody ran the check.
Phase 2: Weeks 3–8 Building:
What happens? : The content and structural work.
- Comparison pages against your most-encountered competitors
- An alternatives page for the largest
- Basic facts made extractable what it does, who for, what it costs, what it integrates with
- Schema implementation
- Answer-first restructuring on key existing pages
What changes in the numbers ?: Still nothing, or close to it. Pages need to be crawled, indexed and incorporated. Retrieval systems do not respond immediately.
This is where engagements fail: Two months in, real money spent, real work done, and the metric has not moved. The pressure to change direction peaks here.
What to hold onto?: The work in this phase is what produces movement in phase three. Abandoning it at week seven means paying for the cost and none of the return. If your agency has not warned you about this phase in advance, that is a problem with the agency, not with the channel.
Warning sign? : If you were told to expect results by week six, the timeline was sold rather than estimated.
Phase 3: Weeks 8–14 First movement:
What happens ? : Narrow, high-intent queries begin returning your name.
Specifically, and in this order:
1. Alternatives queries “[competitor] alternatives” is usually first
2. Direct brand queries “is [your product] good for X”
3. Specific use-case queries your product for a defined situation
4. Comparison queries you versus a named competitor
What changes in the numbers ? : Appearance rate typically moves first, share of voice follows. Expect movement on three to six prompts out of twenty, not across the board.
What does not move: The generic category query. “Best CRM software” will look exactly as it did in week one. This is normal and should not be read as failure.
Warning sign: If nothing at all has moved by week fourteen, something is wrong. Either the diagnosis was incorrect, the work was not done properly, or you are in a category where the mechanics differ. This is the point at which it is reasonable to ask hard questions.
Phase 4: Months 4–6 Broader movement:
What happens ? : Off-site work begins to land.
Review velocity accumulates. Third-party roundup inclusions publish. Community presence builds. This is the slowest component, and it has the longest lead time outreach conducted in month two produces coverage in month five.
What changes in the numbers ? : Share of voice becomes the more useful metric here, because appearance rate can rise while competitors rise faster. Broader comparison queries begin including you.
What to expect realistically? : If you started at 4% share of voice, 8–12% by month six is a good outcome. Doubling is meaningful. Tripling would be unusual.
Warning sign: An agency that has done no off-site work by month four has skipped the component that matters most and is hardest.
Phase 5: Months 6–12+ Category queries, sometimes
What happens ? : The generic category query the one everybody wanted from the start becomes contestable.
Or it does not. This is genuinely category-dependent.
Where it is winnable ? : Fragmented categories with no dominant incumbent, newer categories where consensus has not settled, categories where the leaders have weak comparative content.
Where it is not ? : Categories with one or two brands that have a decade of accumulated mention volume across every source these systems read. In some categories, a company will not appear in the generic answer regardless of what it spends, because the model has learned a strong association that a year of good work does not displace.
The honest position? : This should be assessed in phase one and stated then. A good partner tells you in month one that the head query is not realistic and focuses your investment on the queries that are. A poor one lets you discover it in month nine.
What affects the timeline?
Six variables, roughly in order of impact:
Category competitiveness: The single biggest factor. A fragmented category moves in months; a two-horse category may never move on head terms.
Starting position: A company with existing content, some review presence and reasonable technical health starts several months ahead of one starting from nothing.
Whether off-site work is happening: On-site work alone plateaus. Companies doing only content work see phase three movement and then flatten.
Content type, not volume: Publishing more explainers extends the timeline by consuming resource that should go to comparison content.
Engine changes: These systems update without notice. A model release can move everyone’s numbers in either direction for reasons unrelated to your work.
Whether the product is actually differentiated: If there is no segment where you are genuinely the better choice, there is nothing for a comparison to say. That is a positioning problem, not a visibility one, and no amount of GEO work solves it
What to expect month by month ?
Month Realistic expectation:
- Baseline established. Diagnosis complete. No movement.
- Content built. Technical fixed. Still no movement.
- First appearances on alternatives and direct queries.
- Movement on 3–6 of 20 prompts. Broad queries unchanged.
- Off-site work starts landing. Share of voice begins to move.
- Consistent presence on specific queries. Occasional broad appearance.
- Comparison queries broadly. Head term contestable in some categories.
- Head term, in fragmented categories. Compounding effects visible.
Why the timeline is what it is:
Three structural reasons, worth understanding rather than accepting on faith:
- Retrieval systems need your content to exist, be crawled, and be incorporated. That alone is weeks.
- Consensus takes time to build. These systems favour sources that multiple independent parties corroborate. You cannot manufacture independent corroboration quickly, and attempts to do so are detectable.
- Trained knowledge only updates on model releases. Half the mechanism responds to live retrieval and moves in weeks. The other half updates when a new model version ships, which is outside anyone’s control.
Questions to Ask Before Signing Anything:
Five questions that separate an estimated timeline from a sold one:
1. Is the generic category query realistic for us, and how did you determine that? A specific answer referencing your category is good. “Yes, with the right strategy” is not.
2. What will have changed at week eight? If the answer is a metric rather than “work completed, no visible movement yet,” the timeline is optimistic.
3. What off-site work happens, and when does it start? If it is not in the first two months, it will not land in six.
4. What would tell you this is not working? A partner who cannot name their own failure condition has not thought about it.
5. What have you seen take longer than expected, and why? The answer tells you whether they have actually done this.
Frequently Asked Question:
Can GEO produce results in 30 days?
Not meaningfully. Technical fixes can be made in days, but retrieval systems need content to be crawled and incorporated. Eight weeks is the earliest realistic point for any movement.
Why do specific queries move before broad ones?
Less competition. Broad category queries have entrenched incumbents with a decade of accumulated mention volume. Specific situational queries frequently have no dedicated content from anyone.
What if nothing has moved after six months?
Either the diagnosis was wrong, the off-site component was skipped, or you are in a category where the head terms are not winnable and effort was misdirected. All three are worth raising directly.
Does GEO work faster for smaller companies?
Sometimes. Smaller companies are more likely to be in fragmented categories or targeting narrow queries where the competition is thin. Size itself is not the variable category structure is.
How long do results last?
Comparison and alternatives pages continue working for years with quarterly updates. Off-site presence is durable. This is a compounding investment, which is why the slow start is worth tolerating.
Should I keep paying through the flat months?
Only if the work is being done and you can see it. Ask for evidence of output pages built, outreach sent, reviews requested — rather than evidence of results, during phases one and two.