You Don't Fail: How to Run Marketing as a Series of Tests

By 4 min read
You Don't Fail: How to Run Marketing as a Series of Tests

Treating marketing as a test means every campaign has a clear question, a small budget, a time limit and a rule for what you will do with the result. If it works, you expand it. If it doesn't, you find out why and adjust. Either way, you come out knowing more than when you started.

John's line is short: "You Don't Fail. You succeed ... or you learn." It sounds like a poster in a break room. It is actually a practical way to run your marketing, as long as you set things up so the learning actually happens.

What does it mean to treat marketing as a test?

It means you stop thinking of a campaign as a bet you either win or lose. You think of it as a question you are asking the market.

"Will homeowners in Mesa book a free estimate from a Facebook ad?" is a question. "Will a postcard to past customers bring back repeat business?" is a question. You run the test, you get an answer, and the answer tells you what to do next.

The difference matters. A bet that loses is just money gone. A test that comes back negative still pays you something: you now know where not to spend, and usually you have a clue about why.

Why do so many businesses learn nothing from a bad campaign?

Because the reaction to a flop is usually blame, not analysis. The agency was bad. Facebook doesn't work for us. The new hire wrote weak copy. Then everyone moves on.

Amy Edmondson wrote about this in Harvard Business Review in Strategies for Learning from Failure. Her point is that organizations learn poorly from failure because they blame people instead of analyzing what caused it. She also separates failures into different kinds: preventable ones, unavoidable ones, and intelligent failures that come from experiments.

That last kind is what a good marketing test produces. It is the kind of failure you should be willing to have, on purpose, in small doses.

In marketing terms, the three might look like this:

  • Preventable: the landing page link was broken, the phone number was wrong, nobody answered the leads. Fix the process.
  • Unavoidable: a platform changed, a competitor flooded the market, monsoon season knocked out a week of outdoor work. Note it and adjust.
  • Intelligent: you tried a new audience or offer with a small budget, and it didn't land. That is a result. Use it.

How do you set up a marketing test that protects your budget?

A test that teaches you something is designed before it launches. Here is the simple version.

  1. Write the question. One sentence. What do you want to find out?
  2. Pick the measure. Calls, booked appointments, form fills, sales. Choose the number that answers the question. Our post on what to measure before you spend more goes deeper on this.
  3. Set the budget and time limit. Small enough that losing it does not hurt. Long enough to get a real answer.
  4. Decide the rule in advance. "If we get this many booked calls at this cost, we expand. If not, we stop and review." Write it down before you see results so you are not tempted to move the goalposts.
  5. Change one main thing. If you change the audience, the offer and the ad all at once, you will not know which one made the difference.

Say you run a cleaning company in Gilbert and want to try offering move-out cleans. Instead of redoing the whole website and launching a big campaign, you run a small ad to renters with a clear offer for a few weeks, and you track calls. That is a test. Whatever happens, you know more and you have not bet the business.

What should you do when a campaign doesn't work?

First, resist the blame reflex. Then ask a few plain questions, in order:

  1. Did it actually run the way we planned? Check the boring stuff first: links, forms, tracking, budget delivery, phone coverage.
  2. Did the right people see it? Wrong audience is one of the most common culprits.
  3. Did the message make sense to them? Was the offer clear and worth acting on?
  4. Was there an easy next step? An interested person who cannot easily call, book or buy is a lost lead.
  5. What did we learn that we can use next time?

Write the answers down. A short note per test, kept in one place, turns into a playbook over time. It also stops you from rerunning the same failed idea a year later because nobody remembered.

Then make a decision. This is how we work: keep what works and expand it, fix what slips, and shut down what fails fast to protect the budget.

How do you know when to stop testing and commit?

Testing is not an excuse to never commit. Once a test hits the rule you set, act on it. Put real budget behind the winner and run it properly.

Keep a little room for new tests alongside it. Markets shift, and what worked last year may not work next year. But the bulk of your money should go to what has already proven itself, not to an endless string of experiments.

Also remember that numbers are only part of the picture. A test can look flat on paper and still teach you something important about your customers, which is why marketing metrics aren't everything.

The bottom line

Marketing is not pass or fail. Each campaign either brings in business or teaches you how to. The trick is setting it up so the lesson is clear and the cost is small, which matters when you are an East Valley small business watching every dollar.

Ask a clear question. Keep the budget small. Decide the rule ahead of time. Look for causes, not culprits.

If you want help designing tests that fit your goals and your budget, start with a Strategy Session. Or let's talk strategies about what you are trying now.

Frequently asked questions

How do you test marketing on a small budget?

Write one clear question, choose the number that answers it, and set a small budget and a time limit. Decide in advance what result means expand and what result means stop. Change one main thing at a time so you know what made the difference.

What should I do if my marketing campaign failed?

Check the basics first: links, forms, tracking and whether anyone answered the leads. Then look at the audience, the message and the next step, and write down what you learned. Focus on causes rather than blaming a person, platform or vendor.

What is an intelligent failure in marketing?

Amy Edmondson distinguishes intelligent failures, which come from experiments, from preventable and unavoidable ones. In marketing, that is a small, planned test of a new audience, offer or channel that does not work but gives you useful information. It is a cost worth paying in small doses.

How long should a marketing test run for a Mesa or Gilbert small business?

Long enough to get a real answer, which often means several weeks for local ads, depending on budget and how often people buy. Set the time limit before you start so you are not tempted to quit early or let it drift. The right length depends on your service and sales cycle.

Can a Best Shot Marketing Strategy Session help me plan marketing tests?

Yes. The $495 Strategy Session covers your goals, the message that resonates and the most efficient, profitable direction, which is where good tests start. From there we help you decide what to try first and how to judge it.

Keep reading

Start with strategy. We will help you build from there.

Tell us where you want your business to go. We listen, then build a customized program just for you.

Let’s talk strategies