Case study: TopFlightApps — copy (EN)

Mobile development · healthcare · US market

TopFlightApps: 13 to 3,131 Top 10 keywords in five and a half years

Lead: 1,340 tasks — the most of any project we’ve run. 4,865 hours. Not one month off since May 2020. Growth that doesn’t rest on a single good year.

• 889 keywords in the Top 3, up from six
• 91.3% of traffic non-branded
• Inquiries from the US rose 124% in two months

Small print under the bar:

Ahrefs data, May 2020 to September 2025. Ahrefs estimates traffic from its own index, so traffic figures are an estimate rather than a counter reading; positions are measured directly.

1,340 tasks is the highest count of any project in the agency.

What the client brought

Top Flight Apps builds custom mobile applications. The market is the US, and the competition there is exactly what you’d picture: hundreds of studios, the same promises on every site, directories like Clutch doing most of the lead generation.

May 2020: 1,176 keywords ranking, 13 of them in the Top 10, an estimated 80 visits a month. Half that traffic was branded.

What that meant

The site existed in search results without winning any of them. A thousand keywords producing thirteen top positions is the profile of a site that ranks for a bit of everything and wins nothing.

Build an organic channel that produces development inquiries — in the US market, where every competitor is trying to do the same thing.

Which meant:

• find a segment that can actually be won, instead of fighting for “app development”;
• get the site technically clean and keep it that way for years;
• build a link profile without one-off purchases;
• measure the outcome in inquiries, not positions.

The one case in the set where the framework ran without its first step. We normally start with a competence map — with the question “where can you win.” Here the client already knew: healthcare. So WEDGE became EDGE — evaluate, decode, guide, expand.

E — Evaluate whether we can get in

“App development” is a wall. Hundreds of studios, identical promises, directories like Clutch doing most of the lead generation. A site with 13 Top 10 keywords isn’t getting in there.

“Telemedicine app development services” is a specific market with a specific buyer and visible competition.

The second screen produced not several directions but depth inside one: healthcare unfolded into telemedicine, EHR/EMR integrations, mental health, fitness, medical billing.

⭐ D — Decode who signs the contract

In 2022 we ran a separate audience study: who healthcare startup CEOs actually are, what they read, where they can be reached.

Not guesswork — SparkToro data, written up as a task and turned into decisions.

Today the most valuable pages on the site are the healthcare ones:

G — Guide them to a decision

“Content Reoptimization Task” repeats in the tracker every month from 2020 through 2026. Content that brought inquiries last year may not bring them this year.

Since May 2025, adjusted for AI results. That month the task is named, for the first time, “Content Reoptimization Task for AI Overview.” From June 2025 onward it runs monthly as “(AI included)” — sixteen straight months.

And the key point about the funnel: the page-level attribution shows inquiries come from articles, not service pages. After the home page every top source is in /ideas/ — how-to and how-much. People work out the problem first and look for a supplier second.

E — Expand the coverage

Technical health as a routine, not a one-off audit. The 2021 audit was delivered in six parts — not one eighty-page document nobody implements, but portions small enough to finish. Then quarterly technical mini-audits, separate tasks for implementation and separate tasks for verifying it happened.

The link profile was cleaned, not just grown. A link audit in 2022, monthly link building audits through all of 2023, a disavow pass in January 2024, tier 2-3 work, link exchanges, and a standing monthly task for vetting outreach targets that still runs today.

Two products in parallel. Since early 2025 the client’s second product, Specode, runs on the same rhythm — its own content plans, internal linking and technical audits.

Counting inquiries, not rankings

Since January 2024 the project carries a separate monthly task: the lead report. Not “report,” not “traffic” — leads, its own line, every month without a gap.

Positions are an intermediate metric. That isn’t what the client is paying for.

Volume: 1,340 tasks and 4,865 hours — our highest task count on any project. An average of 62 hours a month for six and a half years.

Evenness: 146 tasks in 2020, 209 in 2021, 201 in 2022, 225 in 2023, 227 in 2024, 221 in 2025. No sprint year, no collapsed year. That’s what a 226× outcome actually looks like from the inside.

Team: a core of seven, the largest single share at 22%. Load spread wide enough that the project doesn’t stall when someone takes leave.

Two products in parallel: since early 2025 the client’s second product, Specode, runs on the same rhythm — its own content plans, internal linking and technical audits.

A market where everyone’s pages look alike

US mobile development is a niche where competitors say the same things in the same words. You don’t get out of that by writing better copy.

What worked instead: narrow to healthcare, study the actual buyer with data rather than intuition, and hold the direction for six years.

Reoptimization never ends

“Content Reoptimization Task” repeats in the tracker every month from 2020 through 2026. Content that brought inquiries last year may not bring them this year — results pages change, competitors change, and since 2025 the shape of Google’s answer changes.

Projects where content gets written and forgotten don’t survive in this niche.

Two things to watch instead of one

Since 2025 there are two products in play. That isn’t double the budget or double the team — it’s a planning requirement, so the second product doesn’t quietly eat the first one’s hours.

Positions, Ahrefs:

Estimated traffic, Ahrefs: 80 → 18,056 a month. That’s 226×, the largest multiple in our portfolio.

Caption:

Per Ahrefs. Traffic there is an estimate from its own index; the position figures are measured directly.

Inquiries. The project ran monthly lead reporting. Across the fifteen months held in full (January 2023 – March 2024):

Inquiries from the US — the target market — rose 124% in two months in early 2023. The client’s own note in the same report: “a 45% increase in intake forms filled versus January for leads in the United States, the highest in a year.”

Page-level attribution. The client recorded which page each inquiry first landed on — 1,588 contacts attributed over the period:

Which gives the project’s main finding: inquiries come from articles, not service pages. After the home page every top source is in /ideas/ — the how-to and how-much material.

Link profile: 1,329 referring domains, 12,007 backlinks.

Work delivered: 1,340 tasks, 4,865 hours.

Depth in one direction instead of width. The client knew they needed healthcare. Our work began at the second step: assessing where inside that direction we could actually get in — and unfolding it into telemedicine, EHR integrations, mental health.

Six years at one pace. Between 201 and 227 tasks a year. No bursts, no gaps. You don’t get 226× out of one good quarter.

Inquiries as the unit of reporting. A monthly lead report since January 2024, with page-level attribution. Positions are an intermediate metric.

In a crowded niche, narrowing beats spending. Healthcare instead of “app development” — and the site’s most valuable pages today are exactly there.

Research the audience; don’t picture it. A standalone study of healthcare startup CEOs, built on SparkToro data, gave a focus that held for six years.

A technical audit in parts gets implemented. A whole one doesn’t. Six parts, each finished, beat one large document every time.

Link profiles need cleaning, not only growing. Disavow passes and monthly link audits are dull work that only becomes visible when nobody did it.

Content reoptimization is a routine, not a project. Monthly since 2020. Since May 2025, adjusted for AI results.

In a long deal cycle, articles produce inquiries — service pages don’t. Of 1,588 attributed contacts here, every top source after the home page is how-to or how-much material. People work out the problem first and look for a supplier second.

The engagement continues. Two of the client’s products are in active work, with monthly content reoptimization for AI results and monthly lead reporting.

Sound familiar?

A crowded market, a US audience, a need for inquiries rather than rankings. We’ve been running exactly that for six and a half years.

Питання