Marketing metrics that matter: from reach to the register, before vs after

When the Instagram dashboard stops being enough?

Your platform dashboard stops working for you the day your cash register sits flat while your numbers climb, and that day arrives sooner than you expect.

The Peruvian restaurant in Bogotá closed March with 412,000 video plays, 1,900 new followers and flat sales: the giveaway is not the reach, it is the gap between that reach and the ticket. Meta hands you impressions, saves and video retention because those are the metrics it produces itself; it does not know, and never will, whether your guest came back the following Thursday, because a restaurant converts offline, at a table, with no trace of the campaign that brought it. Your POS does know. With a marketing budget that can reach 10% of sales in a new restaurant according to Toast, measuring the wrong variable for six months costs real money, not opportunity.

Alternative 1: cost to acquire one new guest

Divide every marketing dollar you spent that month —paid media, production, the photographer, the fee of whoever runs your social accounts— by the number of guests who walked in for the first time, then hold that figure against the contribution margin of an average ticket: if it comes out higher, you are buying customers at a loss. It fits the owner of one to three locations who already has a POS and has no analyst, because the math lives on a single sheet and takes twenty minutes at month end. The switching cost is close to zero in money and steep in discipline: somebody has to flag the new guest at the point of sale, every day, without exceptions. With an optimal food cost between 28% and 35% (National Restaurant Association), the room you have left to pay for acquisition is narrow, and that calculation done once has killed more bad campaigns than any content audit.

Alternative 2: 90-day return frequency

How many of January's new guests ate with you again before April? That question, answered with a number, beats a year of reach reports. An urban daily-menu restaurant in good health runs at 35-45%; a special-occasion place, at 12-18%. Below those ranges the problem does not live in marketing but at the table —the dish, the wait, the server— and pouring more into paid media is water into a punctured bucket. The profile that gains most from this metric is the daily-menu or neighborhood operator, whose business rests on repetition rather than novelty. Third-party data backs the bet: existing customers spend on average 67% more per order than new ones, according to Restroworks, and 47% of adults order takeout every week (National Restaurant Association 2025), so the returning guest has plenty of occasions to come back. Guest lifetime value comes out of three figures you already own: average ticket, visits per year, and estimated years of loyalty.

Alternative 3: guest lifetime value, the number that sets the budget

A guest of 22 dollars who comes ten times a year for three years is worth 660 dollars in sales, and it is that number —not the single ticket— that tells you how much you can pay to bring them in. I got this wrong for years: I recommended fifteen-indicator dashboards to single-location operators who had nobody to fill them, the dashboard died within three weeks, and the owner concluded that measuring was useless. It was not; my design was. FOUR well-filled cells beat fifteen half-empty ones. Loyalty programs are the cheap route to lifting that value: members visit more than 40% more often than non-members (Paytronix Loyalty Trends Report 2024) and spend 38% more per visit than a walk-in customer (Paytronix 2025). Set the sales of your campaign window against the same window without a campaign a year earlier, adjust for holidays and price changes, and keep the difference: those are incremental sales, and they are the only thing that justifies paid media.

Alternative 4: incremental sales during the campaign window

A location billing 400,000 dollars a year and spending 24,000 on marketing needs at least 80,000 incremental dollars for the investment to breathe on a 30% contribution margin. This metric is built for the owner already negotiating with an agency or an influencer who needs a referee; without it, the conversation goes to whoever talks best. US brand spending on influencer marketing hit 10.52 billion dollars in 2025, up 23.7% (Socially Powerful), and a good share of that money moves with no counterfactual behind it. Before you touch the media budget, look at the business's Google listing, because in 2026 restaurant discovery is decided there and inside AI-assistant answers, which recommend three places instead of ten. Listings with more than 100 photos get 2,717% more direction requests (The Media Captain, Google Business Profile Stats 2025), and that work costs one afternoon with a photographer, not a monthly campaign.

The alternative almost nobody measures: your map listing

Diego F. Parra keeps pressing the same order at Masterestaurant: listing, reviews and a visible menu first; paid media after. If your restaurant is missing from the shortlist the assistant dictates to a diner, no volume of Reels rescues it. According to Sheryl Kimes, professor emerita at Cornell's School of Hotel Administration, the right unit in this business is revenue per available seat hour, because you sell time and space, not only plates. Changing metrics does not cost software, it costs routine, and that is why most owners never do it. You need three things: a POS that tags new versus returning guests, somebody who exports that report on the first day of each month, and four figures living on one sheet reviewed in the same meeting where you review purchasing. Extra tooling usually lands under 100 dollars a month in a mid-size location; the attention bill is the one that stings.

What switching dashboards really costs?

And the risk of skipping it is concrete: cash flow is the leading cause of financial stress and closure among small businesses, according to Inc., and a restaurant buying customers above its margin burns cash while celebrating its reach.

One month of delay in that diagnosis gets paid out of payroll. If you opened less than six months ago, stay where you are: you do not yet have a guest base large enough for 90-day return frequency to mean anything, and lifetime value built on four months of history is a statistical illusion that will push you into bad decisions. The same holds for the seasonal operator who sells at a festival or on the coast three months a year: that cycle does not accommodate 90-day windows. In those two cases the useful pair is acquisition cost and incremental sales alone, with the rest parked until the first full year closes.

When NOT to change anything?

There is a third honest case: if your restaurant is full Tuesday through Sunday and billing at the top of its capacity, better measurement will not lift your cash —a price change or an extra turn will.

Start tomorrow: export last month's new guests from the POS and divide your marketing spend by them. ACQUISITION COST PER NEW GUEST. Divide everything you spent on marketing during the month —ads, production, photographer, the fee of whoever runs your social— by the number of guests who walked in for the first time. If the result exceeds the contribution margin of an average check, you are buying customers at a loss. That single calculation has killed more bad campaigns than any content analysis. 90-DAY RETURN FREQUENCY. Of January's new guests, how many ate again before April. An urban daily-menu restaurant lands around 35-45%; a special-occasion venue, 12-18%.

The four metrics that actually move the register

Below those ranges the problem sits at the table rather than in marketing, and pouring more into ads fills a punctured bucket. GUEST LTV. Average check times annual frequency times contribution margin. A guest who spends USD 22, comes seven times a year and leaves 63% gross margin is worth USD 97 annually. With that figure in hand, paying USD 12 to acquire them stops looking expensive. Without it, every acquisition cost looks like an excess. INCREMENTAL SALES FOR THE WINDOW. Not total sales during the campaign month, but the difference against the same period last year, adjusted for seasonality and price changes. It is the most uncomfortable metric because it usually deflates the agency's story, and that is precisely why you calculate it. WHAT IS NOT A BUSINESS METRIC, however proudly the vendor puts it on slide one: reach, impressions, new followers, saves, likes and —yes— engagement rate too.

The four metrics that actually move the register — in practice

They measure content health. None answers the only question that matters on day 30: did more money come in than went out? ONLINE REPUTATION AND DELIVERY CONVERSION enter the dashboard as business metrics rather than communication ones: new reviews per month, response speed, and on delivery the margin after commission, which on aggregators runs between 15% and 30% of the check.