Finance & administration

Five systems open and it still does not add up: how to get your finance back in line

CRM, webshop, time tracking and bookkeeping run separately and contradict each other. Here is how to work out which number is right and build one system you can actually steer on.

By Ricardo TheijsMay 18, 20267 min read

Short answer. When five systems are open and the numbers contradict each other, the fault rarely lies in one package. The problem is that each system keeps its own version of the truth and nobody has decided which one leads. You solve it by appointing one source system and connecting the rest to it.

I have had plenty of business owners across the table with five tabs open: the CRM, the webshop, the time tracking, the bookkeeping package and an Excel sheet alongside to make it all add up. And still it does not add up. Revenue in the CRM is a different number than in the bookkeeping, the webshop says something else again, and the hours that should have been invoiced sit somewhere in a tool that talks to none of the others. That is not a shortcoming on your part. That is a system that has grown to the point where nobody knows which number leads anymore.

In this post I explain why that happens, how to find out which number actually is right, and how to work towards one place where the truth lives. Not by working harder, but by setting things up differently.

Why your numbers do not line up between systems

The core is simple: the moment a piece of data lives in more than one place without an automatic link, those places start to drift apart. A customer is created in the CRM, then again by hand in Moneybird or e-Boekhouden, an order lands in the webshop but is only booked days later. Every manual handover is a moment where a typo, a forgotten line or a difference in VAT code creeps in. ResalePartners puts it aptly: if systems are not synchronised, each system keeps its own version of reality.

The result is that you do not steer on numbers, but on a feeling. You know the margin sits somewhere around a certain percentage, but you cannot prove it. That is not a detail. The moment you want to raise a price, make an investment or simply know whether this month was a good one, you need a number you dare to lean on.

First work out which number you are allowed to believe

Before I connect or build anything, I always do the same thing first: I work out which system is supposed to be the truth for each type of number. That sounds obvious, but in practice it has never been agreed. The CRM thinks it knows the revenue, the bookkeeping is certain of it, and the webshop keeps a third set of books.

The rule of thumb I follow:

  • Products sold and orders: the webshop or the point-of-sale system is the source, because that is where the transaction happens.
  • What actually comes in and has been booked: the bookkeeping leads, because that is the fiscal reality.
  • Customer and deal data: the CRM (HubSpot, Pipedrive or similar) is the source for who the customer is and what stage a deal is in.
  • Hours worked and billable hours: the time tracking leads up to the moment they are invoiced.

Only once that is clear can you explain differences instead of arguing about them. And often in this step I discover that there are process steps that only still exist because they once started that way. A manual export nobody reads anymore. An interim Excel administration that does double work. I deliberately remove those, because the smartest system is the one that no longer needs a step at all. That is not a reason to build nothing. It is the reason to build the right thing.

One source of truth, the rest connects to it

Once it is clear which system leads for what, the solution becomes clear: you appoint one source per data stream and the other systems pull their data from there. A new customer in the CRM flows automatically through to the bookkeeping. A paid order in the webshop is booked without anyone retyping anything. Hours marked as "invoiced" appear on the invoice. e-Boekhouden describes that this is exactly what a good link does: data is exchanged automatically, which makes manual entry redundant.

One important detail I always pin down early: the unique key by which systems know they are dealing with the same customer. In practice that is often the Chamber of Commerce number. SaaS Insights rightly points this out: without a shared key, the software does not know that the customer in the CRM is the same as the one in the bookkeeping, and then you get duplicate records instead of a clean link.

When a standard integration is enough, and when it is not

Many packages offer ready-made integrations. Moneybird, SnelStart, e-Boekhouden and Twinfield each have a list of them, and for a standard situation those are excellent. My advice there is always honest: if an existing integration covers your stream exactly, use it.

It grates the moment your process does not fit the standard template. You work with your own pricing logic, you have three administrations that need to be consolidated, your time tracking sits in a tool no integration exists for, or you want a report your bookkeeping package simply cannot produce. That is exactly where I build: an intermediate layer that ties the systems together your way, a dashboard that brings the numbers from all sources together in one place, or an automated workflow that does the step no package does for you. The calculation and the flow have to happen. The only question is whether a standard package happens to do it exactly the way you need, or whether it has to be custom.

The difference between those two is where the gain sits. A standard integration that almost fits costs you correction work every month again. A system built for your situation does the work right once and keeps doing it.

Frequently asked questions

Why are my reports wrong when my systems are not synchronised?

Because without a link, each system keeps its own version of reality. An order, customer or invoice lives in multiple places that drift apart through manual entry. As soon as even two systems are out of sync, that difference carries through into every report you pull from them.

How do I know which system is right when the numbers differ?

Appoint one leading system per type of number: the webshop for orders, the bookkeeping for what has been booked, the CRM for customer data, the time tracking for hours worked. Only once that division of roles is fixed can you explain differences instead of endlessly checking back and forth between packages.

How do I avoid having to enter customers twice?

By appointing one system as the source and connecting the others to it via a shared key, usually the Chamber of Commerce number. You then create a new customer in one place, after which it flows through automatically. Test the link with one customer first before you synchronise the whole database.

Can I connect CRM and bookkeeping with a standard package?

Often yes. Moneybird, SnelStart, e-Boekhouden and Twinfield have ready-made integrations that work fine for standard situations. If your process does not quite fit, for example because of your own pricing logic or multiple administrations, then a custom integration is the solution that actually removes the correction work.

Further reading


I am Ricardo Theijs of RNT Projects. With a background in enterprise process management (UWV, Centric, G4S, MSc Business Process Management) I build systems that turn messy operations into something you can steer on again. I am not an accountant and I do not pretend to be. What I do do: pull a process apart, see where it leaks, and build the integration, dashboard or workflow that a standard package just does not deliver.

Running into this yourself?

I review your process and build the solution where a standard package falls short. Remote, with visible results in two weeks.

Let's talk