Werner Jacobs

The through-line

What connects everything Werner Jacobs says is a demand for explainability. Whether the subject is a booking that an auditor will question, a decision a board is about to take, or a technology everyone is excited about, his test is the same: can you see inside it, and can you account for what it did. This is why he draws the line in AI adoption exactly where the black box begins, insisting "Je weet wat je erin steekt, je weet wat eruit komt, maar niet wat de blackbox doet exact. Dus de core zou ik heel voorzichtig mee omspringen om daar AI te gebruiken" [2], and why he treats transparency toward the board as the precondition for every good decision rather than one input among many [2].

Against that conservatism sits a personal willingness to be thrown into water he cannot yet stand up in. "Ge moet soms durven springen en dan zwemmen, letterlijk. Ik heb echt een tijd echt gezwommen" [2]. The two impulses are not in tension in his telling: you can take personal risk with your own competence precisely because the structures around you, the analytical accounting, the reporting to the board, the auditability of the ledger, are built so that nothing is hidden.

On the university as a miniature society

Jacobs frames the finance job at a university as governing a settlement rather than running a company. "Een universiteit is eigenlijk een groot dorp. Wij hebben met 6000 werknemers, eigenlijk zijn wij een verkleinde vorm van heel de samenleving" [2]. The population inside that village is genuinely heterogeneous, philosophers alongside supercomputer scientists, lab animals alongside oncology research, and the fiscal consequences follow directly from that heterogeneity: multiple VAT regimes, five different personnel statutes, 110 nationalities and roughly 2,000 unpaid collaborators, all sitting on top of some 5,000 active projects [2]. The complexity is not an accident of bad design; it is what happens when one legal entity has to contain every activity society contains.

On keeping oversight inside one legal entity

His answer to that complexity is structural rather than heroic. The university runs an extensive analytical accounting structure organised into six clusters with separate P&Ls, covering base operations, projects, three institutes that each have their own management agreements, and student services, with a large volume of internal recharges moving between them, all still within a single legal entity [2]. The point of the architecture is that the different economies inside the village can be read separately, so that no cluster's performance disappears into the aggregate.

On AI in the periphery, not the core

Jacobs is deliberately unimpressed by the noise around AI. "We moeten wel doorkijken voorbij de hype. Het wordt nu echt gehyped, AI, overal lees je maar" [2]. His division is practical: apply AI in the periphery of finance, in variance analysis within controlling and in procurement, where a suggestion can be inspected and overruled, and stay very cautious in core transactional finance, where every booking has to be explainable to an auditor [2]. For the core he prefers transparent RPA, precisely because its logic can be traced [2]. The same caution shapes his view of technology cycles generally: "Soms is beter van first follower te zijn dan first mover" [1], and companies still have to depreciate the technology investments they already made before jumping into the next cycle [1].

On researchers who become CEOs

On IP valorisation his position is that the technical brilliance which produces a spin-off is not the skill set that runs one. "Het is niet omdat je gewoon goede onderzoeker bent dat je daarna nog goede CEO of manager bent van je eigen spin-off" [2]. Rather than treat that as a reason to hold back, finance takes an active coaching role: researchers are experts in their own field but usually have no finance background, so the function helps them build business plans and supplies management support around them [2].

On transparency toward the board

For Jacobs the most crucial precondition for good decisions is transparency toward the board, and the failure mode he warns about is not deceit but omission. Where transparency is lacking, even unintentionally, decisions end up poorly grounded in the information that was actually available, and this holds equally in an SME and in a listed company [2].

On competing and collaborating at once

Flemish universities, in his description, operate as "con-collega's": they compete within a defined competitive landscape while collaborating structurally, through shared professorships and through a Flemish interuniversity finance council that exchanges experience and formulates joint positions toward governments [2]. Rivalry at the level of students and research does not preclude a common front on finance and policy.

On promoting people before they are ready

He treats his own early promotion as a deliberate instrument of culture change rather than a reward. Belgacom consciously pushed young people into senior roles, making him finance director at 25, as a policy designed to break the old RTT mentality and force a shift in culture [2]. The honest description of what that felt like is the swimming line: you jump, and then for a while you really are swimming [2].

Takeaways

• Put AI where its output can be challenged, in controlling variance analysis and procurement, and keep transparent RPA in core transactional finance, because bookings must be explainable to an auditor and AI is a black box [2].
• Look past the hype cycle and consider being a first follower rather than a first mover, not least because previous technology investments still have to be depreciated [1][2].
• A single legal entity can hold enormous heterogeneity if the analytical accounting is built for it: six clusters, separate P&Ls per activity type, and disciplined internal recharges [2].
• Fiscal complexity at a university is driven by structural facts, multiple VAT regimes, five personnel statutes, 110 nationalities, around 2,000 unpaid collaborators and 5,000 active projects, and has to be managed rather than simplified away [2].
• Spin-offs need finance to coach them on business plans and management, because research excellence does not transfer into being a good CEO [2].
• Treat transparency toward the board as the first condition of good decision-making; unintentional gaps in information damage decisions just as much as deliberate ones, in SMEs and listed companies alike [2].
• Competitors in the same sector can still run shared structures, as the Flemish universities do with shared professorships and a joint interuniversity finance council that takes common positions toward governments [2].