Running a local visibility audit: the eight-step method

An audit establishes a state at a point in time — it does not set up ongoing tracking. Here are the eight steps, with the scoring grids you can use as they are.

A local visibility audit establishes a state at a point in time and produces a prioritised list of workstreams. It runs in eight steps, in this order: scope, profile completeness score, geolocated rank reading, competitor analysis, citations, website signals, reputation, deliverable. Each step produces an element of the final report — if a step produces nothing, it does not belong in the audit.

A word on the order, because it is not arbitrary. Profile completeness comes second because it is the variable most strongly associated with local pack positions 1 to 3 in the available data — 0.71 correlation per Visionary Marketing — and because 67% of profiles are incomplete. Statistically, that is where the work is.

Step 1 — Scope

Four decisions to take before measuring anything, and to write down at the top of the report.

What scope has to fix

  1. The locations in question, with their addresses and profile identifiers.
  2. The queries: three to ten service queries customers genuinely type, not variations of internal vocabulary.
  3. The areas: the neighbourhoods, towns or districts customers actually come from, taken from the customer file rather than assumed.
  4. The sampling plan: twelve to twenty locations for a single site, four to six per site for a network, chosen by a single documented rule.

Deliverable: a scoping page. Without it the audit is not reproducible in six months, and the before/after comparison becomes impossible.

Step 2 — Profile completeness score

The grid below scores out of 20, one point per line, location by location. It is usable as it stands.

#CriterionEarned if
1Business nameIdentical to the real-world name, no added keywords
2Primary categoryMatches the actual dominant trade
3Secondary categoriesAt least one, all genuinely practised
4Address or service areasFilled in and verified
5PhoneLocal number, identical to the website's
6WebsitePoints to the location page, not the group homepage
7Regular hoursFilled in and accurate
8Special hoursCurrent for the next 3 months
9DescriptionWritten, with no keyword stuffing
10Services or productsListed, using customer wording
11AttributesFilled in (accessibility, payment, facilities)
12Logo and cover photoIn place
13Photo volumeAt least 10 genuine photos
14Photo freshnessMost recent under 3 months old
15Booking or ordering linkPresent where the business supports it
16Questions and answersAt least three, including your own
17Recent reviewsAt least one in the last 90 days
18Review response rateAbove 80%
19Pending editsNone outstanding for more than two weeks
20Third-party suggestionsNone accepted without review

Deliverable: a score out of 20 per location, and a named list of the missing lines. Across a network, this score ranks sites from most to least urgent. What each field means is covered in the Business Profile optimisation guide.

Step 3 — Geolocated rank reading

This is the core of the audit, and the step you cannot do by searching for yourself on Google — the results would be skewed by your own position, account and history.

For each keyword × location pair, record: rank in the local pack, rank on Google Maps, and which businesses appear ahead of you. Always keep the two surfaces separate: they do not rank the same way.

Deliverable: top 3 share of voice, visibility radius at both thresholds, and a rank-by-location table. The formulas are in the local SEO KPIs.

Step 4 — Local pack competitor analysis

The competitors to analyse are not the ones the client names: they are the ones actually appearing ahead of them, and they change from one location to the next. Read them off step 3 rather than starting from a supplied list.

For the three to five that recur most: primary category, number of secondary categories, review volume and rating, reviews received over 30 and 90 days, response rate, photo count, profile age.

Deliverable: the benchmark grid, detailed in the competitor analysis article.

Step 5 — Citations and NAP consistency

Two measurements, and the second matters more than the first. The number of citations: where the business is listed. The consistency: whether the name, address and phone are written there exactly as on the profile.

According to Localo, gains from citation volume flatten in a band of 21 to 27: beyond that, adding directories produces little. And per Visionary Marketing, the correlation of citations with positions 1 to 3 has fallen to 0.34, the sharpest decline of any factor tracked.

Deliverable: the list of existing citations, the ones carrying wrong information — to fix first — and the missing directories that matter in the sector. The subject is covered in detail in the local citations article.

Step 6 — Website signals

A local audit does not turn into a technical audit: three points are enough, and they are the ones most often missing.

Avantages

  • One location page per site, with content of its own
  • NAP shown in plain text on that page, spelled exactly as on the profile
  • LocalBusiness structured data consistent with the profile
  • A page that loads properly on mobile — that is where local traffic arrives

Inconvénients

  • A single "our branches" page listing forty addresses
  • Generated location pages, identical but for the town name
  • An address in an image, unreadable by search engines
  • A profile website link pointing at the group homepage

Deliverable: an inventory of location pages that exist, are missing, or are duplicated.

Step 7 — Reputation

Three numbers and one comparison. Review volume and rating, compared to the competitors actually shown rather than a national average. Velocity: reviews received over 30 and 90 days — the most predictive review measure of the top 3 per Visionary Marketing, at 0.64 against 0.41 for total volume. And the response rate, with the median delay.

Deliverable: your reputation standing against the local pack competitors, and the diagnosis: does the gap come from the stock, the flow, or the ability to reply? The method is set out in the review management article.

Step 8 — Deliverable and prioritisation

An audit returning fifty findings without ranking them will not be acted on. Prioritisation runs on two axes, and the table fits on one page.

PriorityEffortExpected effectTypical examples
1 — NowLowHighFill in services and attributes, fix wrong hours, reply to outstanding reviews
2 — This quarterMediumHighPhotos, secondary categories, review collection process, NAP fixes
3 — PlannedHighHighLocation pages per site, structured data, citation plan
4 — WatchVariableUncertainProfile posts, adding directories beyond thirty or so

Every workstream carries an owner, a deadline, and the metric that will say whether it produced anything. Without that last column, the next audit starts from scratch.

Steps 3 and 4 have to be measured, not estimated

The geolocated reading and the analysis of competitors actually shown are the two steps that need a tool. Pinperf records your rankings from the locations in your sampling plan and returns, for each one, your rank and the businesses ranked ahead of you.

Explore Pinperf

You can also hand us the whole audit: we run it to this method.

Half a day to a day for a single location once the rank reading is available, with most of the time going into steps 2, 4 and 8. Across a network, budget roughly a day per ten locations, the completeness grid being identical everywhere.
No. An audit establishes a state at a point in time and produces workstreams; tracking measures the effect of those workstreams over time. An audit without tracking never tells you whether the fixes produced anything.
Scope, then the profile completeness score. Completeness is the variable most strongly associated with positions 1 to 3 in the available data, and 67% of profiles are incomplete: statistically, that is where the work is.
Seven steps out of eight, yes, with a spreadsheet and the grids in this article. The third — the geolocated rank reading — needs a measurement tool: searching for yourself on Google returns a result skewed by your own position, account and history.

Sources

Read next

All posts