What a Business Profile can and cannot tell you
The Business Profile optimisation tutorial ends with a routine: read the profile's performance section every month. That section shows how people found the listing, including the search terms they used, branded and unbranded, and what they did with it, from calls and direction requests to website clicks, messages and bookings, with the trend over time. Business Profile Help, listed in the sources, is the reference for the exact metrics and their definitions, which change from time to time; this page does not quote them.
What the section cannot do is follow anyone past the action. A website click is a click, not a booking. A direction request is an intention, not a visit. A call is a call, not a customer. The measurement plan framework starts from that gap: decide the questions first, then put each answer where it can actually be recorded.
| Question | Where the answer lives | What it cannot tell you |
|---|---|---|
| How many people saw the listing, and for which searches? | The profile's performance section | Whether they were already customers |
| What did they do on the listing? | The profile's performance section | What the action led to |
| What did profile visitors do on the website? | Analytics, once the profile's links are tagged | Anything about visitors who never clicked |
| How many calls came from the profile, and which were enquiries? | The phone log or a call-tracking system | Calls from people who saved the number months ago |
| Which new customers chose the business because of the listing? | A how-did-you-hear answer at booking | Anything exact; memory is imperfect |
A chain reads the same table for every site; the multi-location hub explains how to report each measure per location and compare sites fairly. For store networks that also buy media, the measurement plan for store visits reads profile actions per location beside the media delivered into each catchment. This tutorial is the single-location version and the building block for both.
Before you start
- Owner or manager access to the profile, so you can read its performance section and edit its links.
- Editor access to the analytics property for the website, and agreement from whoever owns it before you add or change key eventsA GA4 event marked as important to the business, the property's measure of success..
- The business's UTM naming convention, or a decision to write one. The UTM and naming conventions guide sets out the five parameters and the rules that keep them consistent.
- The phone arrangements: which number the profile shows, whether a call-tracking service is in use and who owns the business's listings, because a change to the number on the profile touches every listing.
- The booking or enquiry records, in whatever form they exist: a booking system, a CRM, a diary.
- The profile's own figures for as far back as they go, or a note of when the profile started, so seasonal comparisons are honest from the start.
The steps
Write down the questions the numbers must answer
Start with the few questions the owner actually asks, written in plain words: are more people finding us than last year at this time; are the people who find us calling, booking or turning up; which of the profile's actions turn into customers; and did the change we made last month help. Each question names the measure that answers it and where that measure lives, which is the first column of the table above. A question nobody will act on is dropped. The point of the list is the later conversation: when a number moves, the business already knows which question it answers and what it would do about it.
Record a baseline from the performance section
Before changing anything, record the profile's own figures as they stand: searches and views, then each action, by month, for as far back as the section shows. This is the BaselineAn estimate of what a metric would have done without a given change, campaign or promotion, against which the effect is measured. every later month is read against, so note anything unusual that happened in it, such as a closure for a holiday, a wrong phone number that was later fixed or a burst of reviews. Write down which metrics the section offered when you took the baseline. If the profile later renames or redefines one, the note shows where the series breaks, and the guide to comparing periods explains how to read across a break like that rather than mistaking it for a change in demand.
Tag the website and booking links with campaign parameters
Google Analytics Help's page on URL builders, listed in the sources, describes adding campaign parameters to a link so that analytics records the visit's source, medium and campaign. Add them to the profile's website link and to any booking, menu or quote link it carries, so each can be told apart. Use values from the business's own convention, never ad hoc ones, and keep them in lower case: the naming conventions guide explains why case and separators matter, and the UTM link request SOP shows how a team builds, tests and logs each link rather than typing it by hand. Point the website link at the right page; for a business with several sites, the profile tutorial sends each profile to its own location page rather than to the homepage.
The profile's links are links the business places itself, which is the case manual tags are made for. The comparison of UTM parameters and auto-tagging explains where each method applies, so the profile's tags do not collide with an ad platform's.
Check the tagged visits arrive as their own source
Open the tagged link from a phone and a desktop, then look for the visit in analytics under the source and medium you chose, and check the grouping and the landing page. On grouping: analytics turns source and medium into a Channel GroupingRules that assign each visit or conversion to a channel, such as organic search or email, from its source, medium and campaign values. by documented rules, and the naming conventions guide explains that a medium the default rules do not recognise falls into Unassigned, as the diagnosis for Unassigned traffic sets out. Decide whether profile visits should sit inside organic search or have a channel of their own in a custom group, and choose the values accordingly. On the landing page, the tags must survive any redirect, or the visit arrives as direct, the leak described in the diagnosis for too much direct traffic. The page itself should carry a self-referencing canonical tag, as the guide to location pages and listings at scale recommends, so the tagged address is never treated as a second copy of the page.
Mark the key events a profile visitor can complete
A profile visitor who reaches the website can do a small number of things that matter: book, send an enquiry form, start a chat that captures a request or buy. Mark those, and only those, as key events, following the key events guide. Resist marking a tap on the phone number. The guide to measuring calls and chats treats a phone tap as a diagnostic rather than an enquiry, because it shows intent to call, not a call that connected or a customer who called. If the business takes most bookings by phone, the website key events will understate the profile's work, and the next step matters more than this one.
Decide how profile calls are counted once
Calls are where profile measurement most often double counts or does not count at all. The profile counts taps on its own call button; a call-tracking service, if one is used, counts calls to its numbers; the phone log counts everything. Pick one count of record. If a tracking numberAttributing phone calls to the marketing source that produced them, usually by showing different tracking numbers per source or visitor. is put on the profile to tell its calls apart, the profile tutorial describes the documented arrangement: the tracking number as the primary number with the main number kept as an additional one, so the business's details still match its other listings. Agree any such change with whoever owns the listings first, as the calls guide asks, because the citations and NAP guide depends on the main number appearing the same way everywhere. Then use the calls guide's rule for which calls are enquiries, so the monthly figure is enquiries from the profile, not rings.
Add the how-did-you-hear answer to bookings
No tool sees the person who read the reviews on the listing, drove past twice and then walked in, or the one who saved the number from the profile last spring. Ask. The guide to self-reported attribution covers wording the question and coding the answers; for a local business the useful answer codes are usually the listing or Maps, a search, a recommendation, the premises seen in passing, social media and other. The local content and community marketing guide uses the same question to measure community activity, so one question at booking serves both.
Build the monthly table
One row per month, with columns in the order a profile visitor moves: searches and views from the performance section; calls, direction requests, website clicks, messages and bookings from the same section; sessions and key events from the tagged links in analytics; enquiry calls from the count of record; new customers who named the listing at booking; and, from the reviews and reputation guide, the month's new reviews and the rolling rating, because those are what profile visitors read before they act. Add a notes column for anything that changed: new photos, edited hours, a closure, a new service, a burst of reviews. Without it, next year's comparison cannot tell a cause from a coincidence.
Read it against the profile's own history and act
Compare each month with the same month last year and with the baseline, never with another business's figures, which the profile does not show and no cited source publishes. Local counts are small, so a month's movement can be noise: the guide to reading small numbers explains why a rate built on a handful of calls jumps around and how to read a range rather than a point. Look for a change that persists for more than one month and that the notes column can explain. Then decide one action, such as answering reviews faster, moving photo and service updates ahead of the busy season, or fixing a booking link that loses visitors, and write it in the notes so next month's row can show whether it helped. If actions fall while nothing in the notes explains it, the local marketing troubleshooting pages start from the symptom.
The local presence hub sets this measurement beside the reputation and community work it is meant to judge; neither replaces the other.
A bike repair workshop measures its profile for a season (illustrative)
Ibexquay Bike Workshop measures its profile through spring
Ibexquay Bike Workshop, an invented one-site bicycle repair shop that takes bookings online and by phone, had read its profile's performance section now and then but could not say what the listing produced. In February the owner records a baseline from the previous year: 1,840 profile views in April, 212 website clicks, 101 calls and 140 direction requests, with the note that the booking link was missing from the profile until May.
She tags the profile's website link and its booking link with the shop's convention (source google, medium organic, campaigns gbp-website and gbp-booking), so profile visits sit inside organic search but can be filtered by campaign, and checks both on her phone. Online bookings and the contact form are the only key events. The profile keeps the main number, and calls are counted from the phone log, where the person answering marks new repair enquiries. Every booking now asks how the customer heard about the shop.
| Month | Profile views | Calls (profile) | Booking-link sessions | Online bookings from profile | New enquiry calls | Named the listing |
|---|---|---|---|---|---|---|
| March | 1,420 | 71 | 58 | 19 | 34 | 22 |
| April | 2,010 | 104 | 131 | 47 | 52 | 41 |
| May | 2,260 | 118 | 149 | 55 | 57 | 46 |
Two things stand out. Most bookings from the profile start on the booking link rather than the website link, so the owner checks that the booking page works well on a phone and brings the profile's services up to date before the busy weeks. And April's views are up on the baseline while its calls are about level, which the notes explain: last April had no booking link, so more of this year's demand books online instead of calling. In June she adds photo and service updates to the calendar for the fortnight before each seasonal rise, and records the change in the notes column.
Common mistakes
- Untagged profile links. Profile visits then disappear into other search traffic or direct, and the listing's contribution to the website cannot be separated.
- A medium nobody checked. A value the channel rules do not recognise sends every profile visit to Unassigned.
- Phone taps as key events. A tap is intent to call, not an enquiry; counting it inflates the profile's results.
- Counting calls in three places and adding them up. Pick one count of record and reconcile the others against it.
- A tracking number replacing the main number everywhere. Keep the main number on the profile as the documented arrangement allows, and agree the change with whoever owns the listings.
- Comparing with a competitor or a published average. The profile shows only its own figures; read them against its own history and the same month last year.
- No notes column. Without a record of what changed, a rise or a fall cannot be explained a year later.
- Reading one month as a trend. Small local counts move by chance; wait for a change that persists.
