A Restaurant Video Got Views. How Many Became Actual Diners?
A full comment section and a quiet dining room can exist at the same time. Restaurant teams need a way to distinguish attention from visits before deciding whether a post worked. The useful starting point is a small measurement chain, not a guessed conversion percentage.

Photo: Heritage Singapore / Eventful Singapore
Start with six different outcomes
Views describe exposure under each platform’s definition. Profile visits suggest further interest; link clicks indicate a next step. Confirmed bookings, attended covers and return visits are separate outcomes. A booking for six is not six bookings, and a cancellation is not six diners served.
A simple worksheet can hold a row for each campaign and separate columns for these outcomes. Missing data should remain marked unknown. Filling the gaps by multiplying views by an assumed conversion rate creates a confident-looking number without evidence.
For a packaged food product rather than a table reservation, the same measurement discipline applies: label product-page visits, checkout starts and completed orders separately. A marketing team should compare the outcome it can verify, not rename every click as a sale.
Give the campaign a recognisable link
Google Analytics supports campaign parameters such as source, medium and campaign. A restaurant might label a link with utm_source=instagram, utm_medium=organic_social and utm_campaign=weekday_lunch_nov. Consistent labels make comparisons easier than a new naming style for every post.
A team should test that the destination opens on mobile and that the intended parameters appear in its actual reporting. Tags help identify referred traffic; they do not automatically establish that a reservation was completed. Booking platforms, cross-domain journeys and reporting configurations can leave gaps.
Reconcile bookings with the service record
Where the restaurant’s systems support it, the campaign report should distinguish confirmed, cancelled and attended reservations. The comparison belongs at the campaign level, with the agreed reporting window written down before results are reviewed.
A staff question such as how guests heard about the restaurant can add context, but recall is imperfect. Someone may have seen several posts and then booked through search. Self-reported discovery and tracked link attribution should therefore remain separate columns, not be added together as unique conversions.
Keep personal details out of campaign tags
Google explicitly prohibits personally identifiable information in Analytics. Campaign URLs should contain campaign labels, not a diner’s name, email address or phone number. A restaurant does not need those details in a public link to distinguish a lunch campaign from a new-menu campaign.
This is also a practical reason to keep a shared marketing worksheet aggregate. Staff responsible for reservations can work in the proper booking system rather than copying customer records into a content report.
Make the next decision from the weakest observable step
Strong viewing but few next steps can justify checking whether the offer and destination were clear. Clicks without completed bookings can justify testing the mobile booking path, available times and outlet match. Neither pattern alone proves the cause.
The next experiment should change one plausible friction point and compare like-for-like periods where possible. Holidays, weather, capacity and concurrent offers belong in the notes. The goal is a better operational decision, not claiming every occupied table as a social-media win.
A related conversation for F&B owners
F&B owners exploring these questions can read about the Heritage networking evening on 3 December, from 7–10pm at L^IFE Camping Ground, Level 2, *SCAPE. Entry is free and strictly for F&B owners, with approval required. No food or drinks are provided. Current registration details are on Luma.



