How a new restaurant gets found in the first 90 days

Opening is the one period where an owner has no history, no reviews, and no patience. It is also the period where the decisions are cheapest to get right, because there is nothing to undo yet.

The mistake I watched most often was sequence. People spent on advertising in week one and claimed the profile in week six, which is exactly backwards.

Do nine out of ten restaurants fail in the first year?

No. That number is folklore, and it was disproved in a peer-reviewed journal twenty years ago.

H.G. Parsa and colleagues published "Why Restaurants Fail" in the Cornell Hotel and Restaurant Administration Quarterly in August 2005. They found the 90% figure had become widely accepted with no data supporting it. What the data actually showed:

• First-year failure rate for independent restaurants: 26.16%.
• Second year: 19%. Third year: 14%.
• Over a longer horizon, failure rates were roughly comparable between franchise chains (57.2%) and independents (61.4%). Being a chain does not save you the way the folklore implies.

The arithmetic disproof is the part worth carrying. If 90% of restaurants failed in year one, a market of 1,500 units would shrink to 254 units in twenty years, a loss of 84% of existing restaurants. That is the opposite of the observed industry, which has grown continuously. The National Restaurant Association put industry sales at $1.5 trillion in 2025 with employment reaching 15.9 million, the second largest private-sector employer in the country.

About one in four in year one is still a serious operating fact. It is just a different one, and it should change how you spend in month one. You are not sprinting away from a 90% cliff. You are building a business with a real but ordinary risk profile, and panic spending is a bad response to an ordinary risk.

What should be done before the doors open?

Everything that takes time to verify.

Claim the Business Profile. Verification is not instant and it is not always smooth. Start it well before your soft opening so you are not waiting on a postcard while your first guests are trying to find you.

Use your real name, exactly as it appears on the sign. Not your name plus your cuisine plus your city. Keyword stuffing in the business name is against policy, and it invites the kind of edit war that costs you the listing later.

Pick the most specific primary category that describes what you sell. Whitespark's analysis of 1.8 million Google Business Profiles across 4,209 categories, published in August 2026, found the exact match category outperforms adjacent categories across almost the entire dataset. That is a four-minute decision with a permanent effect. Whitespark sells local SEO software, so label it as vendor research, but the finding matches every edition of its ranking factors survey, where primary category has finished first every year.

Set hours you can actually keep. Soft opening hours are different from full service hours. Use special hours for the transition rather than publishing a schedule you will break in week two.

Buy the domain and stand up a real page. Address, phone, hours, menu as text, and a way to reserve or order. It can be one page. It cannot be a splash screen that says coming soon with a form on it.

The full field by field version of the profile work is in the complete Google Business Profile setup for restaurants. Do that once, properly, before you are busy.

What actually matters in the first thirty days?

Reviews and accuracy. In that order, and almost nothing else.

You open with zero reviews against competitors who have hundreds. You cannot close that gap in a month and you should not try. What you can do is start a pace and never stop, because pace is what a new profile has going for it.

Whitespark's 2026 Local Search Ranking Factors report, in which 47 local search experts scored 187 factors, ranks recency of reviews eleventh at 164 and a sustained influx of reviews over time, rather than bursts, fourteenth at 154. Quantity of native Google reviews with text ranks ninth at 170.

The benchmark is straightforward. Darren Shaw of Whitespark puts it this way: "How often are your top competitors getting new reviews? If the answer is twice a month, you should also aim for that, plus one." Go look at the profiles above you, count their recent reviews, and set your target from that.

Two of his other findings matter more to a new room than to anyone else. A negative review is better than no new reviews at all, because recency raises rankings regardless of sentiment. And a business that asks every customer should expect roughly thirty positive reviews per negative one, which is the number to hold in your head the first time a two-star lands in week three.

Ask every guest, every service, with a system rather than a mood.

Do not route happy guests to Google and unhappy guests to a private form. Google's Maps content policy bans it in a single clause: merchants may not "discourage or prohibit negative reviews, or selectively solicit positive reviews from customers." Google also bans incentives outright, and bans asking staff to hit review quotas or asking reviewers to name a staff member. On the federal side, the FTC's rule on consumer reviews carves out a safe harbour for generalized solicitations to purchasers, which is exactly the ask-everyone pattern. Ask everyone, condition on nothing.

Google's own guidance names the mechanism it wants you to use: print and display the review QR code generated from your profile. Note the trap you will hit on opening week, since that QR code can only be generated in a computer browser, not on your phone.

Accuracy is the other half. Hours, phone, address, and menu should match everywhere, and in the first month they will drift constantly because the business itself is changing. Check them weekly during the opening period, not monthly.

Photos go up from day one. Exterior in daylight, the room during service, six clean dish shots. Add more as you go rather than all at once.

What changes between day thirty and day ninety?

You stop building and start reading.

By day thirty you have some data. Which searches bring people. What they do when they land. What the phone is actually asking for. That last one is the most valuable and the least captured, because the answers are sitting in the head of whoever answers the phone.

Maple's analysis of 1.2 million calls across more than 1,000 US restaurant and local business locations found 58% of reservations still originate from a phone call, and 68% of all calls arrive during the lunch and dinner rushes. Maple sells restaurant phone software, so treat it as vendor research with a disclosed sample. For a new room with one host and no habits yet, that is the single most predictable failure point of the opening.

Pay attention to the questions you did not plan for. If eight people a week ask about parties of ten and your website says nothing about groups, that is a page you should have. If callers keep asking about drop off orders for offices, you are being told there is a business you are not yet equipped to take, which is exactly what a missing catering page costs a restaurant.

This is also when you fix the site properly. The first version was built in a hurry and it shows. A PDF menu, a homepage that is one video, a contact form where a phone number should be. Those are website mistakes that cost covers and they cost more once you have traffic to lose.

Do not change your name, your concept, or your domain during this window. Whatever ranking you have built is fragile, and a rebrand this early throws it away. If you truly must, do it in the correct order, which is laid out in handling a name change without losing rankings.

What can a marketing retainer cost you in covers?

Run this before you sign anything during an opening.

Say an agency quotes $2,000 a month. Your check average is $38. Use your own numbers.

The National Restaurant Association's cost picture for a typical independent puts food at roughly 33 cents of every sales dollar, so each incremental cover contributes about $25 before any labor.

• $2,000 divided by $25 is 80 incremental covers a month just to cover the fee, and that ignores the extra labor those covers need.
• Across a 30-day month, that is fewer than 3 covers a night.

Then decide honestly whether the work being proposed produces three extra tables a night in month two. Profile completion and a review system plausibly do. A content retainer, a social calendar, or an AI visibility package almost certainly does not, on that timeline.

The point of the arithmetic is not that agencies are expensive. It is that the hurdle is small enough to state, so you can hold the proposal to it. Ask the person selling to you what the covers-per-night number is for their scope. If they cannot answer, that is the answer.

What can I safely ignore for now?

More than you would expect.

Paid ads before the profile is complete. You are paying to send people to a listing that cannot answer their questions.

Any agency pitch involving a special file, special markup, or an AI package. Google states plainly that none of that is required to appear anywhere, including in AI answers, and that no third party tool has access to its internal ranking or AI systems.

Rank tracking as a daily habit. A new listing is volatile and watching it hourly will only make you anxious and reactive.

Blogging. A new restaurant does not need articles. It needs an accurate profile, real photos, and reviews arriving.

Loyalty programs, apps, and email automation. All fine later. None of them help a stranger find you in month two.

What to do this week

If you have not opened yet, start profile verification today. It is the only item with a waiting period.

Write your hours down for the next ninety days including every day you know you will close, and enter them as special hours.

Count the recent reviews on the three profiles ranking above you, set your monthly target at their rate plus one, and write it on the wall of the office.

Put a review ask into the closing sequence of every table, with the same wording every time, and make one manager responsible for whether it happens.

Take the exterior photo and six dish photos this week. Upload them.

Then check the two or three searches you most want to win, logged out, from a phone off your network, and write down where you actually land. That is your starting line, and you will want it in three months.

Questions about your opening?

Email me at eric@seod.com with your opening date, your concept, and your city. I will send back a one page sequence for your ninety days, in order, with the two items that have a waiting period marked so you start those first. No call required.

I ran restaurants for sixteen years and I answer these myself. If your opening plan is already sound, I will tell you what to cut rather than what to add.

Or keep reading more on restaurant marketing.