I just added a static, non-AI feature to my AI-powered link platform. Was that the wrong call?
I spent the last few weeks building a link-in-bio feature for trimy.io, and I'm still not 100% sure it was the right move.
Quick context: trimy started as a link shortener, but honestly the AI layer is what the product is about now, it writes your campaign summaries, tells you which A/B variant is actually winning and why, scores your campaign health so you're not staring at raw click charts trying to guess what matters. Marketers using it don't really think of it as "a shortener" anymore, they think of it as the thing that tells them what's working.
Then I went and shipped link-in-bio. Which is... a static page with some buttons on it. No AI anywhere near it. It's the least "smart" thing in the entire product, and it's aimed at a completely different crowd; creators, Instagram/TikTok people not the marketers the AI stuff is built for.
My logic at the time was pretty simple: I already had all the link infrastructure sitting there, bio pages are basically "a list of your links," so why not. Plus free users get a "Made with trimy" footer on their bio page, which is a nice little viral loop I didn't have before.
But a few days after shipping it I started second-guessing myself. Am I now two products glued together, a smart insights tool for agencies, and a plain link-in-bio page for creators? Does that confuse people the moment they land on the homepage?
Genuinely asking, not fishing for reassurance:
- Has a feature ever pulled your ICP in a totally different direction like this? What did you do about it?
- Does "all your links, AI included" actually land as one pitch, or do people just get confused about what the product even is?
- If you found a tool that did both — real AI analytics and bio pages — would that read as "wow, a lot of value" or "wait, what is this thing for"?
The feature's live either way and I'm going to A/B test the positioning regardless. Just curious if anyone else has been stuck at this exact fork.
The feature is only a second product if it earns a different buyer and roadmap. Treat link-in-bio as the capture surface for your existing analytics: the same marketer creates the page, distributes it, then learns which destinations and variants work. If creator signups never activate the analytics layer, keep the viral page but separate it from the core homepage promise.
I really appreciate how you've explained this, it's been super helpful. The idea of a "capture surface for existing analytics" really makes sense to me now. I was thinking about bio pages on trimy.io and how they already track clicks just like any other link, so it's basically just another way for the AI to get more data to summarize. It's funny, I was thinking of it as a "creator feature" before, but I can see how that might be the wrong way to look at it. This explanation has really clicked with me, thanks for sharing your insight. It's great to have a new perspective on this, and I'm excited to think about how we can use this to improve our analytics.
The part I'll steal directly is the activation test. If people who sign up through bio pages never touch the analytics side, that tells me everything.
I'm thinking about this one thing: how long should we wait before making a decision about that group? Users who look at bio pages might need several weeks to get enough traffic for the analytics to be useful, so I'm not sure if 14 days of not being active means we've got the wrong buyer or if it's just too soon to tell. Maybe we should give them more time to see what happens.
Use an exposure window, not just a calendar window. Fourteen days with 20 visits proves nothing; start the activation clock after a bio page reaches a minimum such as 100 unique visits, with a 30-day cap so the cohort still closes.
The exposure window fixes exactly the blind spot I had, I was about to judge that group by a date, when really the question is whether the page even got enough traffic for the analytics to say anything. Where'd you land on 100 uniques though, is that from experience with something similar, or just a reasonable starting number you'd tweak once you see the real traffic spread?
It was a reasonable starting point, not a universal threshold. I'd set the first floor high enough that one or two conversions can't swing the result wildly, then revise it once you see the actual traffic and activation rate. If bio-to-analytics activation is 5%, 100 uniques only gives you about five activations: enough for a directional read, not a final decision. I'd keep the 30-day cap and move the exposure floor once you know that base rate.
That reframes it well, the 100 wasn't really the point, the base rate is, once we actually see what bio-to-analytics activation looks like for us, the floor kind of calculates itself and the directional read vs final decision thing is useful too, I was treating 100 uniques like a verdict when it's really just "enough data to look at the group at all."
Appreciate you walking through this, going to build the base-rate check first before I even touch the number.
Just keep it off the homepage. Let marketers stumble into it inside the product and let the "Made with trimy" footer pull the creators in. You keep the loop and the homepage still says one thing.
That split makes sense, keep the homepage saying one thing and let people find the second use case inside the product. Hadn't actually connected the "Made with trimy" footer to this until you said it. One thing I keep wondering though, once branding removal becomes a paid perk, doesn't that flip the incentive? Your best bio page users are the ones who'll want to pay specifically to turn that footer off, so the free virality loop and the upgrade path start pulling against each other a bit.
It's worked very well for Lovable, as if you are able to have a community play then when be share the products or what they have made with it it acts as promotion for your company.
Good example, but I think the two loops aren't quite the same strength. When someone shares a Lovable build, the thing itself is the promotion, whoever sees it is looking straight at what got made with the tool. A bio page footer's more like a signature at the bottom of a link list, easy to just skim past even when it's sitting right there so the loop probably still works, just weaker than something like Lovable's.
Makes me think the free tier needs to stay good enough that plenty of creators just sit on it for a while before upgrading, since that's really what keeps the footer around instead of it disappearing the second anyone gets traction.