Growth Newsletter #051

Even if your Facebook video ads have strong hook, that doesn’t mean users will stick around to watch the whole thing.

And if your ad can’t sustain their interest, people probably won’t convert.

To find out if your video ads are generating interest, look at their average play time. Then look at their performance data in Ads Manager. There's a chart that shows the percentage of video plays for different seconds of your video ads. Use it to find out at exactly what times users drop off—then map these points in your video.

Consider why users might not be motivated to keep watching at these points:

• Is it after the first cut?
• Does text appear?
• If so, how long is it?

You're not looking to compare it to a play time benchmark—video ad lengths vary drastically between companies.

But you can look for the most intense drop-offs and consider editing your video ads to hold attention at those critical time periods.

A common misconception about experimentation is that it’s all data, all the time.

Data is a critical part of it, yes. But there’s a human element to testing too. After all, you are trying to understand your users. On both ends, people are involved. While researching, hypothesizing, designing tests, and analyzing results, judgment is key.

One area where judgment comes in: when to make the call to override A/B test results.

Imagine you run a copywriting test:

1. One version of the copy is true and honest. You accurately portray your value props.
2. Another version slightly overstates the benefits of your product.

And the second version wins the A/B test.

You'll earn more short-term profit by going with the data and implementing the second version of the copy. But doing so could actually harm your customer lifetime value—those who purchased might feel deceived and likely won't purchase again.

You're better off overriding A/B test results when your winning test variant is harmful to your long-term sustainability and growth.

Another example:

When Netflix was testing Grace and Frankie (show) promos, they found that users clicked more on an image with just one co-star, instead of the two co-stars together. Netflix went with the image of the duo anyway.

Learning: You should also consider overriding your findings if they present ethical or legal concerns, don’t align with your company values, or run the risk of demoralizing your team.

Most companies don't measure leads generated from content.

If you don't measure, you can't calculate customer acquisition cost (CAC)—there's no way for you to tell whether your content marketing is profitable.

Here's an easy way to measure acquisitions from content marketing using a simple ‘thank you' page and Google Analytics.

Two steps:

1. Configure your opt-in forms so they point to a unique 'thank you' page after a user enters their email. Example: https://growandconvert.com/thank-you/ Anyone who lands on this page must have opted in through your lead capture form(s). Nobody else will see this page.
2. Set up a new goal in Google Analytics: Once your lead capture forms are set up to send opt-ins to that thank you page you made, you’ll need to configure the page as a new goal in GA. Here's a 2-minute setup video that shows you how.

GA will track how many users reach your thank you page (a “Goal Completion”). Measure that number against the cost it takes you to create your content, and you'll be able to calculate your CAC.

Fluff muddies messaging.

It's also kind of inconsiderate: It makes readers put in more effort than they need to understand your point.

Here are some free resources for keeping your copywriting fluff-free:

• The Handy List of Human Words—A good list for converting robotic language into human language (e.g., "deactivate" —> "turn off")‍
• Use Simple Words and Phrases—Another good list, this one from a group of federal workers who believe government writing should be plain and clear.‍
• Hemingway Editor—Copy your writing into the editor to see what its reading level is. (6th grade is good; 10th is too hard.) Check for too many adverbs, complex phrases, instances of passive voice, and hard-to-read sentences.‍
• Readability Test Tool—Copy text or a URL into the tool to check its readability according to the Flesch-Kincaid formula, and get helpful stats like average words per sentence.