When I moved from SEO into a Growth PM role, another growth PM gave me some advice: always figure out what the smallest possible version of a change is, and then do that. The point is to spend as little time and resources as possible to find out whether it’s a good idea. If it works, you build it out further. If it doesn’t, you haven’t spent too much time on something that doesn’t work.
I still think that’s good advice in most cases. But I’ve found there are exceptions.
Early this summer, product marketing at Funnel released the newest version of our product positioning. Web is one of the most visible places positioning shows up, so it was on me and my team to make sure what we say on the website matched.
Most of that was straightforward: page by page, replace the old way of talking about Funnel with the new narrative. Easy, especially because I added the positioning doc to a Claude project, fed it all our web content, and let it flag which sections needed updating.
Then I got to the homepage.
We learned over the past year that a homepage can’t be updated section by section. It isn’t a stack of independent sections, or at least it shouldn’t be. It works best if it tells one clear narrative. And a narrative doesn’t survive being edited one paragraph at a time. Do it that way and you get a patchwork: sections written at different moments, each anchored to whatever the positioning happened to be back then.
So instead of fixing the old sections one by one, my designer and I stepped back and looked at the homepage as a whole. We rebuilt it as one coherent piece, worked closely with our product marketer, then presented it to our CEO as a full new homepage instead of a few separate edits.
We got the green light. It goes live very soon.
The smallest-possible-change advice still holds for most growth work: test a new channel, try a different way of explaining the product, see whether a feature gets traction. But it breaks down when the thing you’re changing is really one story split across sections, like a homepage or an onboarding flow. You can’t patch a story incrementally. At some point you have to step back and rebuild the whole thing at once.