Monday, 9am. The leadership meeting opens with revenue for the month. Sales reads a figure from the CRM. Finance has a lower one, because two deals marked as won haven’t been invoiced yet. Delivery points out that one of those projects hasn’t even been set up. Twenty minutes later nobody has decided anything, and everyone turns to you to say which number is right.
You probably do know which one is right. That’s the problem.
How does a business end up with five tools doing the job of one?
One sensible purchase at a time. Sales needed a pipeline, so someone chose a CRM. Delivery needed to track work, so a project tool arrived. Finance already had its accounting software. Then came a time tracker, a shared drive of spreadsheets, a chat tool and a forms app somebody signed up for on a free trial. Each choice made sense on the day. Nobody was asked how they’d fit together, because nobody owned the whole picture.
At 15 people the gaps get filled from the founder’s memory. Somewhere past 30, the gaps are where work goes missing.
What does it look like day to day?
A composite of what Mallard sees in founder-led scale-ups looks something like this.
- The Monday numbers argument. Three systems give three figures for the same thing, and the first half of the leadership meeting goes on reconciling them instead of deciding anything.
- The onboarding checklist that lives in three places. There’s a version in the project tool and an older one in a shared document. The real one is in the head of the account manager who onboards most new clients. New starters follow whichever they find first.
- The client who gets two different answers. They ask how much budget is left. The account manager quotes the project tool. A week later the invoice, built from the time tracker, says something else. From then on the client checks everything you send.
- The side spreadsheet. Someone senior keeps their own tracker “just in case”, because they don’t trust the systems. It’s often the most accurate record in the business, and nobody else can see it.
Each of these looks like a software problem. Usually they’re one structural condition – duplicated work, dropped handovers and constant messages with no shared context – turning up in several places at once.
What does it cost?
The subscriptions are the cheap part.
- Decisions slow down. Every decision needs a debate about the data before it can be a debate about the choice.
- Work gets done twice. People re-key information between systems and chase colleagues for updates that should already be visible.
- You become the integration layer. You’re the one who knows the CRM is out of date and the real status is in someone’s inbox. So the questions come to you, and the business can’t run without you in the middle.
That last cost matters most to a founder trying to step back. Every “which number is right?” that lands on your desk is another small piece of founder dependency.
Why won’t a new tool fix it?
A new platform will faithfully reproduce whatever confusion already exists about who does what. If sales and delivery don’t agree what a handover includes, moving both teams into the same software just moves the argument to a new screen.
Mallard’s Simon Wakeman puts it bluntly in his Build newsletter, where he wrote that “the software tool is not the transformation”. What changes things is behaviour. Who owns each piece of data? Who keeps it current? Does the leadership team actually work from one version in the room? Good software helps once those habits exist.
This is also why Mallard’s model puts people, practices and technology in that order. There’s more on how the three fit together in our guide to building a business operating system.
Which system should hold the truth?
Pick one home for each core fact, and make everything else either read from it or stop holding a copy. For many founder-led scale-ups, a first version looks something like this:
| Core fact | System of record | Owner |
|---|---|---|
| Who the client is and what they’ve bought | CRM | Head of sales or accounts |
| What work is in flight, and its status | Project tool | Delivery lead |
| Time spent on each client | Time tracker | Operations or delivery |
| Money invoiced, received and owed | Finance system | Finance lead |
Your table will look different, and that’s fine. What matters is that everyone in the Monday meeting knows where each answer lives and who to ask when it looks wrong.
Where do you start?
You can do most of this in an afternoon with your leadership team.
- Write your own system-of-record table. List the facts the business argues about most – revenue, pipeline, project status, hours, cash – and name one system and one owner for each. Where the honest answer is “it depends”, you’ve found the gap.
- Map one workflow end to end. Take a client who signed recently and trace them from first enquiry to first invoice. Note every place the same information was typed in again and every point where it changed system. That map shows you which connection to fix first.
- Retire one copy. Pick the duplicate that causes the most arguments, often the onboarding checklist or a side spreadsheet. Archive it with the owner’s agreement and point everyone to the surviving version.
- Run next Monday’s meeting from the agreed sources only. If a number doesn’t come from its system of record, it doesn’t get discussed. It will feel awkward for a few weeks, and it’s the quickest way to find out which systems aren’t being kept up to date.
Hold off on buying anything new until it has an owner and a clear place on that workflow map. Retiring two overlapping tools and connecting the rest properly often does more than a big migration.
When is it worth getting help?
When the table won’t come together, or the workflow map shows problems with ownership rather than software. Disconnected tools are usually a symptom of a business operating system that grew piece by piece, so the fix sits across people and practices as well as technology.
The Find diagnostic looks at 262 operational data points across your people, practices and technology in 14 days, including how your tools and data fit together. It ends with a prioritised roadmap of what to fix first, rather than a shopping list of software.
Contact us