Restaurant SEO in India: how to win Google Maps when the aggregator owns the search

Restaurant search has a property that most local verticals do not: an enormous share of it is immediate. Someone is hungry, they are nearby, they are deciding in the next four minutes. Google knows this, which is why hours are not a piece of profile housekeeping for a restaurant — they are a gate.

When a customer filters to open now, or simply searches at nine at night, a restaurant whose hours say it closed at eight is not ranked lower. It is removed. This is the single most common way Indian restaurants lose search traffic they had already earned, and it happens invisibly, because the owner is standing in a full room at nine and has no way of knowing about the twelve people who never saw them.

The problems compound in this trade specifically. Kitchens close before the restaurant does. Lunch service ends and dinner service starts, and a profile with a single continuous window tells a customer at four in the afternoon that you are serving when the kitchen is cold. Sunday is different. The bar runs later than the kitchen. Almost none of that is reflected in the average Indian restaurant profile, which carries one set of timings entered on the day the listing was created.

Then there are the exceptions, which matter more here than anywhere else. Holi, Diwali, a local bandh, a staff wedding, the day the municipality cuts the water. Special hours exist as a field precisely for this and are almost never used, so customers arrive at a closed shutter, and a portion of them leave a one-star review about it — which costs you far more than the closure did.

Fix this once with a structure that matches operations rather than aspiration: separate lunch and dinner windows if that is how you actually run, a different Sunday, and a standing habit of entering special hours the moment a closure is decided rather than the morning it happens.

It is worth understanding why this field carries so much more weight for restaurants than for other trades. Google is not merely displaying your timings, it is using them to decide whether you are a valid answer to a query being asked right now. A hardware shop that is shut at nine at night is still a reasonable result for somebody planning tomorrow. A restaurant that is shut at nine is not an answer to a person who wants to eat at nine, so Google removes it rather than ranking it low. That is why hours accuracy behaves like an on-off switch here and like a quality signal everywhere else, and why it is the first thing to fix rather than the last.

Every local business benefits from fresh photographs. For restaurants the effect is sharper and the decay is faster, for a reason specific to the trade: your customers photograph their food. A dentist collects almost no user images. A busy cafe collects them continuously, which means the gallery a prospective customer scrolls is substantially not yours, and it drifts further out of your hands every month you do not add to it.

That cuts both ways. Customer photographs are credibility you could never buy, and they are also the badly-lit plate shot taken at eleven at night that now sits at the top of your gallery. You cannot remove it. You can out-supply it, and that is the entire strategy: a steady flow of your own images, taken deliberately, so that what a stranger sees first is the version of your restaurant you would choose.

Shoot what the customer is deciding on. The room as it looks when it is full, not empty at eleven in the morning. The signature dish as it is actually plated, not as it was styled for the launch menu. The counter, the frontage at night so somebody can find the door, the seating so a family can tell whether there is room for six, the parking situation. These answer real questions. A plate on a slate against a black background answers none of them.

Cadence beats production value here. A restaurant that adds a few honest photographs every week reads as open and busy. One that posted twenty beautiful images at launch in 2023 and nothing since reads as closed, and a meaningful number of customers genuinely assume it is. In a trade with a high failure rate, looking alive is itself a ranking-adjacent asset.

The same logic applies to posts. A restaurant has genuine news every single week — a seasonal ingredient, a festival special, a new dessert, extended weekend hours — which is more than most businesses can say, and it is why the profiles that stay active in this vertical tend to be restaurants that decided once to make it somebody’s job rather than restaurants with more time.

A large share of new Indian food businesses are not restaurants in the traditional sense. They are two or three delivery brands running out of one kitchen, or a counter inside a shared facility, or a dine-in restaurant that also operates a separate delivery-only label. This is where Google’s rules stop being generic advice and start being the difference between a listing that ranks and a listing that gets removed.

The rules are more permissive than most operators assume, and stricter in the specific ways that catch people out. Multiple businesses at one address may each have their own profile if they are genuinely distinct businesses with different names and clearly visible separate signage. Multiple virtual brands out of one kitchen are permitted, subject to additional verification. Co-located food brands each need permanent separate signage, and should only display the address if they offer pick-up to all customers.

The delivery-only case is the one operators get wrong most often. A brand that exists only for delivery is permitted a profile if it has distinct branded packaging and a distinct website, and it must add service areas and hide the address rather than displaying a kitchen customers cannot visit. Operators who instead publish the kitchen address for four brands, with no signage and no pick-up, are creating exactly the pattern Google removes listings for.

The temptation in a low-margin trade is obvious: more listings look like more visibility. In practice a cluster of thin, near-identical profiles at one address competes with itself, dilutes whatever reviews you collect across several averages, and puts every one of them at risk together. One brand with genuine signage, real packaging, its own site and a real review base outperforms four ghosts, and it survives.

If you run a legitimate multi-brand operation, the work is documentation rather than cleverness. Photograph the signage. Keep the packaging distinct and photographable. Give each brand a real website, not a page anchor. Then verification, when it comes, is a formality rather than a crisis.

It is worth closing with arithmetic, because restaurant owners are asked to invest attention in things whose payoff is described vaguely, and the payoff here is not vague at all.

The local pack shows three results. Attention drops steeply after the first, and the searcher in this vertical is unusually decisive — they are hungry, nearby, and choosing in minutes rather than comparing over days. Moving from position six, where you are on a second screen nobody scrolls to, into the visible three is not an incremental improvement in impressions. It is the difference between being in the choice set and not being in it.

Put a number on it in your own terms. If the change brings five additional walk-in covers a day at an average bill of six hundred rupees, that is three thousand rupees a day of revenue on which you paid no commission and no discount. Across a month that is around ninety thousand rupees, at a margin closer to your dine-in margin than to your delivery margin. Against that, the work in this guide is an afternoon of setup and about twenty minutes a day.

The reason to be sceptical of your own estimate of where you currently stand is that you almost certainly measure it wrong. Owners check their ranking by searching from inside the restaurant, where they are standing on the pin and always look excellent. The customer who matters is two kilometres away in a different part of the neighbourhood, and your position in their result is a number you have never seen.

Measuring rank across a grid of surrounding points is what turns this from an argument into a spreadsheet. You find out which streets you own, which you are invisible on, and whether the competitor you worry about is actually beating you or merely closer to where you happened to check. Almost every restaurant that runs this the first time discovers their catchment is smaller than they believed — which is unpleasant, and is the only honest starting point for improving it.

One last piece of arithmetic that owners rarely run: the covers you win this way are not marginal covers, they are the most profitable ones on your books. A delivery order arrives with a commission already deducted, no beverage attached in most cases, and packaging costs on top. A walk-in orders drinks, stays longer, occasionally brings four people instead of two, and pays the full menu price. If your delivery contribution margin is materially thinner than your dine-in margin, and for most Indian independents it is, then five walk-in covers a day are worth considerably more than the five delivery orders they superficially resemble. That gap is the actual return on the work in this guide, and it is larger than the revenue line suggests.