Build #2 · Week 1 · Launched Jun 5, 2026
LiveMyNubs
Free calculators and honest care guides for dog and cat owners. Built to be the answer AI search actually cites.

The idea
What it is & why I built it
MyNubs is a pet brand, not a calculator dump. The bet is pretty specific. Calculators are one of the few pages AI search hasn't gutted. If you want to know how much to feed a 40-pound dog, an AI summary doesn't cut it. You want to punch in your own numbers and get your own answer, so you still click through. Plain blog posts are losing a third of their traffic to AI answers right now. Calculators barely feel it.
So the whole site is built to be the thing AI assistants quote for dog and cat math. Real calculation logic, a named methodology, sources right on the page. Then send those people somewhere they actually come back to. That's the funnel. The real product is a mobile app I'm building after this. A companion that knows your pet by name and holds their whole history. Weight, vaccines, meds, vet visits, photos. It reminds you Bella's heartworm dose is due Tuesday. An AI can tell anyone the calorie count for a 35-pound dog. It can't do that. That's the point.
Year one is unglamorous on purpose. Own the calculator searches, become the source that gets cited, build an audience. The app and the subscription come later, once there's real traffic to point at. Slow burn by design.
The build
How it was built
This stretch was a content sprint. MyNubs went from a handful of calculators to an actual library. 179 pages total. 128 articles and 24 calculators (13 dog, 8 cat, 3 both). The articles are the stuff pet owners actually search at midnight. Breed care guides, breed-vs-breed comparisons, the why-does-my-dog-do-this explainers, symptom guides, grooming how-tos. Plus the smaller stuff that makes it feel like a real site instead of a tool. A breed comparator, a should-I-call-the-vet symptom checker, quizzes, printable reference cards.
The boring-but-important layer is everything aimed at getting cited by AI. Structured data on every page so a machine knows who wrote it, a clean sitemap, an llms.txt, a robots file that actually lets the AI crawlers in instead of blocking them. The calculation logic is original, tied to a named methodology, not formulas copied off some other calculator site. Some articles carry affiliate links now too, so the content can start paying for itself the moment traffic shows up.
Plumbing stays cheap on purpose. Domain's at Namecheap, no upsell circus at checkout, and the whole library runs on a basic static host.
The stack
Tools used
Hand-built static pages build
179 pages, each its own file. Loads instantly, very little that can break.
Search + AI plumbing discovery
Structured data, sitemap, llms.txt, open to AI crawlers. Built to be the source ChatGPT, Perplexity and Google's AI answers quote.
Original calculation logic methodology
Real formulas with a named methodology behind them, not recycled from other calculator sites.
Affiliate links money
Live on some articles now, so the funnel can start earning while rankings climb.
Mobile app on Supabase the real product
A companion app that knows your pet by name and tracks their whole history. The actual product, coming once the funnel has traffic.
The numbers
Analytics so far
One week in, indexed and showing up. 328 search impressions with zero promotion. Zero clicks though, because nothing's cracked page one yet. Everything's sitting on page two or three, which is exactly where a brand-new site starts. The early front-runners are the ones the whole strategy is built around. The dog food portion calculator pulled 189 of those impressions and is already near the top of page two, with the British Shorthair vs Persian comparison right behind it. About ten real people showed up too, around 32 seconds each. Tiny numbers. But they're the right shape.
The recap
What worked & what I'd do better
What worked
- Indexed fast and already pulling 300-plus search impressions a week with zero marketing. The calculator-plus-citation bet is showing the early signal I wanted.
- The front-runners are calculators and a breed comparison, exactly the pages the strategy's built around. Nice to see it hold up in the actual data.
- The whole library cost almost nothing to stand up. About $11 for the domain, $9 a month to host.
What I'd do better next time
- Zero clicks so far because nothing's on page one yet. The job now is climbing from page two to the top. Sharper titles, more internal linking, time.
- I hadn't promoted any of it. That changes next. Social, and chasing the links and mentions that pull rankings up faster than just waiting.
- The named author and methodology pages matter more than I'd treated them. For pet health that's a real trust signal, and part of why an AI decides to cite you. Building those out properly.
- A slow week one was always the plan. SEO compounds. I'm not stopping after seven days, I'm stacking the next seven on top.