Case study: Sigma Software University — copy (EN)

IT education · Ukrainian market

Sigma Software University: 9× organic traffic in two and a half years

Lead: Three and a half years of work, 637 tasks and more than forty documented meetings. A project where the client handed us not just access but an ICP and a tone of voice — and it shows in the result.

• Traffic from ~600 to ~5,600 visits a month at peak
• 145 keywords in the Top 3
• 987 referring domains, +524 in the past year

Small print under the bar:

Ahrefs data, March 2023 to September 2026. Ahrefs estimates traffic from its own index, so traffic figures are an estimate rather than a counter reading; positions are measured directly.

Sigma Software University is Sigma Software’s education arm: courses and programmes for people entering IT or moving up inside it.

The market is Ukraine. Worth saying immediately, because the rest of our portfolio works on the US and Europe. Different competition, different seasonality, and an entirely different audience pattern.

March 2023: roughly 600 visits a month.

Directions were chosen through our WEDGE framework. This project shows it best, because both documents survived — the map and the analysis built on it.

W — Where the client can win

Sigma Software University’s competence map was compiled in October 2024: services, industries, directions inside the education product.

The map comes with its own legend explaining how to fill it in, and the client fills it. We don’t assess how strong the team is at instructional design — the team knows that. We bring market movement and data on who already holds the results pages.

⭐ E — Evaluate whether we can get in

A month after the map came a document titled “Competence map analysis” — step E in its purest form.

What’s in it: directions from the map turned into specific pages, each with a competitor example so the client can check whether they offer something similar. And, separately, an assessment of reachability.

Directions that passed the screen:

And an immediate note on reachability. On How to Measure Training Effectiveness the analysis says plainly: “fairly competitive, links will be needed to promote it.” The direction gets taken, with the understanding that content alone won’t close it.

The direction that didn’t pass

This is the important part, and cases usually leave it out.

For industries in Ukrainian, the tools showed no demand. The analysis puts it this way:

“The tools don’t show demand for equivalent topics in Ukrainian, but it may not be entirely absent — it may be low and therefore hidden.”

The decision wasn’t “we don’t do it” but “we do it paired with the English page” — because low demand and no demand are different things, and the tools can’t tell them apart.

That’s what a second screen looks like when the answer is honest: not “this direction is bad” but “the data is insufficient, so we hedge.”

Then the screen was applied again

The map doesn’t sit dead in a folder. The tracker shows people returning to it:

• November 2024 — map and keyword analysis for Instructional Design
• May 2025 — research into the B2B AI services direction
• June 2025 — traffic and keyword analysis for AI-track lectures
• April 2026 — research into the UpSkilling direction

The queue of directions stays alive: close one, take the next.

D — Decode who decides

The client handed over what clients usually don’t: a dedicated ICP file and a dedicated tone of voice. Most clients say “write it like our site.” Here both the audience and the voice were written down.

Personas don’t sit in a document. The project has two GPT agents: one adjusts the brief to the persona, the other checks finished texts against the personas.

G — Guide them to a decision

The project’s structure was built from competitor analysis and from the keyword analysis of the competence map — that’s how the task is worded.

Each direction got a service page plus blog material covering the earlier stage.

E — Expand the coverage

987 referring domains, +524 in the past year — the profile more than doubled at a mature stage of the project.

Plus meetings every one to two weeks for three years running: more than forty documented meeting notes from 2023 through March 2026.

9× across two and a half years to the peak.

Current snapshot: 389 keywords ranking, 145 in the Top 3, traffic value $1.7K.

Link profile: 987 referring domains, 9.1K backlinks.

Work delivered: 637 tasks, 1,719 hours.

There was a decline after the peak

From roughly 5,600 in mid-2025, traffic fell to ~4,300 by September 2026. We don’t hide the figure — the chart is public.

We’re still establishing the cause. In education both seasonality and changes to search results can be in play.

The Ukrainian market runs on different arithmetic

Smaller query volumes, sharper seasonality tied to the academic year, and far fewer publications available for links than in English-speaking markets. An approach lifted from a Western project won’t transfer.

Directions came from analysing the map, not from taste. Seven directions passed the second screen; one was held back because the data on Ukrainian-language demand was inconclusive. Both documents — the map and the analysis — survived.

Meetings every one to two weeks. More than forty documented sessions across three years.

Distance. 1,719 hours across 3.5 years at an even pace.

Give your agency the ICP and tone of voice if you have them. It’s the cheapest way to shorten the approval cycle.

A regular meeting costs less than correspondence. Every one to two weeks, with decisions recorded.

The Ukrainian market needs its own approach. Different seasonality, different volumes, fewer placements.

The project is ongoing. Current focus: establishing why traffic came off the 2025 peak, and restoring the trend.

Sound familiar?

An education product, the Ukrainian market, a need for steady registrations.

Питання