A marketing report that ends the guessing

Last month, an owner paid for a leaflet drop, boosted two social posts, sent an email to regulars, upgraded a listing on a booking platform and sponsored the local school fair. Bookings were up on the month before. Which of those five things worked? The honest answer, in most restaurants, is that nobody knows, and next month's money will be spent the same way, on the same guesswork.

A marketing report ends that guessing, or at least narrows it. On one page, it sets out each channel's spend, the response it produced, the bookings and covers that can reasonably be linked to it and what those covers left behind after their costs. It also states, every time, how sure each line is, because some channels can be measured closely and others cannot.

This guide covers what to put in the report, how to link bookings to channels as honestly as possible, the caveat about attribution that belongs on every report and how to use it to decide what to keep and what to stop. Every venue and figure in it is an example.

A marketing report exists to answer one question: which of the things we pay for bring in guests, and which do not? Everything on it should help answer that, and anything that does not, however interesting, belongs somewhere else.

It is a decision tool, not a scrapbook. A report that lists every post, every like and every email sent, without connecting them to guests and money, looks busy and decides nothing.

Most restaurant marketing reports guess because they measure what is easy rather than what matters. Likes, followers, opens and impressions are easy to count. Bookings, covers and margin that came from a channel are hard to count, so they are left out.

The result is a report that shows activity rising while nobody can say whether it brought a single extra guest. The fix is not more figures but different ones: fewer lines, each one closer to the guest in the chair.

Put every channel on one page, in the same order every month. For each, the same lines: what it cost, the response it produced, the bookings and covers linked to it, what those covers left behind and how sure the link is.

Keeping every channel on the same lines makes them comparable. The email that cost almost nothing and the leaflet drop that cost a great deal can be set side by side, and the question of which brought more guests for the money can at least be asked.

Spend is the easiest line and the one most often wrong. Include everything each channel costs: the printing and the delivery of the leaflets, the boost and the time spent making the post, the platform's fees and any commission per cover, the photographer, the free desserts given away with an offer.

Time counts too. A manager who spends four hours a week on social media is spending money on it, even if no invoice arrives. Put a sensible value on the time, or at least note it, so the cheapest-looking channels are not flattered.

Response is what each channel produced before anyone booked: clicks on an email's links, visits to the website from a post, calls after a leaflet drop, enquiries from a listing. It shows whether a channel reached anyone at all.

Response is not the result. An email with many clicks and no bookings has reached people without persuading them. Keep the response line, but never let it stand in for bookings.

Bookings and covers are the lines that matter. For each channel, record the bookings that can reasonably be linked to it, and the covers those bookings brought.

Covers matter more than bookings, because a booking for eight is worth more than a booking for two. And seated covers matter more than booked ones, because a booking that never arrives brought nothing but an empty table.

The final line for each channel should be margin: what the covers it brought left behind after their food and drink costs, any discount and the channel's own cost. Revenue alone flatters channels that bring in guests at a low spend or with a heavy discount.

A channel that brings forty covers at a deep discount may leave less behind than one that brings fifteen at full price. Only margin shows the difference, and it is the line that should decide where next month's money goes.

The hardest part of any marketing report is attribution: deciding which channel deserves the credit for a booking. A guest may see a post, receive an email, walk past the leaflet on a noticeboard and then book because a friend mentioned the venue. Which channel brought them?

There is no perfect answer, and every method has its bias. Crediting the last thing the guest saw favours email and search. Crediting the first favours social media and print. Asking guests relies on what they remember. The report needs to say which method it uses.

State the caveat on every report, in one or two sentences, so nobody reads more certainty into it than it holds. For example: "Email bookings are counted from guests who clicked the booking link within a week of the send. Social and leaflet bookings are from guests who mentioned them. Some bookings came from more than one channel and are counted once."

Mark each line with how sure it is: measured, estimated or unknown. A reader who sees "leaflets: estimated, from guests asked" weighs that line differently from "email: measured, from clicks", and rightly so.

The more each channel has its own way in, the easier the counting. Separate booking links for each channel, where your tools allow them, a different phone number on the leaflets, or a landing page for each campaign all help separate one channel's guests from another's.

Even without them, the booking link in an email can be counted through its clicks, and a guest who clicked and then booked in the following days is a reasonable match. The closer each channel's path to the booking, the more of the report can be measured rather than guessed.

Asking guests how they heard of the venue is the oldest method and still one of the most useful for channels that cannot be measured any other way: print, word of mouth, local events, a review in the paper.

Ask in the same way each time, at booking or when guests arrive, and record the answer. It is imperfect, because guests forget and simplify, but over a month it shows the broad shape of where new guests come from.

A named offer for each channel, such as a dessert for anyone who mentions the leaflet, makes a channel countable: every guest who mentions it is counted. It works well for print and local events, which otherwise leave no trace.

Keep the offer modest, and count its cost in that channel's spend. An offer generous enough to bring in guests who would have come anyway lowers the margin on every one of them, and the report should show that too.

A channel deserves credit only for the guests who would not have come without it. That needs a baseline: what bookings would have been without the campaign. The simplest baseline is the same weeks last year, or the weeks just before the campaign, allowing for anything else that changed.

If bookings usually rise in the spring, a spring campaign will look successful whatever it did. Comparing with the usual rise, not with a flat line, keeps the report honest.

For a single campaign, compare the bookings and covers in the week or two after it with the same weeks before, and with the same weeks last year. If both comparisons show a rise beyond the usual, the campaign probably helped.

If only one shows a rise, look for another cause: the weather, a local event, a good review elsewhere. A rise that appears in every comparison is far more convincing than one that appears in only one.

Campaign reporting, following a single email or campaign from the send to the seated guest, feeds the marketing report its most measurable lines. Sends, delivery, clicks on the booking link and the bookings that followed can each be counted.

Our guide to campaign reporting: from send to seated guest covers how to follow each step and where the chain usually breaks.

Promotion reporting asks a harder question: did a promotion bring in guests who would not otherwise have come, at a margin worth having? A set menu or offer that fills the room with guests who would have come anyway, at a lower spend, has cost the venue money.

The marketing report should carry each promotion's covers, its cost and its margin, with the baseline. Our guide to promotion reporting: did it actually work? covers the method in full.

Email is usually the easiest channel to measure, because each send goes to known guests and each click can be counted. Report sends, clicks on the booking link, bookings from those who clicked within a set window and the covers they brought.

Report unsubscribes too. A campaign that brings a few bookings but loses many subscribers may cost more in future than it earned this month.

Social media is harder. Reach and likes are easy to count, but they say little about guests. Report the visits to the website or booking page from social posts, where you can count them, and the guests who say they came because of a post.

Separate paid posts from unpaid ones. A boosted post has a cost and should be judged like any other paid channel. Unpaid posts cost time, which the spend line should note.

Booking platforms and listings often report their own bookings, which makes them look easy to measure. But a guest who would have booked directly, and booked through the platform because it came up first in a search, is not a new guest: the platform has charged a fee for a guest the venue already had.

Report platform bookings with their fees per cover, and note how many of those guests are returning regulars. A platform that mostly brings regulars is costing commission rather than bringing new guests.

Print, local sponsorship and word of mouth are the hardest to measure and often the most valuable for a neighbourhood venue. Use named offers, a separate phone number or a question at booking to give them a line, and mark the line as estimated.

Do not drop them from the report because they cannot be measured closely. A channel with an estimated line is still a channel the owner is paying for and deciding about.

Separate the guests each channel brings into regulars and new guests. A channel that mostly reaches regulars is retention: keeping existing guests coming back. One that mostly brings new guests is acquisition: finding guests who did not know the venue.

Both matter, but they are judged differently. Retention should cost little per cover, because the guests are already known. Acquisition can cost more per guest, if those new guests come back and become regulars.

For channels meant to find new guests, the most useful figure is the cost per new guest: the channel's spend divided by the new guests it brought. Set it against what a new guest is worth over their first year, if they return.

A channel that costs a lot per new guest may still be worth it if those guests become regulars. One that brings new guests who never return is expensive at any price.

Report monthly. Weekly figures are too noisy for most channels, because a single large booking or a rainy weekend can swing them. Quarterly reports come too late to stop money being wasted.

Look at each campaign as it ends, for the measurable channels, and gather everything into the monthly report. Over a year, twelve monthly reports show which channels earn their place and which only seem to.

The marketing report and the marketing budget belong together. The budget decides where money goes; the report shows where it worked. Each month's report should change next month's budget, moving money from the channels that bring little towards the ones that bring guests at a good margin.

Our guide to a marketing budget a single venue can defend covers how to set the budget and split it between channels.

The most valuable decision a marketing report supports is what to stop. Most venues keep paying for channels out of habit long after they stopped working. A report that shows a channel bringing few covers at a high cost, month after month, is the case for stopping it.

Stop one thing at a time and watch the next month. If bookings do not fall, the channel was not bringing much. If they do, the report has shown its value, and the money can go back.

Marketing results move with the seasons. A campaign in a busy month will look stronger than the same campaign in a quiet one, because some of the guests it seems to bring would have come anyway. Compare each month's report with the same month last year as well as with the month before.

Note the season beside any campaign that looks unusually good or bad. A spring email that filled the room may simply have gone out in the weeks when the room always fills, and the same email in the winter is the real test.

A partner or investor reading the marketing report wants three things: what was spent, what it brought and how sure you are. Keep their version to the totals for each channel, the margin they produced and the caveat, with a sentence on what you will change next month.

Resist the temptation to present every estimate as a measurement. An investor who later finds the figures were guesses presented as facts will doubt every report after it; one who was told plainly which lines were estimated will trust the rest.

A marketing report should fit on one page and take five minutes to read. If it grows, move the detail of each campaign into its own note and keep only the channel lines and the caveat on the page.

A short report gets read by the people who decide where the money goes. A long one gets filed, and the guessing it was meant to end carries on.

An invented brasserie spends on five channels in one month. Its report shows email: very low cost, measured from clicks on the booking link, forty covers from regulars. Boosted posts: moderate cost, estimated from guests asked, twelve covers, most of them new. A booking platform: commission on sixty covers, of which half were regulars who had booked directly before.

Leaflets: high cost, a named offer counted eight covers. The school fair: modest cost, unknown effect. The owner keeps the email and the boosted posts, encourages regulars to book directly to save the platform's commission, stops the leaflets and keeps the fair for the goodwill, noting it as unmeasured. The next month's report shows covers holding steady at a lower total spend.

servearound does not join marketing spend or channels to bookings, and it credits no booking to a campaign, so the caveat in this guide applies to its figures too. Campaign analytics show emails sent, the delivery rate, the guests who clicked, the click rate and where the clicks went, including the booking link in each email, along with list growth and unsubscribes; they count no opens. Reservation analytics show bookings and covers by source: the booking page, the booking widget on your own website, phone, staff entered and walk-in, with covers by day and hour for the baseline.

Guest analytics show new and returning guests by day and the return rate by the week guests were first seen, which shows whether a month's new guests came back. Social media, print and listings do not appear as sources, and there are no tracked links for a venue's own channels, so their lines come from your own counts. See Analytics and Campaigns.

For each channel: spend, response, bookings and covers, margin and how sure each figure is.

Because a channel bringing guests at a deep discount may leave less behind than one bringing fewer at full price.

You often cannot know for certain. Use booking links, named offers and asking guests, and state your method.

A short note on every report saying how bookings were credited and which lines are estimated.

Monthly, with a look at each campaign as it ends.

Yes, with their fees, but note how many were regulars who would have booked anyway.

Keep them on the report with an estimated line, using named offers or asking guests.

Deciding where next month's money goes, and above all what to stop.

A marketing report puts every channel on one page with the same lines: spend, including time, the response it produced, the bookings and covers linked to it and the margin those covers left behind. It compares each channel with a baseline, separates regulars from new guests and states the attribution caveat every time, marking each line as measured, estimated or unknown.

Use booking links, named offers and asking guests to make more channels countable, and feed the report with campaign reporting and promotion reporting for the measurable ones. Report monthly, let each report change the marketing budget and use it above all to decide what to stop.