How long local SEO takes before anything visibly changes

The reason this question matters is not curiosity. It is that owners cancel good work at week six and double down on bad work at week twelve, because they had no framework for what should have happened by then.

I have sat on the other side of this in restaurants. When you cannot tell whether a change is working or just slow, you start changing things at random, and then you can never attribute anything to anything.

Which changes show up fastest?

Profile field changes, and they are the reason to do them first.

Categories, hours, services, the description, and photos are structured data Google already has a record for. Editing them is a direct update rather than something that has to be discovered. Google reviews changes before they go live on a profile, and it notes that approved service area edits can take around 48 hours to appear, which gives you a rough sense of the review queue.

Whether the change moves your position is a separate question with a slower answer. A category change can be visible on your listing the same day and take several weeks to show up in where you rank, because ranking is recalculated against everyone else in the pool rather than against your listing alone.

This is why the sequence matters. Do the fastest, most consequential field first, and the primary category is the strongest field on the profile by a wide margin, scored first at 227 points in a 2026 survey where 47 local search experts ranked 187 factors. Then change nothing else for a few weeks so the result is readable.

One change at a time is not caution, it is the only way to learn anything.

What takes months no matter what I do?

Anything that accrues rather than toggles.

Reviews. A review flow is a rate, not an event. In the same 2026 survey, recency of reviews ranks eleventh at 164 points and a sustained flow rather than bursts ranks fourteenth at 154. Two separate entries, which tells you the panel treats the shape of the flow as its own signal. Darren Shaw's summary of the mechanism is the sentence to remember: "The moment you stop getting new reviews, you're going to see your local rankings start to slip." Joy Hawkins documented the reverse in a case study where a client's rankings fell after his review flow flat lined and recovered as the reviews resumed.

Pages. New pages need to be found, indexed, and then evaluated against pages that have existed for years. Weeks before they are eligible, months before they are competitive.

Authority and mentions. Local news, associations, neighborhood blogs, and genuine partnerships accumulate slowly by nature. This is also where most wasted money goes, and citation building being partly theatre is worth reading before you buy a package that promises three hundred listings in thirty days.

Distance. Never. Proximity ranks second at 225 points in that survey and it does not improve with time or effort. If you are losing because a competitor is closer to your customers, no amount of patience changes that.

There is one shortcut everybody eventually gets offered, and it goes backwards. Routing happy customers to Google and unhappy ones to a private form looks like an accelerator and is prohibited, plus it makes you rank worse, because it strangles the flow it was supposed to speed up.

What should I look at at thirty, sixty, and ninety days?

Different things at each mark, and none of them is a single rank number.

At thirty days, look at whether the work was done. Do not evaluate results yet. Evaluate execution, because most programs fail here rather than in the algorithm.

At sixty days, look at movement across your service area rather than at one position. Compare against the geographic baseline you took at the start. If you did not take one, take one now and start the clock again, which is a real cost of skipping it.

At ninety days, look at actions. Calls, form fills, direction requests and website clicks, compared to the same period last year if you have it, and to your own trailing quarter if you do not. This is the first point where the question of whether it is working has an honest answer. Compare it to your own numbers and nothing else, because there is no current published benchmark for Business Profile performance: the only large study is from 2019 and Google has since redefined several of the metrics it was built on.

What does the thirty day execution check actually look like?

How do I tell slow from broken?

By whether the leading indicators are moving, even when the position is not.

Slow looks like this: reviews arriving weekly, impressions rising, profile actions creeping up, position stable or improving in the parts of your service area that are realistic. The lagging number has not turned yet, but everything upstream of it has.

Broken looks like this: no new reviews, no completed work, nothing changed on the profile in six weeks, and a report full of charts about keywords. Or a profile that is invisible for your own business name, which is a technical problem rather than a patience problem.

Slow can also be structural rather than yours. Sterling Sky's 2026 field data records AI powered local packs on roughly 7% of tracked keywords, showing one or two businesses instead of three with no call buttons, and local pack ads growing from about 1% of mobile reports in early 2025 to about 22% by December 2025. If the pack in your market is being replaced or crowded by paid results, your work can be succeeding while your visible position stands still.

Why does a weekly progress check tell me nothing?

Because the measurement noise is larger than the movement, and this is provable rather than a matter of opinion.

Owners increasingly check progress by asking an AI assistant whether it recommends them. It feels like a fast read. It is not a measurement, and the research on why is specific. A 2026 study from the University of St. Gallen ran identical prompts across ChatGPT, Gemini, Google AI Mode and Perplexity, both daily over a 45 day window and repeatedly within a single day. Repeated runs of the same prompt on the same day shared only 32% to 43% of their cited sources. ChatGPT was the least stable, at a source overlap of 0.233. Day to day, roughly 65% of sources changed overnight.

The paper's prescriptions are the useful part. At least seven runs per prompt per day, and reporting on a two to four week rolling aggregate rather than week over week. With a standard error near 0.10 at seven runs, a visibility score moving from 40% to 45% in a week is statistically indistinguishable from no change.

So a weekly "we went up five points" is not a small result. It is a reading from inside the noise band. The same logic applies to checking your map position once from your office on a Tuesday. One sample, one location, one moment. Before you treat any of it as data, the tools that claim to track AI visibility deserve an honest comparison.

Posts belong in the same frame. They are worth doing for the customer reading them rather than for a ranking effect, and what profile posts actually affect is a smaller claim than most people make for them.

What about a guarantee?

There is not one, and Google says so on its own help page.

"There's no way to request or pay for a better local ranking on Google." That is the platform operator, in writing, on the page it publishes for this exact question. A guaranteed position in a set number of days is a promise about a system the seller does not control, made to a business that cannot verify it.

What can honestly be committed to is work and measurement: which fields get fixed, in what order, on what dates, measured from which points, against a baseline captured before anything changed. That is a contract you can hold someone to. A ranking date is not.

What to do this week

Write down today's date and take a geographic baseline for your top three search terms from several points across your service area.

Record last month's calls, form fills, direction requests and website clicks. Put both in one file you will still be able to find in ninety days.

Make one change: the profile field you know is wrong and that matters most. Note what you changed and when.

Start the review ask today and make it daily rather than campaign shaped, then run the two counts from the execution check at day 30. That input has the longest runway, so it should start first.

Questions about whether yours is working?

Email me at eric@seod.com with two things: the date the work started, and a list of what has actually been changed since then. I will run the execution count myself and tell you whether what you are seeing is normal for that timeline, early, or a sign that the work is not happening, plus which single indicator I would watch next.

I answer these myself. If the honest read is that it is too early to judge, I will say so plainly rather than find you a problem to buy.

Nobody can give you a responsible ranking date, and this article exists partly to say so. Search visibility gives you regular reporting on position, calls and the next priority, so progress is judged on what is actually changing rather than on a promise. There is more on local search in the library.