What is a low-competition keyword on a quick commerce app?

One where no brand holds most of the first ten unpaid results and few brands hold paid slots on it. We call a keyword open when it is among the least crowded third of an app’s keywords by those measures, and crowded when it is among the most crowded third.

How do I know if a keyword has demand when no app publishes search volume?

Read the signs the apps do show: how many brands held paid slots on it, whether it is a head word or a long-tail one, and how often the app lists products shoppers searched for and could not get. None of them is a volume, and together they rank keywords against each other.

Should a smaller brand start with head keywords or the long tail?

Usually the long tail. Head keywords are where the largest brands and most advertisers are. Long-tail words in your own category are often open, and a first-screen place there is within reach.

Can I see where competitors rank on each keyword?

Yes, from outside the app: type the keyword as a shopper in many cities and count where each brand’s products come up among the unpaid results. A rank worth quoting is the usual place over many searches, never one search.

Guide · Search intelligence

How to find keywords you can win on quick commerce search

Which keywords are open and which are crowded, how to read demand when no app publishes search volume, and why the long tail is often where a brand starts.

A brand cannot be first on every word. The useful question is narrower: which keywords, on which app, can my products realistically reach the first screen of, and which of those do shoppers actually type? Both halves can be read from outside the app, if they are read carefully.

Half one: how crowded is the keyword?

Take one keyword on one app, type it in many cities, and look at the first ten unpaid results each time. Three counts describe how hard it is to get in:

The leading brand’s share. If one brand holds most of the first ten slots across the searches, the keyword is effectively taken.
The first three brands’ share. A keyword where three brands hold almost everything is closed in a different way: a contest between incumbents.
How many brands hold paid slots on it. Paid slots sit above or among the unpaid results. A keyword where many brands hold them leaves less of the first screen to the unpaid results.

None of these is meaningful alone, and none is meaningful across apps. Compare a keyword with the other keywords on the same app: open (the least crowded third), contested, or crowded (the most crowded third).

Half two: is anyone searching for it?

No quick commerce app publishes how often a word is searched, so a search volume for these apps can only be an estimate. What the apps do show are signs of demand:

Brands holding paid slots on it. The apps place paid slots on the searches shoppers make, so a keyword where many brands held one is a keyword that is searched.
Head or long tail. “Milk” is typed more than “lactose free milk 1 litre”. A list of keywords can be split into head words and long-tail ones, as a judgement, and the split is worth stating as one.
Searches the app could not fill. Some apps list products a shopper searched for and the store did not have. A keyword that often shows that list is one where shoppers want more than they find.

Putting the halves together

The keywords worth trying are the ones with some sign of demand that are not crowded, in a category you already sell in, where your products are not already in the first three. Three patterns come up again and again:

1. Just below the first ten

Your product keeps coming up at places 11 to 20. A listing fix, a better title or availability in more stores can be enough to cross onto the first screen.

2. The open long tail

A specific word in your category (a pack, a flavour, a need) where no brand dominates. Few shoppers type it compared with a head word, but the ones who do are close to buying, and the first screen is within reach.

3. A head word where your rival is weak

A crowded keyword is usually not worth chasing, unless the brand holding it is the one you compete with every day and you want the meeting with the category manager. Then the evidence is the city-by-city place of both brands.

How to read rank without fooling yourself

Results move from city to city and from one search to the next. A rank worth acting on is the usual place over many searches, with the city-by-city places under it. Count paid results apart: a product in a paid slot was not ranked there.

On Shelf, the keyword finder on the Search screen sets each keyword beside these measures: how crowded it is, its signs of demand, your usual place and the first five brands’ places.