A study of the finance departments of ten of the Netherlands' largest e-commerce companies, read from their own job posts. Not what they say, who they hire.
You don't have to wait for an AI strategy to see where a finance department is heading. You can already read it from the people they hire. This study reads those signals, and sketches the finance you can build with them.
Everyone is talking about AI in finance. Almost no one about the question underneath: are your numbers right enough to let a machine loose on them?
In thirteen years of accountancy, first at Deloitte, then at the growth companies we now serve at Actuals, I have seen one pattern come back again and again. The ambition runs ahead of the foundation. And that is exactly where the promise falls apart.
So we didn't look at what companies say about AI, but at something more honest: their job posts. A finance department getting ready for the AI era is already hiring the people for it. Whoever doesn't hire those people isn't building the foundation, however hard the board hammers on AI.
This report shows what we saw at ten of the Netherlands' largest e-commerce companies, and what it means for you. Not to call chains to account, but to show where the market stands, and where you can take a step. Because you don't buy the finance of tomorrow. You hire it.
The future first, not the fear. What AI can make of the finance function, and why almost every CFO wants to go that way.
For the first time, the department that always reported after the fact can look ahead. Close in real time. Forecast in seconds. Ask questions of the ledger as if it were a colleague who knows everything and never gets tired.
This doesn't replace the CFO, it enlarges them. And the numbers show that almost everyone wants to go that way.
Because those same studies show a second pattern. Where AI in finance falls apart, it is almost never the model. It is the data underneath. And that is what this study is about: not the ambition, but the foundation that has to make that ambition real.
of CFOs call AI extremely or very important for finance in 2026.
of finance functions already used AI in 2024, and adoption is growing.
prioritize AI integration, while they struggle with trust.
of AI projects without AI-ready data will be scrapped before 2026.
name data quality as the biggest blocker to AI success.
are concerned about data security and privacy with AI.
Read them as one story. The will is there for almost everyone. The brake is not in the technology, but in the data underneath: quality, readiness, trust. The company that invests there now won't be buying AI later. It will be picking it.
You are not your tooling. You are the person the board takes at their word, because your numbers are already right before anyone asks. That is what it's about for you. Not AI. The authority underneath it.
Think of the moment an investor looks at you and you don't have to search for a second. Of the close that is over before your team has given up an evening for it. Of the CEO's question you answer while they are still asking it. That is not a different person. That is you, without the noise that still holds you back.
AI promises you exactly that version. But AI on top of numbers that aren't right doesn't make you stronger; it magnifies your biggest doubt, and puts it on a dashboard. So the question is not what AI can do. The question is whether you become the one who dares to build on it.
The hours that now go to reconciliation go to the question behind the number.
Numbers that are right before anyone asks. The board takes you at your word.
A close that no longer surprises, because the truth is already there.
Same promise, different drive. One wants to be seen strategically, another wants to be untouchable when the auditor calls, a third wants to scale without burning out their team. On the next page is yours.
A CFO is not a job but an identity. These are the profiles we see most often in high-volume environments, and what a reliable data foundation with AI unlocks for each of them.
Profiles from the Actuals CFO audience analysis based on behavioral profiling (Chase Hughes). A real CFO is usually a mix of a primary and a secondary type.
Before we zoom in on the ten companies, the picture at market level. Dutch e-commerce runs on enormous transaction volumes, multiple PSPs and international flows. Exactly the environment where AI on finance promises the most, and breaks the hardest without a foundation. What we see in the job posts is a market that embraces the promise but doesn't yet staff the foundation.
Millions of micro-transactions a day, but the data layer under finance is rarely staffed.
Adyen, Mollie, ERP and data warehouse side by side. The truth lives in four places at once.
The market hires administration and control, not the engineers who build the future.
Why that better version of you doesn't exist without a foundation that holds, and what becomes possible once it's there.
AI is an amplifier. Give it reconciled data and it magnifies your best judgment. Give it fragmented data and it magnifies your mistakes, with the same conviction.
Validated, reconciled and booked before the ledger. Truth Inside.
As long as four systems give four numbers, every AI output stays an opinion with a dashboard around it. Not because the model is weak, but because the source is. That is why there is no endpoint where you are "ready for AI" as long as that one source is missing: the order is the reverse of what the hype suggests. Truth first, cleverness only after.
Not doom-mongering, but an honest map. Every hurdle can be cleared. Together they decide whether AI on your numbers is something to build on or something to be wary of.
Adyen, Stripe and Mollie side by side, each with its own payout rhythm and currency. Three truths about the same money.
Flows across borders and entities. One intercompany error and the consolidated number is fiction.
Hundreds of thousands of orders a day. One percent unmatched is no longer noise, but a gap that AI carries through.
Truth that only emerges at the end of the month. Steering mid-quarter on numbers that aren't final yet is gambling.
As long as four systems give four numbers, there is no truth to put AI on. This is the hurdle that makes the other four irrelevant the moment you clear it.
Get the foundation right, and the promise from page four simply becomes your workday. Not someday, but daily.
This is not a wish. It is what remains once you remove the noise. The rest of this study is about who is building that foundation now, and who is not yet.
There is a gap between the board's AI ambition and the data that ambition has to run on. That gap has a name: reconciliation. It is the invisible work that decides whether a transaction you see is also the transaction that actually happened. As long as that gap exists, every AI output is a gamble with authority.
of AI projects without AI-ready data will be scrapped before 2026. Not because the models fail, but because the data underneath can't be trusted.
Payouts, fees and chargebacks come in separately, on a different rhythm than the order.
Whatever doesn't match gets corrected by hand, under time pressure, at the end of the month.
One percent unmatched across hundreds of thousands of orders is not noise. It is a structural gap.
We read the live job posts of ten large Dutch e-commerce companies. Not what they say. Who they hire.
The future of a finance department is written in its job posts. A company building its data foundation hires the people for it. On 11 June 2026 we opened the live finance careers page of each company and placed the open roles into one of three groups.
Open finance roles that build data and systems inside finance: finance data, finance systems, automation, reconciliation at scale.
Open finance roles that run the existing machine: administration, classic control, credit and procurement.
Almost no finance hiring. No signal that the finance function is being built, in any direction at all.
A snapshot of public job posts on a single day. A directional signal about where a company is investing right now, not a verdict on the whole finance department. A company can have a strong base and simply not be hiring. That is why every card shows the source and the date.
Of these ten, two (bol and Booking) are now hiring for people who bring data and systems into the finance function. The other eight hire to run the present, or barely hire at all. No shame in that, it's a snapshot. But a lead is building now, quietly, in the job posts, long before it becomes visible in the numbers.
One overview of all ten companies and the six role types that matter now. Green is an open job post of that type, read on 11 June and 10 July 2026. The top three rows build the future, the bottom three run the present.
| Role type | bol | Coolblue | Picnic | Belsimpel | Wehkamp | Booking | Takeaway | AH | HEMA | Rituals |
|---|---|---|---|---|---|---|---|---|---|---|
| Finance systemsdata engineering in finance | ||||||||||
| Data analysis in financedata that feeds finance | ||||||||||
| Automationmanual work out of the close | ||||||||||
| Reconciliation & paymentsmatching at scale | ||||||||||
| Classic controlcontroller, credit, capex | ||||||||||
| Administration & paymentkeeping the machine running |
Read the matrix vertically and the story is there: the top three rows, where the future is built, stay largely empty. Only bol and Booking turn green where data and systems go into the finance function. The rest hire in the bottom rows, today's machine.
Per company: what you see in the job posts, what that could mean, and which people would lift it to the next version.
The controllers bol is now looking for carry data and reporting in their title. And there's a data analyst open specifically for finance, separate from central BI. Finance is getting its own data capacity here.
Exactly the move AI-readiness demands: not buying AI, but staffing the data layer under finance. Controllers tied to tech and data domains suggest finance moves with the product. Bol is building the bridge between accounting and data, and that is what AI will run across.
"Data and reporting" says something about the reporting side, not automatically about the source underneath. The question: is the single source of truth under that reporting secured, or is there smart reporting on data that still comes from four systems?
Lock down the reconciliation under the polished reporting just as seriously as the reporting itself.
Turn the match rate across PSPs and entities into a hard, daily number.
An own pipeline that feeds the ledger, not a lean on central BI.
Bol has the scale, the volume and now also the first people to build the finance of tomorrow. If the underlying reconciliation is taken as seriously as the reporting on top of it, this is the kind of company that doesn't have to fear AI on its numbers but can embrace it.
Profile fit: the route of the Strategist and the Scaling CFO. Whoever builds this way earns the position of the board's counterpart and the calm to keep growing without burning out the team.
Seven finance job posts, almost all administration and payment processing. A strong, mature operation. One strategic-analytical role is the exception.
Coolblue runs a tight machine. But the hiring is aimed at processing volume, not at building the data layer underneath. A retailer of 2.46 billion euros staffs its finance with administration team leads, not with finance engineers. The capacity to make the numbers AI-ready is not being built internally right now.
Coolblue has strong data and tech teams elsewhere. The signal here is specifically about the finance function: there it's being managed now, not built.
Forge the payment and order flows that are now managed separately into one source.
A reconciliation lead who turns the tight operation into a measurable daily number.
One system owner above administration who lays the foundation for AI.
Precisely because the operation here is so tight, the leap is close. Coolblue has the discipline and the volume; what's missing is the data layer that turns that administration into a source of truth. One well-placed finance engineer makes the difference between a department that runs the present and one that steers the future.
Profile fit: the comfort zone of the Control CFO. The leap takes the nerve to go from managing to building, and that is where the biggest gain lies.
Three finance roles, all three classic: capital control, fixed-asset accounting and a junior associate. No data or systems role in finance.
Picnic is famous for building its own technology, from logistics to robotics. All the more striking that finance is hired classically. The engineering strength sits in the operation, not yet in the numbers. A missed opportunity and a huge reserve: few companies can modernize finance as fast as Picnic, if it turns that strength inward.
Data roles at Picnic may fall under Analytics rather than Finance. For the finance function itself: here the work is classic bookkeeping now, not building.
SQL and Python and accounting, as a bridge to its own engineering culture.
A data engineer who translates the logistics data strength to the ledger.
Forecast on reconciled input instead of on last month's numbers.
No company here has a stronger engineering culture than Picnic. If it brings the same rigor it applies to logistics into finance, in two years one of the most advanced finance functions in the Netherlands stands here. The question is not whether they can, but whether they want to prioritize it.
Profile fit: par excellence the domain of the Builder CFO. Whoever builds here creates something that wasn't there yet, and is seen for it too.
Belsimpel is building data capacity, but on the commercial side. The finance hiring itself stays at traineeship level.
The data ambition is there, and that's more than many companies can say. But it lands on the commercial side, not in finance. The finance function is mostly fed with young talent: good for the future, but not yet senior capacity that lays a data foundation. The ambition is waking up, the foundation under finance is not yet.
Half a billion in revenue from a fast-growing, telecom-driven model means a lot of transaction volume. The junior intake can be a deliberate talent pool, but it doesn't yet cover the senior role that builds the foundation.
Put an experienced layer above the traineeship intake that lays the foundation.
A data analyst specifically for finance, not just for the commercial side.
Bring the growing transaction volume back to one truth before the growth tears it apart.
Belsimpel shows that it dares to invest in data. The leap here is smaller than it seems: the art is to tilt that same data strength from the commercial side to the finance side, and put a senior layer on top that turns it into a source of truth.
Profile fit: a growth story for the Scaling CFO. Laying the foundation now prevents the growth from tearing the department apart later.
One open finance role, and it's procurement. No controllers, no data, no systems.
A department running on its current headcount and not expanding, in line with revenue that has shrunk in recent years. Whoever consolidates doesn't hire. But no capacity is being built now for the next phase. If the market runs on AI-driven finance later, Wehkamp starts further back than the rest.
In a downturn, not hiring is a rational choice. The signal here is not about skill, but about direction: there is no building ahead right now.
Bring the fragmented administration to one truth, before you hire new people.
Take the close off the shoulders of the current, scarce headcount.
Someone who steers on that reliable source, instead of rebuilding it every month.
For a company consolidating, the route is not a large build program, but one targeted move: bring the fragmented administration to one source, so the scarce capacity is spent on insight instead of on reconciling. Precisely in a downturn, a truth that holds is the difference between steering and gambling.
Profile fit: the Control CFO who seeks certainty. A source that holds gives that certainty faster than a bigger team.
Between the classic controllers and tax roles, Booking is hiring something you see almost nowhere else: a product manager for AI in finance, and a specialist who builds the finance systems (SAP). Here the data layer under finance is treated as a discipline of its own.
This is the rare move AI-readiness truly demands: finance data and finance systems as a discipline of their own, not as a byproduct of central IT. An explicit AI-in-finance role means someone becomes owner of how the numbers go into the models.
Booking's finance is global and enormous; these build roles are a minority in a large apparatus, spread across several countries. The signal is the direction, not the scale: they put in place the people the rest still lack.
An AI-in-finance role is only as strong as the reconciled data underneath it. Secure the single source of truth first.
Don't let the SAP specialist only report, but own the matching across entities and currencies.
That way the build capacity feeds a truth that holds, instead of a quick correction round.
Booking is the only one of the new five with the people who literally build the finance of tomorrow. The question is not whether they want to, but whether the source under all that modeling is taken as seriously as the model itself. If so, this is a department that doesn't fear AI on its numbers but steers it.
Profile fit: the route of the Strategist and the Builder CFO. Whoever lays the source under the AI becomes the board's counterpart instead of a supplier of numbers.
Seventeen finance roles, spread worldwide. Almost all classic: business partners, accountants, controllers and consolidation. The exception is in the systems: an Anaplan specialist for the planning tooling and R2R roles that hammer on automation.
Takeaway invests in planning and process systems (Anaplan, R2R automation), and that's more than most. But it builds the tooling above the numbers, not the data layer underneath. The truth AI would run on is managed and consolidated here, not rebuilt as a source.
Finance is organized globally (Amsterdam, Madrid, Milan, UK, Tel Aviv). The NL roles are business partner and consolidation: steering the existing machine, not building a new source.
Anaplan is only as good as its input. Feed it reconciled data, not exports.
Take out the manual work before the ledger, not only at the close.
Millions of orders a day call for matching at scale as a hard daily number.
Takeaway has the discipline and the systems (Anaplan, R2R) that many companies lack. The leap is closer than for most: the tooling is in place, what's missing is the reconciled source underneath. Whoever adds it turns planning from managing into steering on truth.
Profile fit: the comfort zone of the Control CFO. The leap takes the step from managing to building, and that is where the biggest gain lies.
On Albert Heijn's own careers platform there is one finance role at head office: a business controller for e-commerce. Classic control, focused on the online business. The rest of finance sits centrally at parent company Ahold Delhaize.
A business controller specifically for e-commerce shows that the online growth is being steered. But it is control on the existing machine, not a person building the data layer underneath. The capacity to make the numbers AI-ready is not being built visibly at AH level here.
AH's finance is largely centralized at Ahold Delhaize (controllers Franchise, Opex, Business Services Finance). The AH signal is therefore thin; it says something about the brand's job posts, not about the full group finance.
Bring store and online flows to one truth before you steer on it.
Enormous transaction volumes call for matching as a hard daily number.
An own data capacity that feeds the ledger, close to the online business.
Albert Heijn has the volume and the data ambition of a retailer, but the finance hiring at brand level is thin and classic. The leap lies less with AH itself than with the group: whoever secures reconciliation and the source at group level lifts the online business along with it.
Profile fit: a growth story for the Scaling CFO. Laying the foundation now keeps the online growth manageable.
Five finance roles at the office, all on the management side: a risk & internal audit specialist, a senior payroll administrator and classic administration. No data, systems or reconciliation role in finance.
HEMA hires finance to keep the machine running and compliant: audit, risk, payroll. For a retailer in recovery that is a rational choice, but no capacity is being built now for the data layer that AI on the numbers demands.
Five roles is more than some others, but the direction is managing, not building. HEMA has data and tech teams elsewhere; the signal is specifically about the finance function.
Bring the fragmented administration to one truth.
Take the close off the shoulders of the existing headcount.
Someone who steers on that reliable source instead of rebuilding it every month.
For an omnichannel retailer in recovery, the route is not a large build program, but one targeted move: bring the administration to one source, so the scarce capacity is spent on insight instead of on reconciling.
Profile fit: the comfort zone of the Control CFO. Precisely in recovery, a truth that holds is the difference between steering and gambling.
Finance hiring is thin: what's visible is mainly one senior business controller (omni, wholesale and travel). Rituals does build strongly on tech and data (GenAI, data engineering), but on the product side, not inside finance.
Rituals invests visibly in data and engineering, but that strength lands in the product and commerce, not in finance. The finance function itself is hired classically. The leap is smaller than it seems: the data strength only needs to tilt toward finance.
The broad Tech & Data teams (GenAI, data engineering) say something about Rituals as a whole, not about finance. The signal here is specifically about the finance function: there it's being managed now, not built.
Put a data analyst specifically for finance, not only for product and commerce.
Retail, wholesale, travel and online into one truth.
Matching across entities and currencies as a hard daily number, now the brand is scaling internationally.
Rituals shows that it dares to invest in data. The leap here is smaller than it seems: the art is to tilt that same data strength from product and commerce toward finance, and put a senior layer on top that turns it into a source of truth.
Profile fit: a growth story for the Scaling CFO. Whoever tilts the data strength toward finance keeps the international growth manageable.
Ten companies, one yardstick. Across the profiles, four lessons take shape that apply to the whole market, and to your own department.
Everyone wants AI on their numbers. But of the ten, two (bol and Booking) are visibly building the data layer that makes it possible.
The lead doesn't emerge in press releases, but in job posts, long before it becomes visible in the numbers.
Precisely those with the strongest engineering culture (Picnic) still hire finance classically. That is where the biggest untapped lever lies.
Not a single company hires explicitly for automation in finance. The most-discussed promise is the least staffed.
Look back at the matrix and you don't just see who hires where. Above all you see the empty spots. And those empty spots are the opportunity.
An empty quadrant is not a shortfall. It is an open goal. The department that now hires the people who belong here builds a lead that can no longer be caught later.
A checklist to measure yourself, and the people you hire to get there.
No score needed. Check along. Whoever ticks four or five boxes here can put AI on their numbers with confidence. Whoever gets stuck knows exactly where to start.
Every box that's missing is a role you haven't hired yet. They're on the next page.
Of these six future roles, three are being hired now, and always by the same two companies (bol and Booking). The other eight companies are building none of them. That is where the difference lies.
Owns the data pipeline inside finance and turns separate sources into one truth.
Secures matching across PSPs and entities, and makes the match rate a hard number.
Builds the pipelines that feed the ledger, not central BI.
Speaks SQL and Python and accounting, a bridge between both worlds.
Forecast on reconciled input instead of on last month.
Takes manual work out of the close so the team works on the question behind it.
Hiring counts based on the live job posts from the Finance-Future Matrix, read on 11 June and 10 July 2026.
You don't need a report to read your own department. The same yardstick we laid on ten companies, you lay on your own in ten minutes.
Which roles are open now? Write down the titles, exactly as they stand. Not what you want to hire, what you're hiring now.
Do they fall in the top three rows (data, systems, automation, reconciliation) or the bottom ones (control, administration)? Are you building the future, or running the present?
Which of the six future roles is missing? That's the role you'd hire first. Start at the source: reconciliation and systems.
This study reads ten companies from the outside, from their job posts. For your own company we look from the inside.
The High Volume Accounting Scan shows in fifteen minutes where your foundation stands, independent of the year-end close, and where the first gains are on the way to the finance of tomorrow. Not a ranking, a yardstick.
Take the scanactuals.online/trust/
Would you rather have an in-depth report for your own company, with your own PSP mix and entities built in? Request it, and we'll walk through it together.
Every classification comes from the company's live finance careers page, read on 11 June 2026 (first five) or 10 July 2026 (new five). The market figures come from published CFO and data research. The CFO profiles come from the Actuals CFO audience analysis based on behavioral profiling.
bol careers.bol.com/nl/vacatures
Coolblue coolblue.nl/vacatures/finance
Picnic jobs.picnic.app/en/people-finance
Belsimpel werkenbijbelsimpel.nl/vacatures
Wehkamp wehkampretailgroup.nl/vacatures/finance-vacatures
Booking.com jobs.booking.com/booking/jobs
Takeaway.com careers.justeattakeaway.com
Albert Heijn werk.ah.nl/vacatures
HEMA jobs.hema.com/nl/vacatures/kantoor
Rituals careers.rituals.com
Revenue and key figures (new five) Booking Holdings $23.7bn (2024) · Just Eat Takeaway €5.1bn, GTV €26.3bn (2024) · HEMA €2.2bn gross (2024) · Rituals €2.1bn (2024) · Albert Heijn 37.7% market share NL (2024, Ahold Delhaize)
PSP verification Adyen confirmed for Booking.com, Takeaway, Albert Heijn and Rituals (Adyen cases and press releases). HEMA: not publicly confirmed.
AI crucial for finance · adoption Gartner Finance AI Survey, 2025
60% scrapped without AI-ready data Gartner, Feb 2025
43% data quality biggest blocker Informatica CDO Insights, 2025
76% concerned about data security CFO AI survey, 2025
Methodological note: the classification measures what ten companies publicly hire for, as a directional signal about who is building the finance foundation now. It is not an audit of their actual finance stack. A company can have a strong foundation and simply not be hiring.
This report rests on two kinds of sources. We keep them deliberately separate, so you know exactly what is an observation and what is a published figure.
The future-readiness, the matrix and the role mix come from the live finance careers pages of the ten companies, read on 11 June and 10 July 2026. Repeatable: whoever reads the same pages on the same day arrives at the same classification. A snapshot, not an audit of the whole finance stack.
The percentages on AI ambition and data readiness come from published CFO and data research (Gartner, Informatica, CFO surveys 2025). The CFO profiles come from the Actuals CFO audience analysis based on behavioral profiling.
| Company | bol | Coolblue | Picnic | Belsimpel | Wehkamp | Booking | Takeaway | AH | HEMA | Rituals |
|---|---|---|---|---|---|---|---|---|---|---|
| Data certainty | 5/5 | 5/5 | 4/5 | 3/5 | 3/5 | 5/5 | 4/5 | 4/5 | 3/5 | 3/5 |
Lower with thin public data (few open finance roles), higher with rich, unambiguous careers pages.
Actuals is the financial truth layer for high-volume companies. We validate, reconcile and book every transaction before it hits the books, so that AI on your numbers becomes something to build on.
Actuals Research · Dutch e-commerce · June–July 2026. Based on publicly available information and intended as a conversation piece, not as advice about a specific company. Truth Inside.