Recently, I’ve partnered with Hugo Thiphaine, Search Engine Optimization specialist with 10 years of experience in the industry. Together, we mapped where app discovery actually happens now, in a NeoAds article:
Store search is semantic
Three surfaces, one shift.
On top of this, we thought it’d be even better if we actually looked at real apps, and applied the theory to reality.
We picked the US Health & Fitness category. Health apps live with regulated metadata, health-data and medical-claim restrictions, and review scrutiny other categories do not have.
But the sharper problem is on the web, and it’s a double penalty. Fitness, nutrition and wellness content is what Google calls YMYL “Your Money or Your Life” content that can affect someone’s health or finances. Google holds it to a far higher E-E-A-T bar: visible expertise, real credentials, trust signals. Most fitness apps run thin, anonymous, SEO-by-numbers blogs that fail that bar completely.
Here’s the trap: the thin content fails twice. It doesn’t rank on Google, and it doesn’t get cited by AI engines because both lean on the same authority signals.
🎯
The most instructive audits aren’t on struggling apps, but on the ones that look like they’ve already won. 2 patterns keep showing up:
The incumbents built a discovery machine 5 to 8 years ago, it worked, and they’ve barely touched it since. You can watch the slow decay in the charts as newer apps eat their edges.
The newcomers have the opposite problem: they’re winning right now, on product and virality, with nothing underneath.
Four apps, all winning but building discovery debt.
Ladder (#7 US Top H&F Charts) with the abandoned blog
Ladder did the hard thing right. A real content blog, written by a credentialed coach (NASM-certified CPT) which is precisely the E-E-A-T signal health content needs and almost nobody has.
Then they stopped.
The most recent post is dated February 2022. And the whole thing sits on a subdomain, blog.joinladder.com, walled off from the main site. So whatever authority it earns passes almost nowhere useful. They nailed the part most apps can’t do, but did not push there.
BitePal wins in 2026: slick product, viral hooks, a million-plus users, and a homepage that’s a hero, 3 feature cards and 3 weight-loss testimonials.
Credit where it’s due as the site is server-rendered, so AI can actually read it. That’s actually more than most.
But on the downside, there’s no blog, no informational content, nothing for Google to rank or an LLM to cite beyond the brand name itself.
The weight-loss app is making specific before-and-after claims, with no named expert anywhere on the site, and has a “Sources” page that is, well, pretty basic.
The science meant to back the product lives on a simple page, almost hidden . It passes zero authority and barely counts as content at all.
Right now they don’t need the web. But the real question is how long that lasts.
Unrot got a launch sprint, then… nothing
Unrot is an app, pretty similar to Bitepal in its branding, which is meant to replace doomscrolling with healthy habits.
It has more content that Bitepal, and even opened a blog with 5 posts, answering exactly the trend-driven queries its audience types into Google and ChatGPT: “What is brainrot?”, “Dopamine detox doesn’t work”, “Rebuilding your attention span”.
All five went up in a single week in early July 2025, then nothing. And as of writing, only 2 of the 5 are actually indexed. In a topic moving this fast, that’s the worst possible place to stop since AI engines tends to cite fresh content. A blog that stops doesn’t hold and fades.
But the worst finding is in a file they’ve probably never opened. Unrot’s robots.txt blocks GPTBot, ClaudeBot, Google-Extended, CCBot and Applebot-Extended. Except they didn’t do it… Cloudflare did. The block sits inside # BEGIN Cloudflare Managed content, injected at the edge, on by default, most likely in a dashboard toggle they’ve probably never seen.
And it’s the wrong half. Those aren’t the crawlers that cite you: OAI-SearchBot, PerplexityBot, Googlebot and Applebot are all still allowed. Those are the crawlers that train on you. The default is shaped with publishers in mind: protect the IP, keep the citation traffic. Sensible, if you monetize pageviews, like medias. For apps, it’s a different story.
They wrote the right content, for the right queries, and then instructed the machines not to learn from it.
LeanBites, a goldmine locked
LeanBites is an app that scores 30,000+ menu items across the chains people eat at: Chipotle, Starbucks, McDonald’s, Chick-fil-A.
That dataset is the single most valuable content asset in its category. “Healthiest order at McDonald’s”, “Lowest-calorie Starbucks drink”, “Best macros at Chipotle”… These are enormous search volume and AI-answer queries, exactly the kind of specific, comparative, data-backed question a model loves to have a source for.
The thing is that every bit of it is sealed inside the app: no blog, no indexable page whatsoever. LeanBites is invisible for the exact questions it was built to answer.
E-E-A-T is mandatory, especially in health where it’s non-negotiable. So here’s a quick checklist if you have an app in the category:
Put a human on it: a named author with real credentials, a bio, a photo, and a link to their LinkedIn or professional profile. Ladder is the only one of the four that did this.
Mark that human up: person schema with sameAs pointing to their credentials. This is you telling the machine who said it, in the machine’s own format. Without it, a model reading your page has an anonymous claim.
Cite your sources on your own domain, build internal linking to support your own content and facilitate crawlers’ work.
Show your work: original data beats another listicle every time. LeanBites is sitting on 30,000 scored menu items and publishing none of it.
Keep going, as freshness is a ranking and citation signal.
Conclusion
None of these 4 are failing companies. That’s the entire point of this article. I’ve been working with many mature apps, that are struggling to stay on top. Competition is tougher than ever, and old organic strategies are eroding quickly. Paid-and-viral might work for a time, but won’t run forever if you’re missing solid foundations.
Organic discovery is getting more opaque, more crowded, harder to differentiate in. The apps that win the next cycle won’t be the ones with the most keywords. They’ll be the ones that are present, clear, and everywhere intent forms.
✨ Want to stay ahead in Growth for Mobile Apps & Games?
Subscribe now & get the latest trends delivered straight to your inbox!
Subscribe now