Case study: EPAM Anywhere Business — copy (EN)

IT outsourcing · US and Western Europe

EPAM Anywhere: zero to 1,626 Top 10 keywords in 16 months

Lead: A brand-new section on a global company’s site. Handover baggage from the previous vendor. A budget that got cut in month one, and three months to show something. We stopped trying to cover the whole site and backed four verticals instead.

• 15,323 keywords ranking, 337 of them in the Top 3
• 138 pages rewritten to our optimization briefs
• 1,800 referring domains earned over the engagement

Small print under the bar:

Ahrefs data for anywhere.epam.com/business, November 2021 to March 2023. Ahrefs estimates traffic from its own index, so traffic figures are an estimate rather than a counter reading; positions are measured directly. The curve for this window is visible in Ahrefs to anyone with an account — go check it.

Who the client is

EPAM Anywhere is EPAM’s remote work platform, and it runs two very different motions that are easy to confuse.

One brings in engineers. A developer applies, gets interviewed and tested, enters the system, and is then matched with clients. The other one, anywhere.epam.com/business, brings in the projects themselves: companies that need a team built.

We worked on the second one only.

What we inherited

The Business section was effectively new. As of November 2021, Ahrefs recorded nothing for it at all — no ranking keywords, no traffic. Not a low base. An empty one.

Along with the section came the previous vendor’s work: a stack of briefs, of which exactly one section was useful to us, and a site structure that had already been signed off. We decided not to rebuild that structure right away. Get a result first. Come back to the foundation later.

The constraints we started with:

• The budget was cut in the first month. It’s in the September 2021 report.
• Three months to show movement. Not the usual “SEO takes six months to a year” — an actual deadline, after which the project could have been shut down.
• Competitors were global outsourcing firms with budgets we couldn’t match, and weren’t going to pretend we could.
• A hard naming rule. No text anywhere — site copy, guest posts, crowd marketing — was allowed to say just “EPAM”. Always “EPAM Anywhere”. That applied to every one of several hundred pieces.

The client wanted inquiries from companies that needed development work. Not traffic. Inquiries.

Which meant:

• pull the Business section into organic search from a near-zero base;
• show movement within three months so the project would continue;
• grow visibility in topics that bring people with budget, rather than chasing volume;
• do it on a reduced budget.

Directions were chosen through our WEDGE framework, which answers one question: where should a client’s limited resource go to return the most.

W — Where the client can win

At EPAM Anywhere the competence map wasn’t built from scratch — it was reworked. The tracker says exactly that: “Rework the competence map.”

That’s the normal situation at a large company: many directions, some stale, some new, and without a review the resource spreads across everything at once.

We don’t assess anyone’s engineering. How strong EPAM is in Big Data is EPAM’s own judgement. What we bring is different: market movement, who already holds the results pages, how much traffic is there and whether it’s reachable.

E — Evaluate whether we can get in

The second screen produced the choice of verticals and services worth investing in. The tracker names them: “Selecting and re-clustering vertical and service pages — Retail, BigData.”

Then a monthly cycle: keyword research for three or four vertical and service pages, briefs for optimizing them, implementation checks.

D — Decode who signs the contract

Outsourced development isn’t commissioned by one person. We worked out who decides, what they already know, what worries them, and what they need to see to take the next step — and that fed straight into the briefs.

⭐ G — Guide them to a decision

Here the project produced something we’ve seen almost nowhere else: a tool inside the funnel.

The software development cost calculator isn’t an article or a service page. It’s an interactive step for someone already working out a budget. And it produced almost half of all inquiries.

The rest came from fintech material covering the earlier stage: how to build a payment system, how to create trading software, mobile banking, P2P lending. Five of the eight top pages are the same subject.

That’s a funnel working: the article explains, the calculator calculates, the form collects.

E — Expand the coverage

The structure was built as verticals × services × solutions — and the task names show the intersection rather than the separate axes: “vertical and service pages, Retail, BigData.”

The team: six specialists and a project manager. One lead SEO, two people on links and outreach, three on content. The manager created 220 of the 331 tasks, which tells you planning wasn’t something squeezed in around the edges.

The rhythm: a monthly cycle that repeated more than twenty times.

“ project review → monthly plan → link building plan → execution → implementation check → internal QA → client report “

Reporting: a monthly document plus a separate slide deck. On top of that, a time report comparing planned against actual hours, task by task. Link building was billed separately and invoiced up front, so every month its tasks were pulled out of the main report by hand. The client shouldn’t pay twice for the same work.

Internal QA: 21 times over the engagement we reviewed our own tasks before anything went to the client. There was even a separate review of shortlisted outreach sites, run before any money moved.

Analytics: goals configured, a Data Studio report built, conversions tracked in GA4. When the client asked for more than a position export — they wanted our read on the numbers — we produced a standalone analysis of the commercial pages.

The commercial pages were much tougher than the blog

Traffic climbed. The pages that actually convert didn’t move nearly as fast. By January 2023, of 12 commercial “hire a developer” pages, only three held the Top 10 consistently. None held the Top 3.

We told the client why, plainly: traffic potential on those queries is thin, and the queries are partly geo-dependent, so results differ country to country. But their conversion potential beats any blog article, which is exactly why we kept pushing them.

What worked: hire-java-developers reached the Top 10 largely on the back of three links from the home pages of niche sites built on recovered domains. Three links moved that page further than months of content work had.

What got in the way: some of our briefs were implemented only in part. The first recommendation in our own analysis said as much — apply the brief fully, add the blocks we specified, don’t cherry-pick. Nine of twelve pages missed the Top 10, and that’s the main reason.

One line of code erased part of eighteen months’ work

On 26 March 2023, blog pages started serving by mistake. They began dropping out of the index.

April organic traffic fell 35.7% according to the client’s own data. Google happened to be rolling out a core update at the same time, which made it impossible to separate the two effects. The blog pages were already gone.

We documented the cause in both the March and April reports. That’s why the measurement window here ends in March 2023. Everything after 26 March is the consequence of a technical fault on the site, not an outcome of the work.

We only ever saw part of the lead flow

GA4 had exactly one event configured as a conversion: a click on “Contact us” in the header. In Q1 2023 there were 77 of those clicks from 60 users, which was 38.89% of all site conversions.

That’s the top of the funnel, not a signed deal. From there leads went into EPAM Anywhere’s own CRM, which we had no access to. So we can’t tell you how many contracts organic search produced, and we’re not going to imply that we can.

Organic traffic, Ahrefs, /business section:

Positions, Ahrefs, March 2023:

Caption:

Organic traffic for the section per Ahrefs, November 2021 to March 2023. Ahrefs estimates traffic from its own index — an estimate, not a counter reading. The position figures above are measured directly.

Work delivered:

• 138 pages with optimization and re-optimization briefs
• 18 consecutive months of pre-publication review on every article
• 1,800 referring domains across the engagement
• 1,860 hours of team time

Inquiries: a sevenfold increase in six months.

Comparing two quarters of work:

By autumn 2022 the site was producing roughly seven times the inquiries it had at the start.

Where they came from:

The calculator produced almost half of all inquiries. One interactive tool converted on a par with the entire rest of the site combined.

The highest-performing pages were the fintech material: how to build a payment system, how to create trading software, mobile banking, P2P lending. Five of the eight top pages are the same subject. The content direction is confirmed by inquiries rather than by positions.

More than half of all inquiries came from the US, the target market.

Focus beat coverage. Four verticals instead of the whole site. With a reduced budget and global firms on the other side, going wide loses to going deep.

Links were the main line item, not an afterthought. Thirty-nine percent of project hours. Not one-off placements either — maintenance: checking, re-indexing, recovering links that got pulled or turned nofollow.

The client let us work. EPAM Anywhere shared values docs and persona research, signed off briefs, and shipped them. Where implementation was partial, so was the result, and you can see exactly that in the hire-page numbers.

With three months to prove something, narrow the front. Four directions carried through beat twenty done halfway.

A brief nobody checks is a document, not a change. We verified 19 times in 20 months. Where the client shipped only part of a brief, the page stalled short of the Top 10.

Traffic and pipeline grow at different speeds. The section reached 1,626 Top 10 keywords. Over the same period, twelve commercial pages produced three steady Top 10 rankings and zero Top 3. An honest case study shows both numbers.

Placing a link isn’t the same as keeping it. Auditing live placements and chasing down removed or nofollowed ones is its own workstream. Without it the profile quietly erodes.

One line in can undo eighteen months. A stray noindex cost 35.7% of traffic in a single month. Monitoring meta tags isn’t box-ticking.

{ Sound familiar? }

New section, cut budget, a deadline to show results. We’ve run this before.

Питання