Skip to content
GullySales

Notes for owners · Business growth

How to identify your advertising audience

Start with your own last fifty orders, not a workshop. Then translate the description into something a platform or a media owner can actually sell you.

The GullySales team · Updated 21 Sept 2026 · 8 min read

The answer is already in your own records. Before anyone runs a workshop or writes a persona with a stock photograph on it, pull your last fifty invoices, your call register, your WhatsApp enquiries and the delivery addresses. Write down, for each, what was bought, at what value, from which area, how they found you and who paid. Two hours of that produces a truer audience than a week of discussion, and it costs nothing. Then the second job begins, which is the one people skip: turning the description into something a platform or a media owner can actually sell you.

Read what you already have

Your invoice book gives you the value, the frequency and the geography. Plot the postcodes of a hundred orders on a map and most local businesses find that three quarters came from inside a few kilometres, which settles the media question by itself.

Your call register gives you the question people ask first. If eleven of twenty callers asked about price and three asked about delivery time, you know what your advertisement has to answer.

Your Google Business Profile insights give you the search words people used to find you, and how many asked for directions rather than called.

Your sales team gives you the thing no report will. Ask them which customer was easiest to sell to and why, and which kind of enquiry they dread. The second answer is as useful as the first.

Three people, not one

Most purchases have three roles, and they are only sometimes the same person.

Who uses it. Who chooses it. Who pays for it.

A coaching class is used by a student, chosen by a student and paid for by a parent. A tile adhesive is used by a mason, chosen by a contractor and paid for by a homeowner who never hears its name. A hospital scanner is chosen by a radiologist, approved by a trust and negotiated by a purchase manager. Write down those three names for your own product first. A campaign aimed at the wrong one of them fails quietly and expensively.

Now translate it into something you can buy

This is where most audience work stops too early. A description is not a targeting option.

What you wrote downWhat you can actually buy
Small business owners in BengaluruSearch terms for the problems they look up, trade association access, referrals from their CA
Women aged 25 to 35 who want to start a businessSearch intent and interest proxies, not the intention itself; age and gender only outside restricted categories
Procurement heads at large factoriesNothing directly. Vendor registration, marketplace listings, and the engineer who writes the spec
People looking for a first homeBroad geography and search intent only, because housing targeting is restricted
Our past customersA customer list upload where you hold consent, and a lookalike built from it
People who work in that tech parkA postcode radius, the building's own media, a Tuesday
NRIs from our districtA country, a language, and creative that names the district

The right-hand column is your media plan. If a row has nothing in it, that audience is not reachable by advertising and needs a different kind of effort: a sales visit, a registration, a partnership, a presence at the place they already gather.

Where the translation is not allowed

Housing, credit and employment advertising sit in a restricted category on the large platforms. Inside it, targeting by age, gender, relationship status, detailed demographics and narrow location is switched off, and location has to be broader than a locality. This applies to a builder, a home loan agent, a lender and anyone advertising a job.

It is not a technicality to be worked around. Accounts are restricted for it. Where you land in that category, the work moves into search intent, broad geography, creative that states the price and the criteria so people select themselves, and offline media where the rules do not apply.

Say what does not work, out loud

An audience sheet that lists only channels is half finished. Write the negative half too, because it saves more money than the positive half makes.

A procurement head cannot be reached until there is a live requirement. A consultant will not fill in a form. A contractor will not open a PDF. A student cannot sign the cheque. An owner will not answer an unknown number at two in the afternoon. Each of those sentences removes a line item from somebody's budget.

One page per audience

Keep it to a single sheet, in these seven lines: who pays, who chooses, where they are and at what hour, what event sets off the purchase, what you can actually buy, what does not work on them, and how you will know it worked.

For example, a diagnostic centre in Rajajinagar might write one sheet for walk-in patients within two kilometres and a second for referring doctors in the same area. Same business, two audiences, two completely different plans, and neither sheet mentions the other's channels.

Test it before you commit the year

Run each audience for four to six weeks with a small budget, a separate phone number or code, and one question asked of every enquiry: how did you hear of us. Write the answers in a register rather than trusting a dashboard, because two platforms will each claim the same order.

Then compare cost per enquiry, and compare what those enquiries were worth. An audience that produces half the enquiries at twice the order value is the one to keep.

What to do next

Print the seven lines above and fill them in for the audience you spend most on today. The line most people cannot fill is the fourth, which is what triggers the purchase, and that gap explains most of the wasted budget in a small business.

If you would rather go through your own enquiry records with someone reading them for the first time, that is what the free audit is: forty five minutes on where enquiries come from, how fast they are answered and whether they were ever buyers, followed by a written and scored report.

Questions

Questions owners ask.

We have never done this. Where do we start with no data?
With your last fifty invoices and your call register, which every business has. Write down for each what they bought, where they were, how they found you and who paid. Fifty rows of your own history beats any persona written in a meeting room.
Our audience is everybody. Is that a real answer?
It is a real answer for a very small number of businesses, and it is the wrong answer for almost everyone who gives it. Even if anyone could buy from you, the people who buy most often and complain least are a much smaller group. Advertise to them and the rest will still turn up.
Should we pay for a market research report to find our audience?
Only after you have read your own records and spoken to ten customers, because a bought report will not know your town or your counter. Research earns its cost when you are entering somewhere new or sizing a market you have never sold to.
How many audiences should one business have?
Two or three at most, each with its own creative and its own landing place. Beyond that the budget is split so thin that none of them gets enough repetition to work, and nobody can tell which one produced the orders.

From the blog

More notes for owners.

Get a free audit of how you sell, and a scored report of where the work is.

Book a free audit