Less manual work between tools that don’t talk

In some companies, the integration between systems isn’t software. It’s a person. In the morning they pull contacts out of the CRM, in the afternoon they retype the same details into the invoicing software, and at the end of the day they update a spreadsheet so management can see “the real picture”. Nobody designed it this way; it built up over time, with tools chosen at different moments, each solving its own part and none covering the process end to end.

The worst part of working like this rarely shows up in a software proposal. It shows up in day-to-day operations: a tax number corrected in one system and out of date in the others, a customer contacted twice, a document issued with the wrong address, a report that never matches the invoicing figures. When the team is already tired of being the bridge between tools, the request is no longer “we want more technology”. It becomes more direct: we want to stop copying this by hand.

What was happening

In an SME where sales and admin shared the same customers, the day-to-day work was spread across three places. The CRM held opportunities, notes and follow-ups. The invoicing system issued documents and tracked payments. Between the two sat several spreadsheets: one for “to invoice this week”, another to reconcile names and tax numbers, and another for the exceptions neither system handled well.

When a contact’s details changed, the correction had to be made in more than one place. Sometimes it was made in only one, and the rest came to light late, usually at the most inconvenient moment. It wasn’t for lack of effort; it was the price of systems that didn’t talk to each other and of manual work that had become the accepted way of doing things.

What we did first (and what we didn’t)

Before connecting anything, we mapped what actually happened day to day, not the ideal version described in a manual. Who first creates the customer, and where? Which fields are strictly required for invoicing? Who checks the tax number? When an opportunity is marked as won, what has to exist in invoicing, and by when? Where was the spreadsheet still genuinely needed, and where did it exist only because nobody trusted the automated transfer?

That revealed three points where the pain was felt every day: the customer’s tax details, the “ready to invoice” moment, and the payment statuses sales wanted to see without opening another system. It also showed what wasn’t urgent: long histories of notes, elaborate quarterly reports, and marketing automation ideas someone had raised in the first conversation that did nothing for that week’s problem.

Without that map, the temptation would have been to say “let’s integrate the CRM and invoicing” and leave it at that. With the map, the question became concrete: which data flows, from where to where, at what point, and who is responsible when the connection fails.

What we decided to do

Instead of trying to connect everything, we limited the work to three specific changes.

First, the customer data needed for invoicing began to flow automatically from the CRM to invoicing, with clear rules about which system is the source of truth for each field. Second, when a record was marked ready to invoice, that step no longer depended on someone copying the information by hand. Third, the sales team could see whether an invoice had been paid without opening the invoicing software or re-entering the same data elsewhere.

In the same conversation, we set aside three requests that usually come up: syncing every field between the two systems, getting rid of all the spreadsheets at once, and building a large reporting dashboard before the main process was stable. They weren’t bad ideas. They were premature. Widening the scope at the start would have increased the risk and delayed the improvement where the team was already losing time every day.

We didn’t choose the most complete platform. We chose the solution the business could maintain without adding to the confusion.

What changed day to day

Creating and updating tax details stopped depending on a chain of copies. “Ready to invoice” stopped living only in someone’s head or in a row in Excel. Sales could see payments without retyping the same information by hand. The spreadsheets that remain now handle genuine exceptions, not the normal course of work.

There was friction at the start: fields to align, exceptions to write down, someone having to take ownership of the connection when it failed. That doesn’t go away. What stopped being normal was treating manual copying between the CRM and invoicing as the official way of working.

We didn’t invent percentages to prove the result. The useful test is what the team recognises: less time retyping the same data, fewer late corrections to documents, less time comparing lists that should be identical.

What we left out (and why)

We didn’t treat the spreadsheet as the enemy. Where exceptions are rare and call for human judgement, a controlled spreadsheet can be acceptable. The problem wasn’t the spreadsheet; it was using it as the central system because the others didn’t talk to each other.

We also didn’t automate complex exceptions in the first phase. If the team doesn’t yet understand a rule well, codifying it early creates technical debt. It is better to handle those exceptions by hand, against an explicit criterion, and only then decide whether any of them is worth automating.

Nor did we sell this as a complete solution. Processes change and new fields appear. Without someone keeping an eye on it, or a clear handover of responsibility, the connection degrades and the manual work comes back.

If this sounds familiar

If your team is still the integration between the CRM, invoicing and spreadsheets, don’t start with the tool. Start with the process that is slowing work down most this week. Understand where the data comes from, where it has to go, when, and who is responsible if something fails. Connect only what matters. The rest can wait until the essentials are stable.

If you want to talk to us, bring that specific process: where things get copied, where they fail, who gets stuck. The useful conversation starts there.