Case study: Aimprosoft — copy (EN)

IT outsourcing · Java, SAP Hybris, Liferay · US market

Aimprosoft: 14× traffic in 4.5 years, with two strategy pivots inside it

Lead: The longest completed project in our history: 55 months, 894 tasks, 3,094 hours. Over that time we changed direction twice — once when the pandemic broke the market, and once when it turned out the client sold something other than what we were promoting.

• Traffic from ~150 to ~2,200 visits a month
• Pages were chosen for promotion by their contribution to inquiries
• 62% of traffic from the target market, the US

Small print under the bar:

Ahrefs data, June 2018 to December 2022. Ahrefs estimates traffic from its own index, so traffic figures are an estimate rather than a counter reading.

Our longest completed project — four and a half years without a break.

Aimprosoft has a narrow and fairly uncommon specialization: Java development, SAP Hybris, Liferay. An enterprise stack people approach with a specific problem, not with “build us an app.”

June 2018: roughly 150 visits a month by Ahrefs’ estimate.

This is the cleanest example of our WEDGE framework in the portfolio: steps W and E sit side by side in the tracker, month after month.

W — Where the client can win

We opened with a competence map — the task “Build a competence map for the project” is dated in the first weeks of work.

The map is five tabs the client fills in: solutions, verticals, languages and frameworks, technologies, services. Each row carries scores from 1 to 3: how easy this is to sell, what the company earns on it, how the market is moving. The client scores the first two. We score the third.

We don’t audit anyone’s engineering and we don’t pretend we can. To judge how strong a company is in Java you’d have to read the code. A marketing agency isn’t getting in there, and that’s fine. What we bring instead is what the client doesn’t have: market movement, who already holds the results pages, how much traffic is really there.

E — Evaluate whether we can get in

Then comes what sets this project apart. From March to June 2020 the tracker carries a separate task every month: “Market analysis against the competence map.”

Plus a standalone one: “Analysis of growth potential in the Hybris, Alfresco, Liferay niche.”

The second screen produced a result the client didn’t love at first: Alfresco came off the priority list. The project’s own wording — “hasn’t sold well for a long time.” The market exists, the expertise exists, the sales don’t.

⭐ When the market broke, the map was rebuilt in two weeks

March 2020. The pandemic, frozen budgets, and no obvious answer to what anyone should publish.

Instead of guessing we pulled IDC research and identified the industries where spending would grow through the crisis: videoconferencing, intelligent supply, chatbots, e-learning, healthcare, the public sector. And separately, where it would fall.

Then the step that matters most: we handed that list to the client and asked which of it they could actually deliver. Content went into production only after the answer came back.

That’s the division of roles in WEDGE. We bring market movement. The client says what they can do. An article about e-learning, for a company that doesn’t build e-learning, produces visitors rather than inquiries.

D — Decode who signs the contract

In 2022 the client handed over their own Buyer Persona and Decision Process document. Our assessment was blunt: the personas were weak, parts of them inaccurate.

We cleaned them up, kept four, and developed them into push-pull form — what pushes a persona away from their current state, what pulls them toward a decision. That fed the article briefs.

The method matured later: on another of our projects the personas are now built into a dedicated tool that every brief passes through.

G — Guide them to a decision

The tracker carries a task with a telling name: “Updating and building funnels in the content plan, based on the competence map and the pages created.”

The funnel wasn’t built separately from the map. It was built out of it.

And in February 2022 the framework produced the project’s main result. A decision recorded in the project: shift the focus from outsourcing to outstaffing — for a plain reason stated at the time, “because that’s what they’re good at, and what they sell.”

The pivot changed how pages were selected:

“For promotion we can also consider the pages that brought in good leads last year.”

The priority list was named page by page:

This is rare. Promotion lists are usually built on traffic potential. This one was built on what had already produced revenue.

E — Expand the coverage

The site structure agreed in November 2020: home → industries → technologies (Hybris, Liferay, Alfresco and others) → services → portfolio → blog. Plus a dedicated Liferay Consulting page, aimed at the query where what sells isn’t development but advice.

That’s an intersection: not “Liferay” alone and not “consulting” alone.

The breakdown of 894 tasks shows both axes held steady:

Content and links almost equal. Across four and a half years a tilt in either direction becomes structural — here it never happened.

HARO, with a note on style: we advised writing answers to journalists in a more conversational register. A comment written like a paragraph of technical documentation doesn’t get picked up.

When inquiries dropped, we said it wasn’t SEO

August 2021. The client’s inquiries fell. Exactly the situation where an agency can comfortably sell a rescue plan.

We ran a separate analysis and reached a different conclusion:

“There are no changes in traffic, positions or clicks over the period studied… I don’t see the influence of any SEO factors here. Especially since the project has now started growing more actively.”

We checked the CRM export too: it showed not a drop in inquiry quality but simply an absence of inquiries in that period.

The cause sat outside our remit, and we wrote exactly that.

A detail that costs inquiries

In June 2020 we discovered that nobody was monitoring the company’s Skype, although its goals were firing. People were writing in, and nobody on the other side was looking.

Things like that don’t surface in an audit. They surface where analytics meets the actual sales process.

HARO, with a note on style

We advised writing answers to journalists in a more conversational style. A comment written like a paragraph of technical documentation doesn’t get picked up.

14× across four and a half years.

Geography at the end of the period: US 62%, Italy 22%, then the Philippines, India, Indonesia.

Inquiries. Roughly one contact in six gets qualified. A quarter of all inquiries come from articles — the blog carried its own contact form.

Tellingly, a specialist article brings a specialist client: one contact arrived from the SAP Hybris article, and another is noted in the CRM as potential Hybris outsourcing.

Some contacts get screened out by the client’s team as fake — the company exists but lists a different director, or there’s no company at all. So one qualified in six is the honest conversion rate for B2B outsourcing with a long deal cycle.

Work delivered: 894 tasks, 3,094 hours.

A narrow specialization caps the ceiling

Java, SAP Hybris and Liferay are subjects few people search for. The ones who do know exactly what they need.

So 14× here means more than the same multiple in a broad niche: growth came from covering the stack more deeply, not from widening the subject.

The company was selling something other than what we promoted

For eighteen months the focus was outsourcing. By February 2022 it was clear the company sold outstaffing better and faster.

That isn’t a strategic error — it’s information that only appears once you have inquiry data. But a pivot in year four means redoing priorities, structure and the link plan.

The peak came months before the finish

Autumn 2022, around 2,200; December 2022, around 1,700. We don’t hide the figure.

We chose the direction from the map, not from traffic. Monthly market analysis against the competence map led to the outstaffing pivot and to prioritizing pages that were already producing inquiries.

We answered the market with data, not instinct. March 2020: IDC research, a check with the client, and only then content.

Fifty-five months of distance. Four and a half years without a break, content and links in balance.

Promote what the company knows how to sell. Here the focus moved from outsourcing to outstaffing because outstaffing sold better. Traffic without a sale is an expense.

Build your promotion list from inquiries. Not from traffic potential. The difference becomes visible in year two.

When the market breaks, take analyst data and check it with the client. An article about a growing industry the client can’t serve produces visitors, not inquiries.

Check that somebody reads your inbound channels. A Skype account with live goals that nobody monitored was a real finding on this project.

Sometimes a drop in inquiries isn’t about SEO. And it’s worth telling the client that plainly, even when staying quiet pays better.

Work concluded in December 2022. Later changes to the site were made without our involvement.

Do you have a narrow technical specialization?

An enterprise stack, a limited pool of queries, a long deal cycle. We ran exactly that for four and a half years.

Питання