Case study: Anywhere Club (aw.club) — copy (EN)

Online community for IT professionals · part of the EPAM Anywhere ecosystem

Anywhere Club: zero to 2,296 Top 10 keywords in 27 months

Lead: The project we were handed after our results on EPAM Anywhere Business. The content was written by working IT professionals rather than copywriters — which turned out to be both the biggest advantage and the biggest problem.

• 16,672 visits a month, the highest absolute traffic in our portfolio
• 42,928 keywords ranking, from nothing
• 57.6% of traffic non-branded

Small print under the bar:

Ahrefs data, May 2022 to August 2024. Ahrefs estimates traffic from its own index, so traffic figures are an estimate rather than a counter reading; positions are measured directly.

How we got this project

There was no pitch. We were already working on EPAM Anywhere Business, and on the strength of those results the client handed us a second product in the same ecosystem.

Trust expanding inside one company is probably the best reference an agency can offer.

What the product is

Anywhere Club is a community for IT professionals. Not a corporate blog — a separate product with sections for relocation, learning, partner job listings, online events, volunteering and an expert blog.

The client’s goal was stated very plainly: bring a junior developer into the club so they learn there and grow into a mid-level engineer.

The starting line

May 2022. Ahrefs shows zero — no traffic, no keywords, no positions. New product, new site, empty results pages.

Build an organic channel for a product at launch, in two languages at once, for an audience of junior developers — people searching not for a service but for answers about their own careers.

Which meant:

• content that answers career questions instead of selling;
• every article needed in two language versions;
• experts had to write it, or a junior developer wouldn’t believe a word.

A community is a product with no price list and no sales team. The question of where a limited resource should go is just as sharp as on a commercial project, so directions were chosen through the WEDGE framework.

W — Where the client can win

A competence map for a community looks different from one for an agency: instead of services and technologies, the career tracks where the EPAM ecosystem holds real expertise.

The map is adaptive: the set of tabs changes with the type of business, the logic doesn’t.

E — Evaluate whether we can get in

The second screen produced the defining decision: not “what is DevOps” but how to become one.

The logic is simple. Everyone writes definitions, and that’s a wall of encyclopedias and large portals. A career route — “DevOps Roadmap”, “How to Transition to DevOps”, “AWS DevOps Engineer Job Description” — is a query with visible competition, and one only someone who has done the job can answer.

Real topics from the project: DevOps Roadmap · How to Transition to DevOps · AWS DevOps Engineer Job Description · Azure DevOps Career Path · JavaScript Roadmap · JavaScript Career Path · Data Engineer Skills · Cloud Computing Career Path · How to Become a Cloud Architect.

D — Decode who decides

Here the “person who decides” isn’t a procurement committee. It’s a junior developer choosing where to take their career.

The client’s goal was stated very plainly: bring a junior into the club so they learn there and grow into a mid-level engineer. Everything else follows from that — what to write about, in what register, and who should write it.

The answer to that last question defined the project: practitioners write it. A junior developer can tell a practitioner’s text from a copywriter’s inside one paragraph.

G — Guide them to a decision

Every topic shipped as a pair — English plus Russian. At peak, up to 12 pairs of articles in flight at once.

To keep that from turning into chaos, each brief went through an internal chain before the client ever saw it:

“ brief author → "ready for review" → reviewer leaves comments → revision → "cleared for work" → PM sends it to the client “

Small things got checked too. Geo settings on keyword research were verified one by one — “I had New York set, switched it to LA, added 3 keywords.”

And one stage that usually doesn’t exist: a monthly check of the published text against our own brief. More on that under “What was hard.”

E — Expand the coverage

The URL restructure, January 2024. The client decided to drop the category segment from article URLs — editors kept changing categories, and the URLs moved with them.

We built the redirect map from three sources: a Screaming Frog crawl, indexed pages and 404s from Search Console. Every 404 was worked through by hand.

Real constraints: only 1,000 pages could be exported, so part of the redirection had to be handled algorithmically; there were 404s on language versions, so we pointed a missing Ukrainian URL at the live English one; and the ?recId= parameter needed its own canonical treatment.

Two URLs were flagged as having nowhere sensible to point. That’s more honest than redirecting someone to “somewhere.”

Team: nine people, the largest share at 26%. The lead specialist also ran anywhere.epam.com — the same person on the neighbouring product.

Rhythm: weekly follow-ups without a break from May 2022 to August 2024. Monthly cycle: topics → keyword research → briefs → internal brief review → checks on published articles → implementation check → report.

Volume: 364 tasks, 1,019 hours.

Experts write the way they see fit

The defining feature of this project: the content came from practising IT specialists, not copywriters. That buys a credibility you can’t fake. It also creates a problem familiar to anyone who has worked with subject-matter experts: a brief isn’t binding on an expert.

Here’s a real check on an article about facilitation, October 2022:

– keywords not used: “facilitation what is it,” “facilitation in simple terms”; – supporting terms not used: initiative, optimization, technique, motivation, risks, meeting, management; – H1 — we recommended one, another one is live; – description — we recommended one, another one is live; – no FAQ block, though the audit recommended one.

What we did about it. We didn’t argue, and we didn’t send letters about recommendations going unimplemented. We built checks on published articles into the monthly cycle — and then, in January 2024, ran repeat checks on the same articles, replacing the text wholesale on the existing URLs.

In other words we accepted reality: the text won’t match the brief. So work with what got published rather than with what was planned.

The bottleneck was writing speed, not SEO

For the first eighteen months the pace was uneven — between 2.6 and 38.8 hours a month. Not a pause; work ran every week. Just slowly.

The reason was simple: the client couldn’t write fast enough. The constraint was the throughput of expert authors, not the agency and not the budget.

Then, on 1 August 2023, this went into the project notes:

“Agreed a fundamental change of approach with the client — they’ve started focusing on it and have untied our hands.”

The client handed over two decks — Personas and Skills — and the pace rose to around 59 hours a month. Over the following twelve months traffic reached 16 thousand.

It’s the clearest illustration we have that in content projects the limiting factor usually isn’t the agency.

Positions, Ahrefs:

Estimated traffic, Ahrefs: 0 to 16,672 visits a month — the highest absolute monthly figure across all our projects.

Caption:

Per Ahrefs. Two distinct speeds are visible: the slower 2022–2023 stretch and the sharp acceleration after August 2023 — explained under “What was hard”.

Link profile: 201 dofollow domains.

Work delivered: 364 tasks, 1,019 hours.

Content from practitioners. The articles were written by IT specialists. A junior developer can tell a practitioner’s text from a copywriter’s inside one paragraph.

Career paths instead of definitions — that’s what the second screen produced. “What is DevOps” is a wall of encyclopedias. “How to become one” is a query with visible competition that only a practitioner can answer.

We checked what was published, not what was delivered. We compared live pages against the briefs and rewrote them, rather than complaining that recommendations went unused.

Expert content is worth the friction. Yes, the expert will ignore the brief. But the credibility they bring can’t be reproduced by a copywriter, and a brief can be reapplied after the fact.

Check what’s published, not what’s delivered. The most useful ritual on this project was the monthly comparison of live pages against the briefs. Without it, recommendations exist only in documents.

Content velocity is the client’s velocity. For eighteen months this project moved slowly, and SEO wasn’t why. Once the client set priorities, traffic reached 16 thousand within a year.

Bilingual content needs a pipeline, not enthusiasm. Twelve article pairs a month only hold together on a process with internal review before anything reaches the client.

If there’s nowhere to redirect, say so. We flagged two URLs in the map as having no sensible destination. That beats sending someone “somewhere.”

Work stopped on 1 August 2024: the client cut the internal budget and moved all SEO to an in-house team.

A calm ending with no conflict — by handover the channel was working and documented well enough for someone else to run it.

Sound familiar?

A new product, expert-written content, two languages, an audience looking for answers rather than services. We’ve run this before.

Питання