When Google Business Profile website clicks fall, the right response is not automatically "publish more content". First establish what changed, which customer action matters to your business, and whether the outcome you care about changed too.
Recent portfolio data provides a useful reason to review that approach. It does not provide a universal benchmark for every business or prove that AI caused a decline in your own enquiries.
Reviewed 8 September 2026.
What the Q2 2026 data actually shows
In its updated analysis of local search performance, GMBapi reported the following year-on-year changes across its US portfolio for Q2 2026:
| Profile action | Reported change |
|---|---|
| Website clicks | Down 12.5% |
| Calls | Down 11.9% |
| Direction requests | Up 21.1% |
The study describes a changing customer portfolio and retrospective data imports, and explains why earlier Q1 findings were revised. It is not a fixed-cohort experiment. Geography and account mix matter, and the US figures should not be treated as a forecast for an individual Irish, Spanish or UK business.
Our conclusion is to investigate the mix of customer actions, not to declare local SEO dead. A movement in one metric is a starting point for diagnosis.
Know what each Google metric means
According to Google's Business Profile performance documentation, calls count clicks on the call button, website clicks count clicks on the website link, and directions measure direction requests. Performance can include organic results and Google Ads.
A call-button tap is therefore not proof of an answered call. A direction request is not proof of a completed visit. Neither should automatically be assigned the value of a sale.
In your working report, place the observed action beside the business outcome you ultimately need. That makes missing evidence visible instead of hiding it inside a reassuring total called "engagement".
Judge the result against your business model
Consider two illustrative businesses: a shop that mainly serves walk-in customers and a specialist contractor that needs customers to telephone before booking.
For the shop, more direction requests could be encouraging, but we would still want evidence about visits and sales before calling it growth. For the contractor, fewer answered qualified calls could be a serious issue even if direction requests increased.
These examples show why we recommend an outcome-led interpretation. Do not assume that an increase in one action compensates for a decrease in another. Write down the customer journey you actually want to support.
For a training provider, that journey might include researching course suitability, visiting a booking page and paying a deposit. For a restaurant, it might include choosing a location, booking a table or arriving during opening hours. Your report should reflect that distinction.
Make the comparison trustworthy
Start with the same locations, comparable periods and a consistent definition of each metric. Keep a separate view of newly opened or closed locations. Note changes in advertising, opening hours, promotions, contact details and measurement setup.
For multi-location businesses, we recommend reporting both the total portfolio and a like-for-like group. That helps separate growth from adding locations from changes at existing locations. Explain exclusions rather than quietly removing poorly performing branches.
Check the volume behind a percentage. A fall from ten actions to eight and a fall from one thousand to eight hundred have the same percentage change, but create different operational questions. Report counts beside rates.
Also compare the business calendar. Holiday closures, local events or a change in available appointments can matter to interpretation. They are competing explanations to examine, not reasons to dismiss a genuine problem.
Inspect the route to the intended action
Our recommended review is practical: confirm that the profile's phone number, website destination, booking route and opening information match the current business. Follow the customer path yourself rather than relying on a settings screenshot.
On the website, ask whether a visitor can quickly determine suitability and take the next step. For a service page, useful information might include the service area, relevant limitations and what happens after an enquiry. Use accurate information that the business can maintain.
For the phone journey, distinguish button taps from answered enquiries using the records available to the business. Do not assume Google supplies every downstream outcome. A visibility problem and an unanswered-phone problem need different fixes.
Do not blame AI without evidence
The portfolio findings show an observed pattern. They do not isolate the causal effect of AI on your own profile.
Before changing strategy, consider competing explanations: a different customer mix, an advertising change, seasonal demand, an inaccurate destination or operational availability. Record which explanation the evidence supports and which remains speculative.
You can also investigate local AI recommendations separately, using a stable set of relevant customer questions. Our citation tracking guide explains why one answer should not become a confident visibility verdict.
A local SEO report worth acting on
Our suggested monthly report combines observed profile actions, website outcomes where identifiable, and the business's own enquiry or sales evidence. State the main change, its likely commercial importance, the uncertainty and the next action.
For example: "Website clicks fell, but qualified course enquiries were steady. Review the booking route and maintain a comparable baseline before increasing content production." That is more useful than celebrating a ranking while ignoring the booking diary.
For hands-on support, explore UpSearch managed SEO. For a broader measurement framework, read AI search attribution beyond referral traffic.
The right question is not simply whether somebody clicked. It is whether the business remains easy to find, evaluate and choose, and what evidence supports that conclusion.
FAQ
Are Google Business Profile calls the same as completed calls?
No. Google's metric concerns interaction with the call button. Use appropriate business records to understand answered calls, qualified enquiries and resulting bookings.
Are more direction requests enough to prove more customers?
No. They are a useful action signal, but we recommend checking visits or sales evidence before claiming a commercial gain.
Should a European business use the US figures as its target?
No. Build a local baseline with comparable locations and periods. Use the study as a reason to investigate your own customer journey, not as a universal performance target.