Hey Build in Public gang,
Here's a list of silly mistakes I made setting up a Google Ads experiment that you can avoid - all you have to do is read this post!
Mistake #1: Not selecting the right geographies 🌎
You'd be amazed what geography selection can do for your conversion rate - the reason is even with targeted Google Ads you will get spam. If you are selling a B2B product to tech companies, then only show your ads in tech cities (SF, NYC, Seattle, LA, Chicago). I get it - people are remote now - but most tech workers have stayed in their city, or relocated to "another tech city", don't worry too much about missing out on those people who didn't.
Mistake #2: Don't test $#!% you know! 🧠
If you already know certain keywords resonate, or if you know you're not changing your target customer any time soon, then don't bother experimenting with it. Start by what's actionable. One learning we had was "for remote teams" performed well - but we had decided as a company this would be our positioning for at least a quarter.
Mistake #3: NOT re-reading the copy. Twice. Then a few more times.
Google ads will auto-generate copy for you and if you do not change it they will use it. It's not clear to me how - since it's been really good on some campaigns, and really bad on others. I had gotten comfortable with their suggestions, and some line about "trusted moving services" slipped into a text description for a few days because I didn't read the ad copy over again 🤦♂️. Next time, even though the Google suggestions are nice, will write my copy ahead of time.
Mistake #4: Not using the goals feature 🥅
I'm guessing you're trying to get signups. Then make sure you're tracking those signups! I'm sure you can count the emails you got, but if don't add that goal in Google Ads, you'll completely miss out on learning about landing page conversion. You can easily set up a goal by providing them with a URL you want users to convert to. Yes, this means a thank you page, or some separate page after signup, with a unique url. This is by far the easiest way to track this unless you're running analytics out of whatever product you do signups through.
Mistake #5: Testing without a control 🧪
Our Google Ads experiment had one too many copy changes. I was really hoping to learn about "Async Standups" - but I ended up making a few other copy changes too. i.e. "Ooo...stop playing Slack-a-mole is provocative"...For sure, but now you're also testing whether or not Slack-a-mole makes sense to people, on top of testing whether or not people care about "Async Standups." Remember that even if you have two ads, and only one variable is different, you need a control to test against!
Most of this stuff my high school biology teacher, or any entry-level marketing person, would kick me in the butt for.
But I know you all are founders, builders, hustlers, who are wearing many hats - so I thought this might be helpful for any Google Ads newbies out there.
Happy building! 🚀