Skip to content

Measurement

Seasonality and benchmarks: judging restaurant SEO honestly

How to build a baseline before work starts, which comparison periods are safe, and how to separate a genuine visibility gain from seasonal demand.

Alex Opsenica, founder of Moolah Media

Alex Opsenica · Founder, Moolah Media

4 min read
On this page

Restaurant search demand is one of the most seasonal patterns in local search. Christmas group bookings, bank holidays, summer terraces, restaurant weeks and even weather all move the numbers before anyone touches a Business Profile. Reporting that ignores this credits the calendar to the agency in December and blames the agency for January.

Key takeaway

Set the baseline before the work starts, compare year on year rather than month on month, and state the seasonal context in every report.

Set a baseline before anything changes

  1. Export at least twelve months of Business Profile Performance data, or as much as exists
  2. Export the same window from Search Console, split into brand and non-brand
  3. Record covers, average spend and enquiry volume from the reservation system
  4. Take a grid snapshot of map visibility for the core terms and catchment
  5. Note fixed events — refurbishment, closure weeks, menu launches, price changes
Baseline
The measured position before work begins, captured across the same reports that will be used to judge progress. Without it, every later number is an assertion.

Source: Google Business Profile Help — Business Profile performance — Reporting windows and the interactions available for export.

Source: Google Search Console Help — Performance reports — Data retention and processing limits that constrain long comparisons.

Which comparison to use

Comparison periods, and what each can prove

Month on month

Good for
Spotting sudden breaks — a suspension, a site outage
Risk
Confuses seasonality with performance

Year on year, same month

Good for
Judging genuine progress
Risk
Needs twelve months of history and stable trading

Rolling 90 days vs prior 90

Good for
Smoothing weekly noise
Risk
Still crosses seasonal boundaries

Pre- and post-change windows

Good for
Testing one specific change
Risk
Only valid if nothing else changed

Worth avoiding

A local pack screenshot from a single device and location is not a benchmark. Local results vary by searcher position, so a grid measurement across the catchment is the only comparable baseline.

Separating demand from visibility

The useful question is not "did our numbers rise?" but "did our share of the available demand rise?". When discovery searches, branded searches and interactions all move together in a known-busy month, that is demand. When discovery searches and grid visibility rise while branded volume is flat, that is a visibility gain.

Business Profile Performance

Discovery vs direct searches

Direct searches follow brand awareness and season. Discovery searches follow whether Google is putting the venue in front of people who did not know it.

Example · hypothetical

Hypothetical: the good January

Views fall a fifth against December, as they do every year. Against the previous January they are up by half, grid visibility widened across the catchment, and direction requests rose. That is progress, reported honestly, in a month that looked like a decline.

See this in practice

What was recorded, and what it doesn't prove.

Premium steakhouse group · name withheld · Dublin 2

3,699Network covers · recorded in OpenTable

A Dublin 2 steakhouse grew covers 7% while its market fell 13%. OpenTable and Google data side by side, without claiming what they can't prove.

What was recorded, and what it doesn't prove.

What to put in the report

Seasonally honest reporting

  • Baseline stated on every chart, not just the first month
  • Year-on-year comparison wherever twelve months of data exists
  • Seasonal context written in plain English, including bad months
  • Known external events noted — closures, works, weather, local events
  • Search data labelled as correlation and booking data as outcome

Building your own benchmark

Published hospitality benchmarks rarely survive contact with a specific restaurant: a tasting-menu room in the city centre and a suburban brunch operation do not share a demand curve. The only reliable benchmark is your own history, recorded consistently.

  1. Record the same five numbers on the same day each month, from the same reports
  2. Note the operational context beside them — closures, refurbishments, menu changes, holidays
  3. Keep at least thirteen months so every month has a like-for-like comparison
  4. Mark anything that changed on the profile or website, with the date

Example · hypothetical

A hypothetical year-on-year read

Imagine December profile actions down eight per cent against last December, while non-brand impressions are up a third. If the restaurant closed for four days that had traded last year, the fall is an operating decision, not a visibility problem. Without the note beside the number, it reads as a failure.

Worth avoiding

Never compare a month against the month before it in a seasonal business. January against December will always look like a collapse, and August against July will always look like one in a business-district restaurant.

Reading a bad month without panicking

Every restaurant has months that look wrong. The useful discipline is a fixed order of checks before concluding anything about the marketing.

  1. Did trading days change? Closures, private hires, refurbishment, staff shortages
  2. Did the calendar shift? Bank holidays, school terms, a late Easter, an event that did not return
  3. Did anything change on the profile or website, and when?
  4. Is the fall present in both the profile report and Search Console, or only one?
  5. Is it present across the catchment in grid tracking, or only in a few points?

If the answer is in the first two, there is nothing to fix. If it appears in both reports and across the grid, it is a visibility problem worth investigating properly.

Worth avoiding

Reacting to one weak month by changing categories, rewriting the site and rebuilding the profile at once removes any chance of knowing what caused the recovery.

Common questions

How long before restaurant SEO results are readable?

Profile and map movement often appears within weeks, but a defensible judgement needs a full quarter, and ideally a year-on-year comparison.

Is a drop in December to January a problem?

Usually not. It is the normal shape of restaurant demand. Compare against the same month last year before concluding anything.

What if I have no historical data?

Start the baseline now and report month on month with explicit seasonal caveats until twelve months of history exists.

Can weather really move search demand?

For venues with terraces, gardens or heavy walk-in trade, yes. It belongs in the report as context, not as an excuse.

Working on this yourself? Reporting and measurement.

Not sure which numbers to trust?

The audit sets out what your current reporting can and cannot prove, and which measures would actually tell you whether search work is paying.

Get Your Free Visibility Audit

Written by

Alex Opsenica, founder of Moolah Media

Alex Opsenica

Founder, Moolah Media

Founder of Moolah Media, a founder-led restaurant marketing specialist based in Ireland, working with restaurants worldwide on websites, SEO and Google Maps, and Google Ads. Around 20 years in hospitality, including hands-on restaurant management.

More about how Moolah works.

Continue exploring

More on MeasurementAll insights