Is There Anyone to Sell To? Counting Demand Before You Spend a Dollar

Most advertising decisions are made in the wrong order. Someone decides on a budget, then picks a platform, then finds out — eight weeks and several thousand dollars later — whether anybody in their county was looking for the thing they sell.

Do it the other way around. Before you work out what you can afford, work out whether there is anything to buy. It takes an afternoon, it costs nothing, and it answers a question no amount of budget can fix.

There’s a second, practical reason to start here: demand doesn’t just tell you whether to advertise. It tells you where.

Two kinds of demand

Everything you can buy in advertising comes in one of two shapes, and the difference determines how much you’re allowed to trust your own forecast.

Demand that announces itself. Someone types water heater repair near me into a search engine at 7 a.m. because there’s cold water coming out of the hot tap. That person has declared themselves. They exist in a specific quantity, in a specific place, this month. You can count them before you spend anything.

Demand that doesn’t. Someone scrolling their feed is not looking for a kitchen remodel, but they’ve been thinking about their kitchen for two years. They’re reachable — by neighborhood, by homeownership, by income, by the fact that they just moved — but they haven’t raised their hand, and nothing in the data tells you which of them is ready.

Both are real. Both can be profitable. But they give you completely different kinds of knowledge before launch, and confusing the two is how forecasts turn into fiction.

For search demand you can get an actual number: this many searches a month, in this area. Searches, not people — one homeowner comparing three companies searches several times, and the count doesn’t know the difference. It’s still finite and still countable. Once you’ve captured most of it, raising the budget does not create more searches. It buys the same demand at a worse price.

For the other kind you can get reach — how many people could see it — and nothing else that matters. Whether any of them respond depends on your offer and your creative, which is to say: on things that don’t exist yet at the moment you’re forecasting. You cannot forecast that. You can only test it.

Say that out loud when you’re planning, because it changes what a plan even means. On search you build a plan and check it. On social you build a hypothesis and buy the smallest amount of information that will confirm or kill it.

Counting the demand that announces itself

You need three things: the words people use, your service area, and a volume figure.

The words. Not what you call the service — what a homeowner calls the problem. People do not search for hydro-jetting; they search for drain keeps backing up. They don’t search for exterior envelope restoration; they search for house needs painting. Write down what customers actually say when they call you. That list is worth more than any tool, and you already have it.

The area. Be honest about how far you’ll actually drive. A contractor who says “the whole metro” and means “twenty minutes from the shop” will build a plan on demand he was never going to serve.

The volume. Any keyword tool gives monthly search volume by area — the platforms’ own planners are free. The number will be approximate and it will be wrong in the third digit. It doesn’t need to be right in the third digit. You’re about to make a decision that only needs the order of magnitude.

Then do the arithmetic most people skip:

total monthly searches × the share you could realistically win × site conversion × close rate = jobs per month this channel can produce at its absolute maximum

That’s if you took every click in the market. Compare it to how many jobs you need.

If the maximum of the entire channel is below your requirement, stop. Nothing downstream of that matters. Your bid doesn’t matter, your ad copy doesn’t matter, your agency doesn’t matter. There is not enough demand in the room.

What “not enough” actually looks like

A fulfillment business — warehousing and shipping for online sellers — wanted to run paid search. Before quoting anything, we counted the demand.

Across every commercial term that described what they did, in their region: 335 searches a month. Not 335 clicks. 335 times a month that anybody typed anything relevant at all.

At competitive bids in that market, and at the conversion rates the business could actually achieve, the projected cost of acquiring one customer came out to several times what a customer was worth.

There was no version of that campaign that worked. Not a better-written one, not a cheaper one, not one managed by somebody smarter. The channel was too small and too expensive, and both facts were knowable in an afternoon, for free.

So we didn’t run it. The budget went into search optimization instead — one round of work on a single landing page — and the same customers ended up costing about a tenth of what paid would have.

That’s what this exercise is for. Not to talk you out of advertising. To make sure that when you do advertise, you’re pointed at a market that exists.

When it’s close

Most of the time the answer isn’t a clean yes or no. The demand is there but thin — enough to fill half your capacity, not all of it.

Say so out loud, before launch, to whoever is paying. This is the single most common place where an honest project turns into a disappointed client: everyone knew the volume was marginal, nobody wrote it down, and four months later it’s remembered as underperformance instead of arithmetic.

Then go find the rest somewhere else. Roughly in order of what it costs you:

1. Widen the area. The cheapest fix, and often free — you may already be turning down work forty minutes out because nobody asked for it.
2. Widen the service. Adjacent work your crew can already do. The searches for it are separate searches, so this genuinely adds demand rather than splitting it.
3. Go to the demand that doesn’t announce itself. Now it’s a considered choice rather than a default: you’ve exhausted the people raising their hands, so you go find the ones who haven’t yet. Expect a worse forecast and a longer test.
4. Go where demand is created by an event. One remodeling client in Hawaii runs deck-replacement campaigns triggered by storms. Nobody wakes up wanting a new deck. They wake up after a bad night of weather and want the broken one gone. Demand like that doesn’t show up in an annual average — it shows up in a week, and it belongs to whoever is already there when it does.

One caution before copying that last idea. A storm creates real, immediate demand for a damaged deck, and advertising can buy that demand directly. In roofing the same storm works completely differently — the money is in insurance claims, the selling happens on a doorstep, and advertising’s job is to make your company familiar before somebody knocks. Run a roofing campaign expecting deck-shaped results and you’ll conclude the channel doesn’t work, when what failed was the assumption.

Seasonality, which is the same problem on a clock

Annual search volume tells you how much demand exists. It doesn’t tell you when.

A business with sharp seasons doesn’t have a demand problem — it has a demand timing problem, which is worse, because the arithmetic looks fine and the year still ends badly. Half of everything you sell arrives in ten weeks. Crews sit idle either side. Cash is thin exactly when you should be buying next season’s customers.

Two things to do about it:

Look at demand by month, not by year. Every keyword tool will show you this, and almost nobody looks.

Decide, before the season, whether you’re buying ahead of it. The moment competitors start bidding is the moment prices peak — and some of that bidding isn’t strategy, it’s desperation. In markets with big tickets and hard seasons you’ll find companies bidding above what the work is worth to them, because they need cash this month more than they need margin. You cannot win that auction profitably, and you shouldn’t try. Get in earlier, at a sane price, and let them fight over the peak. More on seasonality in home services.

Which comes first, and why it matters

Notice that we haven’t talked about money yet.

There’s a reason this comes before the arithmetic. If there’s no demand, the economics don’t matter — the most profitable business in your county still can’t buy customers who don’t exist. And the shape of the demand determines the channel, which determines what you’ll be buying: clicks, or leads, or impressions on people who never asked. The math is different for each.

So: does anyone want this, are there enough of them, and where are they? Answer that first, and everything after it is arithmetic — which is the two-point test, and it’s where the actual numbers start.

This chapter is drawn from Win in the Spreadsheet First. The full method and the same treatment per trade are in the book.