Data Insights

Why Restaurant Sales Are Dropping: Find the Cause, Don't Guess

Updated on 10 min read

A restaurant sales decline almost never has a single cause. It sits in one location, one weekday, one time of day or one product group – rarely everywhere at once. If you break it down systematically along these four levels, you find the driver in hours. If you guess from your gut, you search for weeks and often treat the wrong symptom. To keep the comparison between locations fair, you need consistently calculated KPIs with clear formulas.

Watch the gap between nominal and real sales: on your bank statement, flat or slightly rising sales can look like growth while the purchasing power behind them shrinks. AI can't make this diagnosis for you. What it can do: automate the data work behind it. Bring the numbers together, break them down by level, and name anomalies before you ask.

This article shows the four levels on which you break down a sales decline, and where AI-assisted controlling takes over that work.

Contents


Check the market first: is it you or everyone?

Before you dig into your own numbers, put your decline in context. Published hospitality statistics for your market tell you whether the whole sector is under pressure — and whether that pressure is real (adjusted for price increases) or only nominal.

That context explains part of a decline at most. If your business is roughly in line with the market trend, that's bad, but not unusual. If you're far below it, the market has nothing to do with it. Then the problem sits in your business, and the industry figure won't tell you where.

Why looking at total sales misleads you

The most common mistake I see in multi-location businesses: someone reports "sales minus 8 per cent" as a fact, when it is really a summary of at least four different stories. One location can be booming while two others collapse. The weekend can hold steady while weekday lunch trade falls away. Total sales hide all of this, because they add everything up into one number.

That one number answers the wrong question. The right question isn't "How much less?", but "Where exactly, and why there?". Only when you break the decline down into its components does a hunch become a verifiable cause. That is also what the location comparison in sales analysis does: outliers only show up in comparison, never in the average.

The four levels, in the order that makes sense: location, time, product group, external factors. Each level narrows the problem further.

Step 1: Are sales falling at one location or everywhere?

The first question is the cheapest to answer and delivers the biggest insight: does the decline affect all locations equally, or is it concentrated?

Take a fictional model business with €1.5 million annual sales across three locations – purely for illustration, not real operating figures:

Location Sales vs. same month last year Assessment
Location A (city centre) −2% small decline
Location B (suburb) −4% moderate decline
Location C (near the station) −22% clear outlier

The total for this model business would be around minus 9 per cent, a number that sounds "bad, but nothing unusual" and is dangerous precisely for that reason. It hides that location C has a problem of its own that the other two don't have. Perhaps new roadworks outside the door, a new competitor, a change in kitchen staff, worse Google reviews. Without the location comparison, you blame 9 per cent on the market – and miss that two thirds of it hangs on a single location.

If the decline is spread evenly across all locations, the cause is more likely outside your business: market trend, regional purchasing power, a seasonal effect. If it's concentrated in one location, the cause is there – and usually fixable.

Step 2: Which weekday or time of day carries the decline?

Once the affected location is identified, the time axis follows. A monthly decline of 10 per cent can come from two completely different patterns: spread evenly across all days, or concentrated entirely on two or three days.

Model calculation for the same fictional location C, compared with the same week last year:

Day Sales vs. last year Pattern
Monday–Thursday, lunch −31% home-working and canteen effect
Monday–Thursday, evening −6% slightly declining
Friday/Saturday −3% stable
Sunday +2% stable to slightly positive

In this model, almost the entire decline comes from weekday lunch trade. Weekend business is practically unchanged. If you only see monthly sales, you'd reflexively question the menu, prices or service – areas that apply equally at the weekend and work there. The real cause probably lies in the lunch business itself: fewer offices nearby, more home working, a new lunch spot next door. Those are different levers from "rework the menu".

The time axis doesn't just show when sales are missing; it also narrows down the guest segment that's staying away. Weekday lunch guests are mostly people working nearby – a different segment from weekend guests who come on purpose.

Step 3: Which product group is losing, and is it volume or price?

Now comes the question most often overlooked: is the location losing guests, or are the remaining guests spending less? That isn't an academic distinction. Each requires a completely different response.

Break the sales decline down by product group, again for the fictional model business:

Product group Volume (number of items sold) Average price Sales effect
Food −4% +2% (price increase in spring) almost neutral
Soft drinks −6% unchanged slightly negative
Alcoholic drinks −34% unchanged strongly negative
Desserts −12% unchanged negative

In this model, the "alcoholic drinks" group carries almost the whole decline – with stable food volume. That's a completely different diagnosis from "guests are staying away". It points rather to changed guest behaviour (less alcohol, more drivers, health trend), a cut cocktail list or a staffing problem at the bar. The response isn't "more marketing", but: review the bar concept, upgrade alcohol-free alternatives, train bar staff.

If you only saw "sales minus X per cent", you'd probably tinker with the food offer – the one area that isn't the problem in this model.

Step 4: Are weather or surroundings playing a part?

Only once location, time and product group are narrowed down is it worth looking at external factors. Pulled in too early, the weather becomes a convenient excuse for a structural problem.

Checklist for external factors:

  • Weather. For locations with a terrace or outdoor area, rain and cold noticeably affect outdoor guest numbers – a wet spring explains part of a decline, but rarely a lasting one.
  • Roadworks and construction. Closed roads or footpaths outside the door reduce footfall, regardless of your own offer.
  • Public holiday timing. A public holiday that fell on a Saturday last year and on a Monday this year shifts the comparison base without anything changing in the business.
  • Local events. A concert, trade fair or street festival nearby can heavily distort a single day – in either direction.
  • New competitors. A new opening within walking distance often explains exactly the decline that is concentrated in one location and one guest segment.
  • Local visibility. Outdated opening hours, few reviews or missing photos on your Google profile cost walk-in trade. How to check your profile is covered in the guide to optimising your Google Business Profile.

The point of this level isn't to push every decline onto external causes. The point is to rule them out or confirm them systematically – before you launch an expensive internal measure that can't solve an external problem.

How AI-assisted controlling finds the cause automatically

Doing these four steps by hand is feasible for one location in one bad month. For ten locations over twelve months, Excel quickly gets messy – and that is where the analysis usually stops in practice. Not because the method is wrong, but because nobody has time to rebuild it every month for every location.

Kiosa automates exactly this breakdown. The AI morning briefing doesn't wait for you to ask why sales fell. It reports the deviation and the most likely explanation of its own accord, as soon as the previous day's POS data has come in. Anomaly detection compares every location, every weekday and every product group against its own historical pattern, not against a flat target. One View shows your daily P&L; with the Invoices module, costs from your supplier documents flow in automatically, instead of waiting three to six weeks for the monthly accounts from your accountant – so a sales decline can be read directly alongside the food cost ratio and labor cost for the same period, not in isolation.

Important: the software delivers hypotheses and quantified drivers, not certainty. Whether the new roadworks outside location C are really the reason or just play a part is your call – the AI shows you where to look first, not the finished answer.

Kiosa pricing is modular, per location per month, with no setup fee:

Module Price/location/month
One View (dashboard, morning briefing, reviews) €99
+ Inventory €49
+ Invoices (documents, approval, accounting export, costs flow into the P&L automatically) €99
+ Loyalty €99
+ Scheduling €99
Full stack €445

To get started, One View is enough to see location and time comparisons automatically. The Inventory module adds the product group level from step 3. The sales forecast, built on the same data, is accurate enough to make staff and order planning proactive rather than reactive.

From analysis to action

What this is about: a diagnostic process for finding the cause when sales are falling.

Put it into practice:

More on KPIs, sales and forecasting

FAQ

What are the most common causes of a restaurant sales decline?

There is no single main cause – that's the core of the problem. Common drivers are falling guest numbers due to competition or changed surroundings (roadworks, office locations), a lower average spend due to changed ordering behaviour (less alcohol, for example), seasonal and weather-related swings, and general consumer restraint across the sector. Which of these applies to you only becomes clear when you break sales down by location, time and product group.

How do I tell whether the decline is down to one location or the whole business?

Compare each location's sales trend individually against the same period last year, not just total sales. If all locations are in a similar range, the cause is more likely external (market, season). If one location clearly stands out, the problem is there – and it's usually also the easiest to fix.

How can I tell whether fewer guests are coming or guests are spending less?

Break sales down into covers and average spend. If covers fall while average spend stays stable, it's a footfall problem – marketing, location, competition. If covers stay stable while average spend falls, it's down to ordering behaviour or the offer itself, such as a cancelled promotion or fewer add-ons sold.

How often should I check sales for deviations?

Daily for the deviation itself, weekly for the cause analysis. A single bad day says little. Only a pattern over two to three weeks, such as a particular weekday consistently below plan, is the point where a deeper analysis pays off.

Can software find the cause of a sales decline automatically?

Software can deliver the breakdown by location, weekday, time of day and product group automatically and quantify the most likely drivers – the data-heavy work that otherwise takes hours. The final assessment, such as whether an external cause like roadworks is really responsible, remains the operator's job. The software delivers the hypothesis and the data behind it, not the final decision.

Your next steps

Are you guessing right now why your sales are falling, or have you already broken them down by location, weekday and product group? To get from hunch to diagnosis:

Book a demo: In 30 minutes we'll show you how the morning briefing explains deviations at your own locations, instead of just displaying them – choose a time.

Related articles: Restaurant sales analysis: why daily sales lie · Restaurant financial controlling: what to measure daily, weekly and monthly

Sven Schneider

About the author

Sven Schneider — Founder of Kiosa and Ragazzi Concepts. 15 years in multi-unit restaurants; helped grow Mama Trattoria from 0 to 10 locations in Hamburg.

Kiosa Dashboard

All your numbers. One view. Instantly.

What you just read, Kiosa does automatically: every day, for every location, without you having to piece anything together.

Book a demo call
All articles

Data: Kiosa Gastro-Index (in German): 269,448 venues across Germany, Austria and Switzerland, measured.