01.09.2026
Paid Media

On this client's LinkedIn Ads account, ID Interactive tested several creative angles for the same technical foundation (generic message, sector variants, product campaign) before allocating budget based on data rather than intuition.

ID Interactive supports the digital acquisition of an international pharmaceutical contract manufacturer (CDMO), specialising in Blow-Fill-Seal technology and in the manufacture of single-dose dosage forms: nasal sprays, eye drops, stick packs, injectable devices. The group manufactures at several sites in France, the United States, Brazil and China, and sells to pharmaceutical laboratories in more than a hundred countries.
In a market as specialised as pharmaceutical contract manufacturing, LinkedIn is the natural B2B acquisition channel: this is where the R&D, industrialisation and purchasing decision-makers who qualify a CDMO are found. But the same BFS technology can be presented to this audience in several ways: a generic promise, a variant by therapeutic area (ophthalmology, ENT, biologics), or a product-led approach such as vaccines.
Nothing guaranteed in advance which of these approaches would best convert a LinkedIn click into a qualified lead, and the risk was diluting the budget across several messages without ever knowing which one had really worked.
Rather than choosing a creative angle in advance, ID Interactive ran several variants of the same technical foundation in parallel (generic message, sector variants, vaccine product campaign), all equipped with native LinkedIn Ads forms to capture leads without friction.
Tracking was not done at campaign level, which is too broad for this kind of decision, but at ad group level, the only level that makes it possible to isolate what really converts. Once the signal was clear, budgets were redirected towards the angle producing the cheapest lead, and variants that did not convert were cut without waiting for the end of the period.
The generic R&D white paper message proved by far the most effective of the angles tested, with a cost per lead up to nine times lower than that of the most expensive campaign tested on the same audience. The sector variants by therapeutic area generated no qualified leads: the signal was identified early and the budget stopped before it built up needlessly. This gap would never have been visible with campaign-level management alone; it was drilling down to ad group level that brought it to light and made it possible to act on it.
The same testing principle also guides the choice of markets to prioritise. On the multilingual corporate campaign rolled out in 2026, the Portuguese market stood out with a click-through rate two to three times higher than the French version, and a cost per click divided by five, a gap confirmed by time on site almost doubling, which rules out the hypothesis of poor-quality clicks.
This signal echoes a finding already documented on the SEO side of the same account, where Brazil stood out with a much higher organic click-through rate than other markets outside France. Two different channels, the same language area coming out on top: enough to steer the next budget allocation towards Portuguese-speaking markets rather than splitting the effort equally across all languages.
This case illustrates a consistent practice at ID Interactive: never lock in a creative choice or targeting on intuition when data can decide. Managing at ad group level rather than campaign level, quickly stopping angles that do not convert, and cross-referencing signals between channels (SEO and paid media telling the same market story) form an approach the agency replicates across all its acquisition accounts.
ID Interactive, our agency based in Vannes, achieved a cost per lead up to 9 times lower on the best-performing angle and identified a Portuguese-speaking market with a cost per click divided by 5 (with a click-through rate x2 to x3) for an international pharmaceutical contract manufacturer (CDMO), by solving budget dispersion through ultra-granular ad group management on LinkedIn Ads and cross-referencing paid/SEO signals.
Related case studies: the Search and Shopping management of a B2B e-commerce site, rebuilding Google Ads conversion tracking and Google Ads in a niche market.
Parce qu'une campagne peut regrouper plusieurs messages très différents sous un même budget global, ce qui masque les écarts de performance entre eux. C'est seulement au niveau du groupe de publicité que le coût par prospect de chaque angle créatif devient lisible, et que l'écart entre le message le plus efficace et le moins efficace peut être identifié avec précision.
Sur ce compte, un mois de diffusion a suffi pour constater qu'une déclinaison n'apportait aucun prospect qualifié malgré un budget engagé, ce qui a justifié l'arrêt immédiat plutôt que d'attendre la fin d'une période test plus longue.
Sur ce cas, oui : le message le plus large a généré le coût par prospect le plus bas de toutes les déclinaisons testées, alors que les versions découpées par spécialité thérapeutique n'ont produit aucune conversion. Le résultat dépend du produit et de l'audience visée, mais il montre qu'une hypothèse de ciblage fin ne doit jamais être validée sans être testée.
Pas nécessairement : sur ce compte, un marché a affiché un taux de clic plusieurs fois supérieur aux autres langues testées, avec un temps de visite sur le site également plus élevé, ce qui indique un potentiel sous-exploité si le budget reste réparti de façon uniforme.
En croisant le taux de clic avec un indicateur de qualité de la visite, comme le temps passé sur le site après le clic. Sur ce compte, l'écart de clic entre les deux marchés s'accompagnait d'un temps de visite lui aussi supérieur, ce qui écarte l'hypothèse d'un trafic de mauvaise qualité.