Tech scale-up: recruitment against uncontrolled Time-to-Skill

DRAFT — to be replaced by author content

Context

The terrain: a French technology scale-up in the cloud infrastructure sector, seven hundred employees in May 2026, founded in 2017, with a Series C closed in autumn 2024 for an amount near one hundred and fifty million euros. The company, present mainly in France but with a team of about one hundred people based in Central Europe, builds a distributed infrastructure management platform for large enterprises and software publishers.

The growth plan validated by the board provided, over the eighteen months following the funding round, the doubling of engineering headcount — moving from three hundred to six hundred people. Six months after the plan started, the HR department alerted the executive committee on a significant gap between actual recruitment and planned recruitment.

Problem

In-depth analysis revealed that the problem was not, as initially supposed, a sourcing capacity problem. Candidate flows were sufficient; conversions were mediocre. A finer examination brought to light three superimposed causes.

The first cause concerned the effective Time-to-Skill of new joiners. The average delay between starting in the role and operational autonomy on the company’s proprietary platform reached ten months — whereas sector benchmarks placed this delay between five and seven months. During this learning period, new joiners weighed on payroll without producing the expected value, and mobilized experienced employees for their training, degrading the latter’s productivity.

The second cause concerned attrition. The annual rate measured on senior engineering profiles reached 28%, against a sector median of 18%. A systematized exit survey identified as the first motive not compensation — comparatively competitive — but the feeling of a poorly designed work organization, with a cognitive load considered excessive and insufficient visibility on evolution trajectories.

The third cause concerned payroll. The conjunction of long Time-to-Skill and high attrition produced a ratchet effect: to maintain production capacity, more recruitment than planned was required, which aggravated pressure on payroll and compromised the profitability horizon targeted by the financing plan.

Intervention

The intervention was designed as a restructuring of the talent chain, articulated around four coordinated moves.

The first move, over two months, consisted in instrumenting Time-to-Skill measurement by technical scope, distinguishing platform components whose learning was critical from those for which autonomy could be reached more quickly. This measurement allowed identifying two particularly time- consuming components, for which internal documentation was notoriously insufficient.

The second move, over four months, consisted in investing massively in the documentation and learning tooling of these two critical components. The investment, valued in engineer-hours diverted from production, represented the equivalent of eight person-months — a significant cost for a team under tension, but lower by a factor of three than the cost of avoided attrition.

The third move, over six months, consisted in overhauling the evolution trajectories proposed to senior engineers. So-called vertical trajectories (toward management) now coexist with formalized expert trajectories recognized — including in terms of compensation — at the same level as managerial trajectories. This overhaul, negotiated with employee representatives, relied on a Pay-for-Agility grid targeted at engineers fastest to upskill on the new platform components.

The fourth move, more continuous, consisted in regaining control of the payroll trajectory by instrumenting a monthly dashboard shared between HR, finance and engineering.

Outcome

At the end of twelve months, the measured indicators were the following. The average Time-to-Skill on the platform had been brought from ten to six months, an improvement of 40%. The annual attrition rate on senior engineering profiles had come back down to 19%, returning to the sector median. Payroll had been brought back to the trajectory provided by the financing plan, with one month ahead of the objective revised mid-year.

The engineering headcount doubling plan was revised downward — moving from six hundred to five hundred and twenty — productivity per head having sufficiently increased to allow reaching production targets with a smaller workforce.

Lessons

Three lessons were drawn from the experience by the HR director, in a feedback presented to the board.

The first lesson concerns diagnosis. In a scale-up in hyper-growth, the reflex is to attribute any failure to a lack of resources. This attribution is rarely accurate: failures almost always come from debt accumulated on recruitment process quality, on internal documentation or on trajectory thinking. Resolution comes through investment, not through increasing means with unchanged process.

The second lesson concerns coherence between compensation and trajectory. As long as the expert trajectory is not recognized at parity with the managerial one, the most experienced engineers will continue to leave for other companies more mature on this point. Parity must be visible, formalized and honored by effective promotions.

The third lesson concerns rhythm. The described transformation took twelve months of intensive work. No acceleration would have been possible without degrading arbitration quality. Holding this rhythm, in the context of a scale-up under pressure from impatient investors, is a governance act that exceeds the sole HR perimeter.