Stab in the Dark or Strategic? Run Every Campaign as a Test

Everything in marketing is a hypothesis until you put it to the test. I got a reminder of that in a movie theater.

In the summer of 2023 I went to see Oppenheimer at the Cinemex VIP in a mall in my area. I went in early, got comfortable, ordered my popcorn and waited for the advertising we all love so much before the show.

The ads started. The usual lineup: cars, drinks, chips. Then one caught my attention. A private jet charter company called Millennium Air.

It got me thinking. Why am I seeing this? The mall sits in a middle-class neighborhood. A charter flight is priced for a very small slice of any audience, and that slice was probably not in my row with a bucket of popcorn.

Maybe the theater has better data on its audience than I do. Who knows. The ad certainly got my attention. Whether it reached anyone who could book a flight is a different question, and it's the only one that matters to the company paying for the screen time.

Buying attention is the easy part

You can buy attention almost anywhere: a cinema screen, a billboard, a boosted post, a radio spot. Every one of them will put your name in front of people. On its own, none of them tells you whether those people can buy from you.

For a business whose sale happens in a conversation, that gap gets expensive fast. A school sells through a tour. A dental practice sells through a consultation. A contractor sells through an estimate. Every person who sees your ad with no reason to book that conversation is budget spent on the wrong crowd.

So the question behind every campaign is simple. Where is your attention coming from, and is it the right crowd?

Every campaign is a hypothesis

A stab in the dark is a campaign with no stated reason to exist. A strategic campaign starts as a sentence you can prove wrong.

Write it like this:

If we show [this offer] to [this audience] in [this place], then [this many] of them will book [this conversation] at [this cost or less] by [this date].

A preschool version might read: if we show a "book a tour this week" offer to parents of children under five who live within a 15-minute drive, on Instagram and Facebook, a set number of them will book a tour each month at a cost per tour we agreed on before launch.

Once that sentence exists, every part of it is something you can check:

• The offer. Is it something this audience wants right now?
• The audience. Can these people actually buy, and do they live where you serve?
• The place. Do they spend time where the ad runs?
• The action. Does the ad ask for the conversation you sell through?
• The number. What result, by what date, would tell you the idea is wrong?

That last line is what makes it a test. When you never decide what failure looks like, every campaign "kind of works," and you keep paying for it month after month.

What a stab in the dark looks like

You'll probably recognize a few of these:

• Boosting a post because it got likes, with no audience set beyond "people near me."
• Running the same ad on every platform and judging it by clicks.
• Changing the headline, the image, the audience and the budget in the same week, then guessing which change mattered.
• Sending ad traffic to a home page with six menu items and no booking button.
• Reading the ad platform's own conversion count and never checking it against the calendar.

Each one spends money without producing an answer you can use next month. The private jet ad might have been a brilliant bet. From my seat, there was no way to tell, and unless the charter company tracked who called after seeing it, there was no way for them to tell either.

Read the result at the conversation

The ad platforms will happily report impressions, clicks and "leads." For a business that sells in a conversation, those are early signals. The result sits further down: a booked call, a completed tour, a consultation that showed up, an estimate that turned into a job.

Reading that result means following each lead past the form. When a booking happens, the ad platforms hear about it, and so do you. Conversion tracking that connects your ads, your site and your CRM gives every lead one story, from first click to signed sale, so the hypothesis you wrote has a real number at the end of it.

The right crowd shows up in that number. Thor Construction was new to residential landscaping in Southern Ontario. Search campaigns reached homeowners already looking for landscaping in Thor's service area, and $1,400 in ad spend turned into $375,000 in pipeline. You can read the full story in the Thor Construction case study. The figure worth your attention is pipeline against spend, because that's the one that decides whether a campaign earns another month.

Run it in rounds

A single test gives you one data point. Direction comes from running tests in a steady rhythm, each one built on the last.

A simple rhythm looks like this:

1. Write the hypothesis. One sentence, with the number that would prove it wrong.
2. Change one thing. The audience, or the offer, or the creative. One per round.
3. Set the time and budget up front. Decide how long the round runs and what it can spend before anything goes live.
4. Keep the winner. Retire the loser and write the next hypothesis from what you learned.

Creative is often the quickest variable to move. New angles pulled from what your customers actually say on calls and in reviews, tested a few at a time, teach you more than the same ad on repeat ever will. That's how ad creative tested in rounds works, and the same discipline applies whether you run one campaign or twenty.

Budget follows the same rule. Once a round shows which audience and which message book conversations, that's where the next dollar goes. Paid ads managed around the booking move money every week toward what's booking and away from what isn't.

Over a few months, the guesses turn into a playbook you wrote yourself, built on your own customers and your own numbers.

Where to start

1. Write one hypothesis for the campaign you're running today. Offer, audience, place, action and the number that would prove it wrong. Any blank you can't fill in is your first test.
2. Check where your attention comes from. Pull last month's leads and mark the ones who could actually buy: right area, right need, right budget. That share tells you more than your reach number.
3. Put booked conversations on the report. Compare the leads your ad platform reported last month with the tours, consultations or estimates on your calendar. If the two numbers disagree, fix the tracking before you change anything else.