What a regular
is worth.

Search 'how much should a restaurant spend on marketing' and you'll get a neat percentage of revenue, usually from overseas sources. I couldn't find an Australian survey that measures it, and the ATO's small business benchmarks for restaurants, coffee shops and pubs, taverns and bars don't include an advertising line at all.

That doesn't mean guess. It means work from numbers you do have: turnover, food and drink costs, what a regular spends and how often they come. A percentage someone else picked can't tell you whether $600 of ads last month paid for itself. Your covers can.

From the ATO's small business benchmarks, which use 2023-24 tax returns. Cost of sales is mainly food and drink. Total expenses is everything the business claims, including wages and rent.

Read the last column the other way round. A restaurant in Newcastle turning over $1.2 million averages 88% in expenses, which leaves about 12 cents in the dollar. For a sole trader or partnership that slice usually has to cover the owner's own pay too, because owners in those structures don't pay themselves a wage. Marketing comes out of the 88% or out of the 12. Either way it has to earn its place.

A regular's value isn't what they spend. It's what's left after the food and drink they eat. Hey You's Daily Brew Report, built with Westpac data, put the average café transaction at $17.70 in June 2025. A regular who spends that twice a week for a year spends about $1,840. At the ATO's 38% average cost of sales for most cafés, about $1,140 of that is gross profit, before wages and rent.

How often people come matters as much as what they spend. Lightspeed's 2024 survey of more than 1,000 Australian diners found the average person eats out around 38 times a year and goes out for a drink 23 times. Your regulars are the ones well above that. In the US, Toast's data suggests up to 50% of a restaurant's order volume can come from just 7% of guests. No Australian equivalent has been published, but most owners can name their top twenty customers without looking.

The cheapest marketing a venue does costs time, not money: a current Google profile, a menu written as text, replies to reviews. Do those before paying for anything. The Google profile page covers what to fix.

Then count the costs you already pay without calling them marketing. A delivery app's commission is an acquisition cost: Uber Eats publishes 30% when it delivers. Paying that once to win a new customer can be fine. Paying it every week for a regular who'd happily order direct isn't, which is what the direct ordering page is about.

• Free first: Google profile, menu, hours, review replies, photos.
• Then the leaks: app commission on regulars, discounts on nights that fill anyway.
• Then paid: ads for the quiet services, with a cap and a way to count bookings.

Flip the question. Instead of 'what percentage?', ask 'how many new regulars do I need this month, and what can I pay for each?' If a regular is worth $1,140 a year in gross profit and you want your money back inside a year, you could pay up to that to win one. Wages and rent still need covering, so aim well under it.

Then test it on one service. Spend a capped amount for four weeks on a quiet Tuesday, count covers against the four weeks before, and ask new faces how they found you. If ten extra covers a week turn into two or three people who keep coming back, the maths works. If not, stop and try the next idea. The 30-day plan lays that test out week by week.