Growth outcome
Piktalent
Building an organic acquisition channel across six languages.
Build the content first, translate it last
- 598 → 69,905organic sessions a month, April 2023 to October 2024GA4 · team outcome
- 496,833organic sessions in 2024GA4 · team outcome
- 76%of measured 2024 organic sessions landed on country pagesGA4 · team outcome
- 46 → 1,016new student sign-ups a month in the first six monthsTeam tracker, not GA4
- Challenge
A new platform, starting from almost zero.
A new platform with almost no organic visibility, competing for students and companies across Europe, in six languages.
- My contribution
Build the pages people search for.
I led the structure of the country and field landing pages, the website UX and conversion paths for students and companies, technical, on-page and content SEO, and the GA4 and KPI reporting behind it.
- Delivery and measurement
Measured · GA4, 2023–2024
Organic Search grew from 598 sessions in April 2023 to 69,905 in October 2024 (GA4). 2024 closed at 496,833 organic sessions. New student sign-ups rose from 46 to 1,016 a month in the first six months, and held between 1,150 and 1,780 a month through 2025.
What this doesn't prove
Team outcome. I led web and SEO; product, sales and operations sat with other teams. The sign-up figures come from the team's internal tracker, not GA4. Organic traffic later declined through 2025 and 2026; part of that decline is where the international SEO investigation begins.
The full curve, including the turn.
April 2023: 598 · October 2024: 69,905 · December 2024: 50,406
Lower in November and December · all 24 months shown
View monthly data (24 months)
| Month | Organic sessions | Comparison point |
|---|---|---|
| January 2023 | 128 | — |
| February 2023 | 476 | — |
| March 2023 | 524 | — |
| April 2023 | 598 | April baseline |
| May 2023 | 2,528 | — |
| June 2023 | 4,146 | — |
| July 2023 | 7,981 | — |
| August 2023 | 10,272 | — |
| September 2023 | 13,021 | — |
| October 2023 | 16,909 | — |
| November 2023 | 18,198 | — |
| December 2023 | 17,392 | — |
| January 2024 | 26,636 | — |
| February 2024 | 21,264 | — |
| March 2024 | 21,140 | — |
| April 2024 | 20,526 | — |
| May 2024 | 35,351 | — |
| June 2024 | 39,438 | — |
| July 2024 | 43,104 | — |
| August 2024 | 45,796 | — |
| September 2024 | 60,136 | — |
| October 2024 | 69,905 | October peak |
| November 2024 | 63,131 | — |
| December 2024 | 50,406 | End of recorded series |
How to read it. GA4 Organic Search sessions, January 2023–December 2024. April 2023 is the selected baseline and October 2024 is the maximum within this series; the line continues through the lower November and December values.
Chart limits
The April baseline is a selected comparison point, not the start of the dataset. The October peak is followed by lower November and December values. All twelve months of 2024 remain visible; the chart does not stop at the peak.
No intervention dates are marked because the approved record does not tie particular changes to particular months. The curve records what happened to traffic; it does not identify the cause of each rise or fall.
Web, search and the paths between them.
A platform across six languages.
How do you turn a new website into a channel that brings in students and companies every month?
The commercial context was an Erasmus+ mobility and internship platform serving students and companies across Europe. The documented remit connects country and field landing-page structure, search acquisition, website UX and conversion paths, with GA4 and KPI reporting.
International landing-page architecture
Structure of country and field landing pages on a platform serving students and companies in six languages.
Technical and on-page SEO
Leadership of technical and on-page SEO alongside the website structure.
Content and search acquisition
Content SEO within the wider web and search remit.
Student and company paths
Website UX and conversion paths for both audiences.
Analytics and reporting
GA4 and KPI reporting behind the website and SEO work.
This is a responsibility map, not a causal model. It does not show that each workstream independently caused the growth. Product, sales and operations sat with other teams.
What the remit record does not show
The breadth of the role is documented; individual technical decisions and their measured effects are not. The approved account does not identify specific template rules, SEO fixes, UX experiments or content-production methods for this period.
Build the content first, translate it last
- In mid-2023, Piktalent's English country and service pages were still being built, and there was pressure to start multilingual early so a French-language site could move across.
- Start translating as soon as Spain was done, or finish and review the English pages first.
- I recommended finishing and reviewing the English pages before translating anything. The business needed Spain in French sooner, so that part moved ahead; I held further Spain redirects and kept building the English country, legislation and service pages.
- A Spanish translation had already been wasted when the content changed, and every later edit would have to be made in every language.
- Multilingual work continued through 2024: French pages were being drafted in June, and in October Spanish was close to finished while the other languages were in review. The record I have shows drafting and review; it does not show publication dates.
Two sources, different questions.
GA4 Organic Search sessions
496,833sessions · full year 2024
The monthly series records sessions classified as Organic Search, January 2023–December 2024. It measures traffic, not unique students, registrations or revenue.
Internal student tracker
46 → 1,016new students / month · first six months
Monthly new student sign-ups come from the team’s internal tracker, not GA4. Their acquisition-channel attribution is unknown. They cannot be treated as organic conversions or divided by these sessions to calculate a conversion rate.
Registrations add a commercially relevant measure of student acquisition alongside search traffic. They do not establish paid placements, company acquisition, sales or revenue. The two sources describe team outcomes; they do not isolate my contribution from the work of other teams.
What the measurements leave open.
Sessions and sign-ups are different measures from different sources, and neither isolates one workstream. The fuller record is below.
Full evidence boundaries
The public evidence does not specify the GA4 reporting configuration, changes to tracking, or the tracker’s registration definition and deduplication rules. There is no channel-level registration breakdown or controlled comparison linking a workstream to an outcome.
The approved account reports a later traffic decline through 2025 and 2026, but this monthly traffic dataset ends in December 2024. The 2025 registration range is a separate tracker observation, not a later-period organic traffic series. A complete monthly 2025–26 series is not supplied: three later observations (January 2025, August 2025 and July 2026) are shown under Later context, and no recovery is quantified.
The August 2026 anchor-profile analysis below is later context. It is not evidence for the cause of the 2023–24 growth, the post-peak fall or the subsequent decline.
- Diagnosed — Done
- Built — Done
- Measured — GA4 and internal KPI tracker
The decline started before the 2026 fixes
- Monthly organic sessions reached their published maximum of 69,905 in October 2024. They were 58,130 in January 2025, fell to 28,710 by August, and kept falling through 2026 (8,563 in July), before the technical investigation began.
- Present the migration work as the recovery story, or report the timeline as it is.
- Report the timeline as it is.
- The decline began in 2025, before the 2026 migration and indexing work. Claiming that work caused it, or reversed it, would be wrong.
- GA4 and Search Console both show the contraction. Search Console omits anonymised queries, so query counts are directional.
What the web still called it /
The repositioning said Erasmus+ mobility and internship provider. The web still used the old name: in the owned anchor profile (31 Aug 2026), 141 referring domains linked to Piktalent as "europe-internship.com", against 63 as "piktalent".
What the brand says
Piktalent
same entity, old name
What the web says
europe-internship.com141
piktalent63
First-seen link dates are not migration dates.
I led work across search, website structure and user journeys, kept acquisition and registration reporting distinct, and can say what the data does not prove. Traffic later declined through 2025 and 2026, and that decline is where the international SEO investigation begins.
Reflection
A channel isn't built once. It has to keep being found.