What Restaurants Can Learn from Amazon: Personalization, Reviews, and Cross-Selling

A restaurant owner in southern Germany told me a few months ago: "Michael, my server last week greeted a regular by name, kept his favorite table free, brought the Aperol Spritz without being asked — and the guest ended up booking a Christmas party for four people. Just like that."

I said: "You realize that's exactly what Amazon is trying to do with 1.5 million employees and an algorithm that drives more than a third of their USD 638 billion in annual revenue? Except your server does it better."

He looked at me like I'd lost my mind.

But that's the point. Amazon didn't build more than a third of its total revenue — over USD 200 billion of its USD 638 billion in 2024 — on a single function by accident: "Customers who bought this also bought…"

Personalization. Reviews. Cross-selling. These are not tech buzzwords. They are the three levers that turned Amazon into the most valuable company in the world. And every single one of those levers works in your restaurant — without an app, without an algorithm, without an IT budget.

I've worked with restaurant operators for 25 years. And if I've learned one thing, it's this: the best restaurant strategies don't come from restaurants. They come from industries that turned customer retention into a science.

What you'll learn in this article:

• How Amazon's "customers also bought" algorithm translates to your restaurant — with your staff as the human recommendation engine
• Why more than 9 in 10 buyers read reviews before purchasing (BrightLocal) — and what the 4.5-star threshold means for your restaurant
• 5 personalization techniques that work without technology and make guests return 56% more often
• Why cross-selling consistently outperforms upselling — and how "Would a dessert go with that?" drives 30% more revenue than "Would you like a larger portion?"
• What customer data Amazon collects about you — and what you should know about your guests (but probably don't)

"Customers also bought": Your staff as a human recommendation algorithm

When you buy a coffee machine on Amazon, the page immediately shows you matching filters, descaler, favorite beans of other buyers. You didn't search for those. But suddenly you want them.

That single function — "Customers who bought this also bought…" — generates more than a third of Amazon's USD 638 billion 2024 revenue. Over USD 200 billion. Annually.

The mechanism is simple: Amazon analyzes what millions of other buyers purchased after the same product and recommends exactly that. Not the most expensive item. Not the highest-margin one. The one that fits best with what was just bought.

What this means for your restaurant

Your service team has the same dataset — only better. A good server knows after three tables which guest takes the Barolo with their steak and which one prefers craft beer. They know the family with kids doesn't order espresso after the main but does order ice cream. They know the couple at the window table is open to a dessert recommendation when the tiramisu is framed as "fresh from the kitchen, especially good today."

That is not an algorithm. That is human judgment. And in one crucial way it beats the algorithm: your server can adjust tone, hold eye contact, and catch the right moment. Amazon can only move pixels.

But — and this is where most restaurants fail — this "human algorithm" only works when it's trained.

3 Amazon recommendation principles for your service team

Principle 1: Complement, not replace.

Amazon never shows you "buy this coffee machine instead." It shows you "this pairs perfectly with…" Your server should never say: "Instead of the schnitzel, I recommend the ribeye." They should say: "With the schnitzel, our potato-cucumber salad pairs perfectly, and many of our guests take a dry Riesling with it."

Principle 2: Social proof, not self-recommendation.

Amazon doesn't say "we recommend." It says "customers also bought." The psychological difference is enormous. "Many of our guests order the Pinot Gris with the fish" hits 3× harder than "I recommend the Pinot Gris." Because the guest thinks: if a lot of people do it, it must be right.

Principle 3: Timing is everything.

Amazon shows you the recommendation exactly when you're ready to buy — in the cart, on the product page. Not on the homepage when you're just browsing. Your server should make the dessert recommendation when the guest is finishing the last bite of the main — not when they're still studying the appetizers. And the wine recommendation comes before the guest orders, not after.

In our advisory work, one client — a 120-seat family restaurant in northern Germany — trained their service team on exactly these three principles. No major investment. Two team meetings of 45 minutes each, a laminated card with five pairing combinations per main dish, and one simple goal: recommend at least one complement at every table.

The result after 8 weeks: the average check rose 22%. Not because guests had to pay more. Because they ordered things they would have wanted anyway — if someone had simply pointed them out.

22% more check at the same food cost. At a monthly revenue around EUR 60,000, that translates to north of EUR 10k extra per month. Around EUR 150k per year. From two meetings and a laminated card.

There's a specific sentence that makes the recommendation land at 4 out of 5 tables. One sentence that separates a server who annoys from a server who advises. That sentence, and the psychology behind it, is something I cover regularly.

What you can do now: For every main dish on your menu, create a "pairs perfectly with" combination — a drink, a side, and a dessert. Print it on a laminated card for your service team. Morning briefing tomorrow: 15 minutes.

95% read reviews: Why Google reviews are your storefront

Imagine you're looking for a restaurant in an unfamiliar city. What do you do?

You Google. You check the reviews. You read 2–3 accounts. Then you decide.

That's what more than 9 in 10 consumers do — on Amazon and for your restaurant (BrightLocal). And in BrightLocal's 2017 survey, 84% trusted online reviews as strongly as a personal recommendation from a friend. That share has since softened, but reviews remain the decisive pre-visit signal.

Meaning: your Google reviews are not "nice to have." They are your storefront. Your business card. And for most potential guests, the only touchpoint before they decide whether to come to you or your competitor.

The 4.5-star threshold

Amazon listings with 4.5+ stars convert significantly better than listings with lower ratings. In hospitality the effect is even more drastic, because a restaurant visit is a higher commitment than an online purchase. You're not investing 20 dollars. You're investing an entire evening.

Below 4.5 stars on Google, you lose half your potential guests before they ever walk through your door.

Surprised? Most operators are. Because they think 4.2 stars are "pretty good." In reality, 4.2 stars are a death sentence for online visibility. The algorithm favors restaurants with higher ratings. Potential guests scroll past. And your empty Tuesday evening has nothing to do with the weather — it has to do with the three negative reviews you never responded to.

What Amazon knows about review management (that you don't)

Amazon built a whole science around reviews. The three most important findings:

1. Quantity counts almost as much as quality.

A product with 4.6 stars and 2,000 reviews sells better than one with 5.0 stars and 12 reviews. For your restaurant that means: 80 reviews at 4.6 stars beat 15 reviews at 4.9 stars. Volume creates trust.

2. Newer reviews weigh more.

Amazon weights recent reviews more heavily than older ones. Google does the same. A restaurant with 200 reviews where the most recent is 6 months old looks abandoned. A restaurant with 80 reviews where 10 are from the last month looks alive. Generating new reviews consistently is more important than having a perfect average.

3. The response to negative reviews matters more than the review itself.

Amazon sellers who respond professionally to negative reviews lose materially fewer potential buyers than those who stay silent. In restaurants the effect is stronger still. A composed, non-defensive response to a 2-star review tells every other reader: someone here cares. How you respond to negative restaurant reviews defines your image more than 50 positive reviews combined.

Across 25 years I've seen a clear correlation: restaurants that actively manage their review profile — asking for reviews systematically, responding to every one, treating negative reviews as feedback rather than attack — raise their new-guest rate by 30–45% on average within 6 months.

30–45% more new guests. Without spending a cent on advertising. Just through what others write about you.

The review math

Quick calculation?

Say your restaurant currently has 4.2 stars across 60 reviews. You start systematic review management and generate 15–20 new reviews per month (that's realistic — my clients hit this consistently). Of those, 85% are 5-star reviews (because you're deliberately approaching satisfied guests).

After 3 months: 110–120 reviews, 4.5-star average.

After 6 months: 150+ reviews, steady above 4.5.

After 12 months: 250+ reviews, stable 4.6–4.7 average.

The conversion lift kicks in the moment you cross 4.5. For a restaurant with 200 table reservations per month via Google, that typically means dozens of additional reservations. Every month. Free.

In one of our client engagements, a Swiss inn in a tourist town went this route. The owner started with 32 Google reviews. Within 2 years he had over 300 — and had climbed from TripAdvisor #46 to #1 in his area. September revenue went from CHF 28,900 to CHF 70,000. That's not luck. That's systematic review management.

What you can do now: Check your Google Business profile tonight. How many reviews do you have? What's your average? And when was the last review written? If the answer to the last question is "more than 2 weeks ago," you have a review problem — not a quality problem.

Personalization without tech: the server beats the algorithm

Amazon invests billions in personalization. Every homepage is different. Every email is tailored. Every product suggestion is based on your behavior over the last weeks, months, years.

The result: personalization drives a meaningful share of global e-commerce revenue. Customers who get personalized recommendations return 56% more often (Twilio State of Personalization 2023).

And now the good news for you: what Amazon tries to achieve with billion-dollar budgets, your restaurant can do better with zero.

Because personalization in a restaurant is not a technology problem. It's an attention problem.

The 5 personalization levels (from "nice" to "unforgettable")

Level 1: Name.

"Good evening, Mr. Weber!" instead of "Table for two?" It sounds trivial. But think about it: how often are you greeted by name at a restaurant? Almost never. That's exactly why it lands. Amazon greets you with your name on every page. Your host should do the same — and not just for the top 10 regulars. For every guest who has a reservation. The name is in the booking system. Use it.

Level 2: Favorite table.

"Your table by the window is free, as always." Amazon remembers which product categories you prefer and shows them first. Your restaurant can do the same: note the table preference, automatically reserve it next time. One entry in the reservation system. Five seconds of effort. Impact: priceless.

Level 3: Favorite drink.

"Your Pinot Gris is ready." The guest sits down, and their drink arrives before they order. That's the moment a restaurant visit becomes an experience. Amazon calls this "anticipatory shipping" — sending products before the customer has ordered. Your restaurant can do it with a notebook and an attentive server.

Level 4: Allergies and preferences.

"I saw you had the pasta without cream last time — I've already let the kitchen know." That's the moment a guest realizes: I'm not being served. I'm being understood. Amazon tracks every interaction, every click, every return. You just need a simple shorthand in the booking system: "LF" for lactose-free, "VEG" for vegetarian, "nc" for no cream.

Level 5: Occasions and life moments.

"Happy anniversary! May I bring you our house champagne?" Amazon knows when your birthday is and sends you offers. Your restaurant can do more: it can know the anniversary, the birthday, the date of the first visit — and celebrate it. Based on documented client portfolio data, birthday marketing done with system can bring 80–200 celebrations a month and EUR 25,000–60,000 of reliable revenue. Not through technology. Through attention, applied systematically.

The details — which gift, which timing, which phrasing separates 5 celebrations a month from 200 — make a 1,000% difference. Timing alone is so decisive that a shift of a few days can halve or triple the result.

The 56% return rule

Customers who receive personalized recommendations buy 56% more often. That's not an Amazon stat. It's a finding confirmed across every consumer-facing industry.

What does that mean for your restaurant?

In our advisory work, the math looks like this: if your regulars currently come twice a month, consistent personalization takes them to three. At an average check of EUR 35 and 100 regulars, that's EUR 3,500 more per month. EUR 42,000 per year. From doing what a good host ought to do anyway: take an interest in the guests.

The key isn't the single gesture. It's the system behind it. One note in the booking system. A short briefing before service: "Tonight Mr. Weber is coming — window, Pinot Gris, no cream. Mrs. Klein is celebrating her birthday — table 7 is decorated." Five minutes. And it turns an ordinary evening into something the guest tells their friends about.

What you can do now: Set up a simple system today. For every regular, one row with name, favorite table, favorite drink, allergies, special occasions. Booking system, spreadsheet, notebook — the medium doesn't matter. Consistency is everything.

Cross-selling vs. upselling: Why "dessert with that?" beats "larger portion?" by 30%

Amazon draws a sharp line between upselling and cross-selling. And that distinction explains why some restaurants systematically raise the check — and others just annoy their guests.

Upselling means: selling a more expensive version of the same product. "Looking at the iPhone 15 — how about the iPhone 15 Pro?" In a restaurant: "Would you like the larger portion?" or "Would you prefer the filet mignon over the ribeye?"

Cross-selling means: recommending a complementary product. "Customers who bought the iPhone also bought this case." In a restaurant: "With the steak, our red pairs beautifully, and for dessert I recommend the chocolate sorbet."

The difference in numbers: cross-selling raises revenue by 30% on average. And research consistently shows it lands more reliably than upselling.

Not a small edge. A different category of result.

Why cross-selling is so powerful in hospitality

The psychological reason is simple: upselling feels like "you should spend more." Cross-selling feels like "I want your experience to be even better."

When your server asks "would you like the larger portion?", the guest thinks: "Is he trying to sell me more?" When your server says "With the schnitzel, our house-made potato salad pairs perfectly — grandma's recipe," the guest thinks: "Oh, that sounds good. I'll try it."

One sentence creates resistance. The other creates appetite.

The 5 most effective cross-sell moments in a restaurant

Moment 1: Before the order — the aperitif.

"While you study the menu — may I bring you our house-made limoncello spritz? We only added it this week, and it's already a favorite." The easiest add-on: a drink the guest didn't plan but gladly takes. Average additional revenue: EUR 6–9 per table.

Moment 2: At the main — the side or the wine.

"With the salmon, our guests most often take the Pinot Blanc from the Pfalz — dry, pairs perfectly." Don't ask if the guest wants wine. Ask which wine fits the dish. That is the Amazon move: not "would you like something else?" — "this goes with that."

Moment 3: After the main — the dessert.

"Our kitchen made tiramisu fresh today — shall I bring two, or would you like to share one?" The either/or question instead of yes/no. Amazon doesn't show "would you like to buy something else?" — it shows two concrete options. Your server should do the same.

Moment 4: With dessert — the digestif.

"With the tiramisu, an espresso martini works beautifully — or a classic grappa?" Two options again. No yes/no. Average additional revenue on this single moment: EUR 8–14 per table.

Moment 5: At check — the takeaway.

"By the way, we also sell our house-made pasta sauces to take home — in case you want to relive the sugo at home." The Amazon checkout moment: "Customers who bought this also bought…" An underused revenue lever almost no restaurant touches.

The cross-sell math

Worked example: 50 tables an evening, 6 evenings a week. If your team recommends a complement worth EUR 8 at every second table — and half of those convert — that's:

25 tables × EUR 8 × 50% acceptance = EUR 100 per evening.

EUR 100 × 6 evenings × 4 weeks = EUR 2,400 per month.

EUR 2,400 × 12 months = EUR 28,800 per year.

And that's the conservative math. In our advisory work, clients who train cross-selling consistently hit check lifts of 15–35%. At EUR 80,000 monthly revenue, a 25% check lift is EUR 20,000 extra. Per month.

There's a decisive difference between a server who experiences cross-selling as "pushy add-on sales" and one who understands it as actual hospitality. That difference lives in a single question every team member should ask themselves before service.

What you can do now: From tomorrow, strike "would you like anything else?" from your team's vocabulary. Replace it with concrete recommendations: "This goes with…" 10-minute briefing before the next service. Print one recommendation per main dish: a drink, a side, a dessert.

Customer data as gold: What Amazon knows about you — and what you should know about your guests

Amazon knows about every one of its estimated 310 million customers (third-party estimates; Amazon does not officially disclose): the last 100 purchases, the average cart size, the preferred delivery day, the return rate, which categories they browsed without buying, and when they typically shop.

Your restaurant? What do you know about your guests?

If the answer is "not much," you're in good company. Most operators know their daily revenue but not their guests. They know how much they made yesterday — but not which 20 regulars account for 80% of that revenue.

The 5 data points every restaurant needs to know

Data point 1: Contact details.

Name, email, phone, birthday. That's the minimum. Without contact details, your guest exists for you only while they're sitting in the restaurant. Afterward they're gone — and you hope they'll come back. Hope is not a strategy. Email marketing is.

Data point 2: Visit frequency.

How often does the guest come? Every two weeks? Once a month? Only on occasions? Amazon tracks not just what you buy but how often. Because frequency reveals whether you're a casual buyer or a loyal customer. In a restaurant, frequency is the best early indicator of churn: if a regular who normally comes every two weeks hasn't been in for 6 weeks — you've lost them. Unless you notice and react.

Data point 3: Average check.

Not the daily average. Check per guest. Amazon knows that customer A spends an average of USD 45 per order and customer B spends USD 180. That changes what products are shown to each. For your restaurant, that means: the guest who regularly orders wine and spends EUR 75 a head gets a different recommendation than the lunch-special guest at EUR 12.

Data point 4: Preferences and allergies.

Vegetarian, lactose-free, prefers fish, only drinks red, likes to sit outside. Every one of those is a personalization opportunity. And personalization — as we've seen — drives 56% more return visits.

Data point 5: Acquisition source.

How did the guest find you? Google? Referral? Social media? A review? Amazon tracks the acquisition channel down to the last click. You don't need that kind of detail — but knowing that 60% of your new guests come from Google fundamentally changes where you put your energy.

The gold is in the combination

Individually, these data points are useful. Combined, they're powerful.

If you know Mr. Weber comes every two weeks (frequency), spends an average of EUR 85 (check), prefers red wine (preference), has his birthday on May 15 (contact), and found you through a Google review (acquisition), you can:

• Send him a personal birthday invitation on May 1
• Chill his favorite red when he reserves
• Ask him to write a Google review (he came in through a review himself — he understands the value)
• Reach out when he hasn't been in for 4 weeks (frequency deviation)

That is not surveillance. That is appreciation. And the guest feels the difference.

Based on documented client portfolio data, the pattern is consistent: restaurants that build a customer database of just 500–1,000 entries raise their revenue 20–35% within 12 months. Not through new-customer acquisition — the most expensive lever there is. But through better care of the guests they already have.

That's the third of four growth levers I work with: bringing guests back more often — the cheapest lever of the four. And data is the key to that lever.

What you can do now: Count tonight how many of your guests you know by name, how many of them you have an email address for, and how many of them you know the birthday of. Those three numbers show you exactly how far you are from a working customer database.

5 Amazon principles your restaurant can implement now

Amazon is a technology company. Your restaurant is a craft business. Yet both rest on the same principles — only the tools differ.

Here are the five principles you can roll out starting tomorrow. No app. No budget. No IT department.

Principle 1: "Customers like you" — use social proof systematically

Amazon shows you on every product: "4.7 out of 5 stars, 3,200 reviews." That's not accident. That's the most powerful selling argument in the world: others tested it and approved.

Your restaurant can use social proof in four places:

• Menu: "Our most popular dish" or "Favorite of our regulars" next to a dish reads like a 5-star review
• Entrance: A small board with the current Google rating and review count
• Service: "Most of our guests take the Primitivo with the lamb" — social proof in conversation
• Online: Embed the best Google reviews on your website — real names, real words

Principle 2: "Recently viewed" — repeat the familiar

Amazon shows you what you recently viewed. Why? Because familiarity creates trust. People buy what they know.

In a restaurant: "Mr. Weber, last time you had the risotto — it's on the menu again tonight, and the chef has refined it slightly." The guest feels remembered. Feels important. And the probability they order is 3× higher than for a dish they don't know.

Principle 3: "Bought together" — build combo offers

Amazon bundles products into packages: "Frequently bought together: camera + memory card + bag. All three for USD 449 instead of USD 510."

In a restaurant this works as a tasting menu: appetizer + main + dessert + paired wine at a package price. Not as a discount — as a curated experience. "Our connoisseur menu: 4 courses with wine pairing, composed by our chef." The guest spends more than on à la carte — and feels cared for rather than squeezed.

Principle 4: "Prime members" — create VIP status

Amazon Prime costs USD 139 a year. For that, you get faster delivery, exclusive offers, and the feeling of being treated better than non-members. More than 220 million people worldwide pay for it.

Your restaurant can use the same principle — without a fee. A "regulars' club" or VIP program giving loyal guests exclusive benefits: priority reservations, invitations to cooking events, early access to seasonal menus, a small amuse-bouche on arrival.

Cost: minimal. Effect: enormous. Because the guest isn't just eating — they belong to a group that deserves special treatment.

In our advisory work, one Swiss client built a VIP club with 360 members. No membership fee. Just an invitation after the third visit: "You're now one of our valued regulars." VIP members visit 3.2× more often than non-members and spend 28% more per visit.

Principle 5: "Leave a review" — actively ask

Amazon asks across multiple touchpoints: "How was your experience? Rate now." Via email, app notification, reminder.

Most operators never ask. Because they think it would be pushy. It isn't. The opposite: it tells the guest that their opinion matters to you.

Timing is decisive: not at checkout (too rushed). One or two hours later. A short message: "Thank you for your visit tonight. If you enjoyed it, we'd appreciate a quick Google review. Here's the direct link: [link]."

Clients who run this system consistently generate 15–30 new reviews per month. Within 12 months it reshapes their entire online visibility and drives a measurable rise in new-guest reservations via Google.

What you can do now: Pick ONE of these five principles — the one where you see the biggest gap. Implement it this week. Not all five at once. One principle, done consistently, beats five principles started half-heartedly.

FAQ

Isn't personalization in restaurants old-fashioned — doesn't a good host already know their guests?

Exactly the point. The "good host who knows everyone" isn't obsolete — it's the original that Amazon spends billions trying to copy. The difference: the good host has the knowledge in their head. If they're sick, on vacation, or leave, it's all gone. Systematic personalization means the knowledge lives in the booking system, not in one person's head. Any team member can access it. It's not old-fashioned — it's the professional version of what good hosts have done for centuries.

What if I barely have any online reviews — where do I start?

With the satisfied regulars you already have. Talk to them at their next visit: "Mr. Weber, you've been coming to us for 3 years — your opinion on Google would help us a lot." Most say yes immediately. They just didn't know you wanted them to. Follow the conversation with a short message containing the direct Google link. Ask 10 regulars, 7 write a review. That's a realistic start — and after 4–6 weeks you'll have 20–30 new reviews.

Cross-selling — won't my guests find it pushy?

Only if it's done wrong. "Would you like anything else?" is pushy. "With the lamb, our Primitivo works beautifully — shall I bring a glass?" is hospitality. The difference sits in the phrasing: complement, not add-on. Recommendation, not question. And in the timing: the right moment matters. When the guest is still puzzling over the menu, don't interrupt. When they're taking the last bite of the main and look content — that's the moment for the dessert recommendation.

Do I need a CRM system or is a notebook enough?

For the beginning, a notebook is enough. Or a spreadsheet. Or an entry in the booking system. Amazon started with a database in a garage. You don't need Salesforce. You need a system you'll actually use. 100 guests in a spreadsheet with name, email, birthday, and preferences are worth more than a CRM with 5,000 contacts nobody maintains. Start small, but start.

Does this work for a small 30–50 seat restaurant?

Especially for small restaurants. 30 seats means 30 guests per evening, each of whom you can personally greet. That's an advantage, not a disadvantage. In a 200-seat operation, personalization is heavy lifting. In a 30-seat operation it's natural. You already know your guests — the system just makes sure the knowledge doesn't disappear when you're not there.

What does this actually get me in money — is it worth the effort?

Best-case-scenario math with four growth levers all firing: if you win 15% more new guests through better reviews (lever 1), lift check by 20% through cross-selling (lever 2), raise visit frequency by 15% through personalization (lever 3), and reduce churn by 10% through better data (lever 4), you get: 1.15 × 1.20 × 1.15 × 1.10 = 1.75. That is 75% more total revenue. At current monthly revenue of EUR 50,000, that's EUR 37,500 more — per month. At EUR 80,000 monthly, it's EUR 60,000 more. In practice, not every lever fires at once — and at half the result you're still looking at EUR 19,000–30,000 extra per month.

My staff turns over constantly — how do I keep the system running?

That's exactly why you need a system instead of gut feel. If the knowledge about guests sits in your best server's head and they quit, you're back at zero. If it's in the booking system, the new hire can say "Good evening, Mr. Weber, your window table is ready" on day one. The system outlasts turnover. Gut feel doesn't.

Amazon has billion-dollar budgets — can I really do this with zero?

The principles cost nothing. Remember a name: 0. Note a favorite table: 0. Make a pairing recommendation: 0. Ask for a review: 0. The only investment is time — and not much. 15 minutes briefing before service. 5 minutes of notes after. 10 minutes answering emails. 30 minutes a day total, for a 30–75% revenue lift. Show me another investment with that ROI.

Bottom line: Amazon thinks in data. Your restaurant can think in people.

Amazon perfected three things: recommendations, reviews, and cross-selling. Together, those three levers account for the majority of Amazon's USD 638 billion in 2024 revenue. Annually.

Your restaurant can use the same levers. Better. Because you have something Amazon never will: human contact.

An algorithm can analyze what millions of customers bought. But it can't see that the guest at the next table looks tired and would prefer something light tonight. It can't hear that the group at the round table is celebrating a birthday and would light up with a glass of sparkling. And it can't sense that Mrs. Klein hasn't been in for three weeks and would appreciate a "we miss you" message.

The 5 lessons:

• Recommendation engine — Train your team to recommend complements, not upgrades. "This goes with…" is the most powerful sentence in hospitality.
• Reviews as storefront — 4.5+ stars, new reviews regularly, professional responses to criticism. Your Google profile is your most important marketing asset.
• Personalization without tech — Name, table, drink, allergy, birthday. Five data points that turn a guest into a regular.
• Cross-selling over upselling — 30% more revenue through complements, not larger portions. Consistently more effective.
• Customer data as gold — A database of 500 entries is worth more than 500 new flyers. Because existing guests are 6–7× cheaper to activate than new ones.

Amazon has 1.5 million employees, over USD 200 billion in recommendation-driven revenue, and the most powerful algorithm in the world. And even then, Jeff Bezos would envy what an attentive operator in a 30-seat restaurant can do: treat a guest so well they feel like the only customer in the world.

You don't need an algorithm. You need a system. And a server who loves their work.

Start today. Not with a tech project. With one question to your next guest: "How can I make your evening perfect?"

Related reading

• More reviews for your restaurant: systems that work
• Email marketing for restaurants: the numbers
• Birthday marketing: 80–200+ celebrations per month
• Negative reviews: how to respond
• What restaurants can learn from Starbucks
• The full series: What restaurants can learn from other industries