Episode 20 recap: You changed five things. Which one worked?

If you missed today's call, here is the whole thing in one post.

𝗧𝗵𝗲 𝗽𝗿𝗼𝗯𝗹𝗲𝗺

Most owners fix everything at once. New headline. New VSL. New page copy. New posting plan. New sales call. All in the same week.

Two weeks later the number moves. Everybody is happy. Then somebody asks the only question that matters: what did it?

Nobody in the room can answer.

Melissa said it better than I did. She went makeup shopping and the girl behind the counter had changed her primer, her base, how many layers she used, and her setting spray. All at once. Now she has no idea which one broke it. She is going to be mixing percentages forever.

That is not testing. That is a Powerball ticket.

𝗧𝗵𝗲 𝗲𝗻𝗴𝗶𝗻𝗲: 𝗟𝗮𝗯 𝗮𝗻𝗱 𝗧𝗿𝘂𝗰𝗸

Organic is the lab. Paid is the truck.

The lab is where you are allowed to be wrong. It costs you time, not money. You change one thing on purpose and you get an answer in days. (most times weeks, depending on volume of traffic)

The truck costs you money every day. It amplifies a message. It never creates one. You leave it alone while it runs.

Nothing rides the truck without an audition. Here is the correction I want to make: that line is not just a content rule. It is the whole testing system. You do not test on the truck.

𝗧𝗵𝗲 𝘁𝗵𝗿𝗲𝗲 𝗿𝘂𝗹𝗲𝘀

1. One variable. Change the hook or the page, not both. If two things moved, the result belongs to neither one.

2. One window. Same days, same budget, same days of the week. A holiday week against a normal week is not a test. It is a holiday week.

3. One floor. About 1,000 visitors or 50 conversions. Under that you are reading noise and calling it a finding.

Then write the guess down before you start. If you cannot say in one sentence what you expect to happen and why, you are not testing. You are redecorating. This is the classic scientific method applied to your business.

𝗧𝗵𝗿𝗲𝗲 𝘁𝗶𝗺𝗲𝘀 𝘆𝗼𝘂 𝘀𝗵𝗼𝘂𝗹𝗱 𝗻𝗼𝘁 𝘁𝗲𝘀𝘁

1. You are below the floor. Forty visitors a day cannot separate two headlines. You cannot split test your way out of an empty room.

2. The traffic is not real. Two accounts we run went from 40 to 60 visitors a day to a thousand in a day. One was bots. The other was junk traffic from messaging that pulled the wrong people. Any test read that week was fiction.

3. You got bored. It works at small scale; it stops being interesting, and you go build the next thing instead of scaling this one. I do this. It is the most expensive habit I have.

Boredom is not data. Most people do not have a testing problem. They have a finishing problem.

𝗧𝗵𝗲 𝗱𝗶𝗮𝗴𝗻𝗼𝘀𝗶𝘀: 𝘄𝗵𝗶𝗰𝗵 𝗵𝗮𝗹𝗳 𝗶𝘀 𝗯𝗿𝗼𝗸𝗲𝗻

Front end. Nobody opts in, nobody books. That is the promise, the traffic, or the page. Do not touch the sales call.

The seam. They book and do not show. That is your confirmation and reminder sequence, and whether you told them what the call actually is.

Back end. They show and do not buy. That is education and sales motion. Do not touch the ads.

Find the half first. Then run one test inside that half.

𝗧𝗵𝗲 𝘀𝘁𝗼𝗿𝘆 𝘁𝗵𝗮𝘁 𝗽𝗿𝗼𝘃𝗲𝘀 𝗶𝘁

A brand we work with was certain the ads were broken. Clicks were fine. Cost was fine. The traffic was real. Add to cart was above average. About 60% made it to the checkout page, which is normal for e-commerce.

Then 0.5% finished.

It was not the ads. It was the offer. Competitors had a money back guarantee, shipping included, better pricing, and real bundles. This brand had all sales final, paid shipping that pushed the cart up by as much as 40%, and a 10% off bundle.

Six months of new creative was never going to find that.

𝗢𝗻𝗲 𝘁𝗵𝗶𝗻𝗴 𝗜 𝘀𝗵𝗼𝘂𝗹𝗱 𝗵𝗮𝘃𝗲 𝘀𝗮𝗶𝗱 𝗼𝗻 𝘁𝗵𝗲 𝗰𝗮𝗹𝗹

Cold and warm traffic do not break in the same place.

Warm traffic already knows you. They show up to the call. They convert off a thin sequence because the trust got built before they ever clicked.

Cold traffic has none of that. So when a funnel works for warm and dies for cold, the ads are usually not the problem. The sequence is. The cold visitor needed the education, the proof, and a reason to show up, and the funnel assumed they already had it.

Before you blame the ad, ask which temperature is failing. If warm converts and cold does not, you do not have an ad problem. You have a sequence problem.

𝗧𝗵𝗶𝘀 𝘄𝗲𝗲𝗸

Run one test. Just one.

1. Answer the two questions and name your half.

2. Pick the biggest lever in that half. Not the easiest one.

3. Change that one thing and freeze everything else for 14 days. 30 days if you are running ads.

4. Go find the thing that already worked and never got scaled. Everybody has one.

Post your half and your one lever in the comments. One line is the whole post.

𝗜𝗳 𝘆𝗼𝘂 𝗰𝗮𝗻𝗻𝗼𝘁 𝘁𝗲𝗹𝗹 𝘄𝗵𝗶𝗰𝗵 𝗵𝗮𝗹𝗳 𝗶𝘀 𝗯𝗿𝗼𝗸𝗲𝗻, 𝗱𝗼 𝗻𝗼𝘁 𝗴𝘂𝗲𝘀𝘀 𝗮𝘁 𝗶𝘁 𝗮𝗹𝗼𝗻𝗲

DM me the word AUDIT.

I will send you my calendar link. No application. No form. We walk your funnel front to back and you leave that call knowing exactly which half is leaking and the one test to run next.

Nothing rides the truck without an audition. Nothing gets fixed until you stop changing five things at once.

See you next Tuesday.

1:05:24

0 comments

Episode 20 recap: You changed five things. Which one worked?