Restaurants and Cloud Kitchens: Hyperlocal Auto Campaigns That Fill Orders

Almost every other medium a restaurant can buy is either too big or too small. A newspaper insert covers a whole edition zone, most of which will never order from you. A hoarding puts the brand on one road, seen by people driving past on their way somewhere else. Instagram reaches the right radius, but reaches it in a feed where a single outlet competes with every food brand in the country. Auto advertising is the rare medium shaped like a delivery business.

Why Does Auto Advertising Fit a Delivery Radius So Well?

Autos are short-haul vehicles. A driver working a locality does school runs in the morning, station and market trips through the day, and residential drops at night, mostly inside a 5 to 8 km working area. Over a month the same auto passes the same colonies, the same market lanes and the same office gates dozens of times. That is the geometry of a delivery catchment, and it is why food is one of the categories autos serve best.

The full impression and CPM working is in our 2026 cost guide. For a single outlet, though, the raw impression count matters less than who those impressions belong to. They belong to residents of the zone you deliver into and people who physically stand near your kitchen, which is the only audience one outlet can serve.

What Is Different About a Cloud Kitchen With No Storefront?

This is the part most food brands get wrong on a first campaign. A dine-in restaurant is advertising a place: someone reads the hood, recognises the name from a signboard they pass daily, and walks in. A cloud kitchen has no signboard and no door. It exists inside Swiggy and Zomato, where it is one tile in a scroll of two hundred.

So the campaign is fighting recognition, not giving directions. The customer opens the app, sees your tile beside four they have never heard of, and picks the name that feels familiar. Repeated street-level exposure is precisely what manufactures that familiarity. Across the campaigns we have run since 2020, the food brands that do well on autos are the ones that accept this and write the hood accordingly: brand name large, cuisine instantly clear, one reason to choose it. A hood that tries to carry a full menu carries nothing.

How Many Autos Does One Outlet Need?

Zone density decides whether a food campaign works. Our planning rule is 25 to 40 hood autos per zone, and for most single-outlet restaurants and cloud kitchens the delivery radius is one zone. Thirty autos circulating a 4 to 5 km catchment means a resident meets the brand several times a week without looking for it. The same thirty autos spread across a metro means nobody sees it twice, and the campaign is over before it starts.

Media only. Printing is a one-time cost on top, and metro union charges are extra in cities like Mumbai, Delhi and Bangalore; we quote all of it as separate line items. Multi-outlet brands scale by zone rather than by adding autos to one pool: three outlets means three catchments planned independently, because each kitchen serves its own circle. The method is in our fleet-size and zone-density planner, and the targeting logic in our guide to hyperlocal zone targeting.

Which Format Should a Food Brand Buy?

For food, hood branding at ₹580 per auto per month is usually right rather than the cheaper sticker formats. The reason is not prestige, it is appetite. Food sells on colour and on the image of the dish, and the hood is the only panel on an auto large enough to carry a photograph that still reads across a road. A 24x6 inch strip at ₹140 per auto per month carries a name and a number legibly, but it cannot make anyone hungry.

The smaller formats earn their place as a volume layer on top. A brand that wants its name everywhere in a locality at the lowest cost per impression can run 30 hood autos as the appetite layer and add 24x18 inch side panels at ₹230 per auto per month on another 40 autos for frequency. That is a budget decision, not a verdict on which format is better. Whatever the format, the rules for a moving panel do not bend for food: one message, high contrast, readable from 30 metres in three seconds, one response mechanism. The detail is in our 30-metre rule for auto creative.

What Should Actually Go on the Hood?

1. 1The brand name, as large as the panel allows. This is the asset the campaign is buying. If a viewer takes one thing away, it must be the name they will recognise later in the app or on a signboard.
2. 2One food image or one cuisine word. Biryani, South Indian, bakery. The category has to be instant, because nobody decodes a logo at a traffic signal.
3. 3One reason to choose you. A signature dish, a price point, or a delivery promise. One only.
4. 4A direct-order number or WhatsApp line. For food this outperforms a QR code, because a passenger in traffic will save a number far more readily than scan a code on a moving vehicle.
5. 5Nothing else. No menu list, no five social icons, no address block in eight-point type.

How Do You Measure Orders That Arrive Through an App?

Honestly, and with expectations set correctly. This is the weakest link for food, and pretending otherwise is how agencies lose clients in month two. Someone who sees your auto on Tuesday and orders on the app on Friday leaves no trace connecting the two. Four things we set up before a food fleet goes live:

• A dedicated number or WhatsApp keyword printed only on the autos. Every call to it is a clean, attributable, commission-free order, and it is the strongest single signal available.
• Aggregator dashboard trends by locality. Both major platforms expose order volume and new-customer counts. Watch the delta for the branded catchment through weeks 4 to 12, not week 1.
• A geo holdout. With outlets in two comparable localities, brand autos around one and leave the other alone for eight weeks. The gap between them is the honest read, and it is the method we recommend in our guide to tracking response on transit ads.
• Branded search volume. Note searches for the brand name in the city for four weeks before launch, then watch the curve. For a cloud kitchen this is often the earliest visible movement.

Set the timeline honestly too. Food is a habit category. First attributable calls usually appear within two to three weeks, but the meaningful change is order frequency among existing catchment customers and recognition among new ones, and that accumulates through months two to four. Three months is the minimum honest run for a food fleet; a one-month test stops exactly when the compounding starts.

When Does This Not Work for a Food Brand?

• A fine-dining destination drawing from the whole city. If customers travel 20 km for a reservation, the catchment logic does not apply and the money belongs elsewhere.
• A kitchen with a capacity or consistency problem. Advertising accelerates whatever the product already does, so a campaign that sends first-time orders into missed delivery times converts trial into bad reviews faster than it builds a brand.
• A brand that will pull the plug in four weeks. Recall in food is cumulative, and month one is the least representative month there is.
• Scattered coverage bought on price. Twenty autos across a metro because the total invoice looked comfortable is the most common way food budgets get wasted.
• A hood carrying the full menu. If it cannot be read in three seconds at 30 metres, it was not advertised.

A food brand on autos is buying one thing: familiarity inside the circle it can actually serve. Get the zone right, put 25 to 40 hood autos in it, and give it three months.