Wrapped our second beta round — 89% retention, bootstrapped, built around a hospital job and college
Wrapped our second beta round — 89% retention, bootstrapped, built around a hospital job and college
Six months ago, I got tired of standing at checkout
With 4 credit cards and no idea which to use. So I
built Divide — a Chrome extension that tells you
Which card to use based on your rewards AND your
Credit health, not just generic rankings.
Just finished our second beta round:
→ 17 testers across two rounds
→ 89% said they'd keep it installed (round 2)
→ 100% trust rate on recommendations
→ Working on Amazon, Wayfair, Target, Walmart,
and 40+ other sites
The honest part: it hasn't been smooth. Every time
I fix a bug on one site, and a new edge case shows up
On another. I'm building this as a solo founder,
bootstrapped, around a hospital job and college.
One tester wrote: "One of the few needed but totally
non-existent apps I've seen in over 1,200 submissions."
That line kept me going through a rough month.
Just shipped v0.4.10, updated the website
(dividepay.ai), and refreshed the Chrome Web Store
listing. UI redesign in progress with our designer.
Next milestone: figure out how to fix the website landing page loading. Sometimes you have to load the landing page twice for it to work. Public launch push.
Happy to answer anything about the build, the beta
process, or the product itself.
Hey Kimani,
First off, congratulations on getting to an 89% beta retention rate while bootstrapping around a hospital job and college. That's an impressive milestone.
One thing really stood out to me in your post though—you mentioned having to reload the landing page before it works.
If you're gearing up for a public launch, that's probably the last thing you want potential users experiencing.
I run Kodawa Labs, and we help businesses make sure their digital experience matches the quality of the product they've built.
Rather than just point it out, I'd be happy to take a closer look at the landing page and see if we can help identify what's causing it and improve the overall experience before launch.
No pressure—I just thought it was worth reaching out because I think you've built something genuinely interesting.
— Brian
What stood out to me is that you're optimizing for decision quality rather than maximizing rewards.
Most card recommendation tools answer "which card earns the most points?" Yours seems closer to "which card should I actually use right now without hurting my overall financial position?" That's a much more practical job to solve.
Exactly that — you put it better than I have. "Which
" Card, should I actually use right now without hurting
my overall financial position" is the real job. The
rewards-maximizing tools technically give correct
answers that are sometimes the wrong move, because
they're blind to utilization and timing. That gap is
the whole reason Divide exists. Appreciate you seeing
the distinction — it's the thing I'm trying hardest to
get people to understand.
Glad it resonated.
Your reply made me think there's one strategic decision sitting underneath that distinction which becomes much more significant as Divide grows, but I don't think I can explain the reasoning properly in a thread without oversimplifying it.
If you're interested, what's the best email to reach you on?
Would love to hear more on it
Kimani@dividepay.ai, or kimanitate@outlook.com
Thanks! I’ve just sent it over at kimanitate@outlook.com.
Looking forward to hearing your thoughts whenever you have a chance.
looking at it now
89% retention is solid for a browser extension — most lose users after the first week. On the landing page loading issue: check if your JS bundle is getting served with aggressive cache headers. Chrome's service worker can serve a stale version on first load if the cache-control is set too far out. Adding a version hash to the bundle filename usually fixes it.
Really appreciate this — both parts. The retention
number surprised me too, honestly; I was bracing for
The first-week drop-off everyone warns about.
And thank you for the cache-control tip — that's
genuinely useful. We've had a landing page load issue
where a refresh fixes it, which lines up exactly with
what you're describing about the service worker serving
a stale version on first load. Going to pass the
version-hash-on-the-bundle-filename fix to my developer
and test it. Didn't expect to walk away from this post
with an actual bug fix — thanks for that.
Really interesting progress, Kimani. The landing-page loading issue you mentioned is definitely worth looking at alongside the redesign, especially since even a small amount of friction at that first step can affect how many visitors actually move deeper into the product.
I’d be curious to see whether the redesign is also being approached from the conversion/UX side, not just visually. Would be happy to share a few things I’d look at if useful.