Back in 2014, a buddy and I opened a brewery called Moody Ales in Port Moody, a suburb of Vancouver. Like a lot of small businesses, we were invoice-driven, and getting paid was a grind.
Someone on our team would sit there, open QuickBooks, look at the invoice, go to the card terminal, type in the credit card number, process it, reconcile it back into the accounting system and send out a receipt. That took two days a week, sometimes three. About twenty percent of the time, the card declined. And we had one to two months of cash tied up in invoices.
When the COVID lockdowns shut our breweries, I had some time, so I dusted off my dev skills and wrote a script to automate it. We recaptured about two months of cash really quickly, and that team member got to spend their time on more valuable work.
None of those numbers were hidden. Two to three days a week. One to two months of cash. But until you actually look at them, it’s just “how we get paid.” Once you do look at them, the decision gets a lot easier.
That’s what good KPIs are for: a few numbers that make the right call obvious.
You don’t need 50 KPIs
MSPs generate a tonne of data: tickets, time entries, agreements, invoices, payments, pipeline. The problem is rarely a lack of data. It’s that the information is scattered.
A useful dashboard answers four questions:
- Are we growing?
- Are we profitable?
- Are we keeping our customers?
- Are we generating cash?
If your KPI dashboard needs three monitors and a data scientist, it’s probably gone too far.
The core financial KPIs
Revenue growth. Is the business expanding or contracting? Watch the trend, not one month.
Gross margin. What’s left after the direct cost of delivering the work. For most MSPs that’s largely labour, so this tells you whether your services are priced and delivered properly.
Net profit margin. What’s left after everything else. Revenue is just one part of the number. What you spend is the other side of the equation.
Days Sales Outstanding (DSO). Roughly how long it takes an invoice to become cash. Those one to two months of cash sitting in invoices at the brewery? That’s DSO. It’s money you’ve earned and can’t use.
Cash reserves and runway. How long you could keep going if revenue or collections slowed down. Your bank balance won’t move in step with your P&L. Something like a laptop purchase hits cash all at once.
Gross revenue retention (GRR). Take the customers you had a year ago, ignore everyone you’ve added since, and ask how much of their revenue is still coming in. Losses and downgrades count. Upsells don’t. When we ran a June-to-June cohort on our own business, the result wasn’t horrible, but it wasn’t amazing either. It’s actually painful to look at a group of customers from twelve months ago and count how many are gone.
Net revenue retention (NRR). Same cohort, but now expansion counts too. If NRR is higher than GRR, your existing customers are growing with you.
MSP-specific metrics that explain profitability
The core numbers tell you the score. These tell you where it’s coming from:
- Agreement gross margin
- Effective hourly rate
- Technician utilization and labour mix
- Recurring revenue per employee
- Customer profitability
- Accounts receivable aging
- Autopay adoption
Revenue per employee is one I think about a lot. Almost all of our revenue goes to pay our people, and then we pay for everything else on top of that. If our average sale price goes up, we can hit our revenue targets with fewer customers and fewer hires. Headcount shouldn’t have to grow in a straight line with revenue.
Outcome metrics versus driver metrics
Outcome metrics tell you what happened: revenue, gross margin, net profit, cash, DSO. Driver metrics help explain why: close rate, churn, agreement margin, utilization, revenue per employee.
If gross margin is dropping, that’s the outcome. The cause might be poor utilization, unprofitable agreements or rising labour costs. Knowing the result without understanding the cause doesn’t tell you what to fix.
Churn is a good example. Some customers leave because they sold the business or switched accounting systems. Some just go quiet, and if they won’t tell you why, they’ve probably moved to a competitor. And some leave because they assume your product still only does what it did on day one, so they go find another product that does what you already do. Three different causes, three different fixes.
Outcome metrics tell you what happened. Driver metrics tell you what to change.
Your data already exists – it’s just scattered
- Accounting system: revenue, margins, expenses, cash flow
- PSA: time, utilization, agreement and customer profitability
- CRM: pipeline and conversion
- Payroll: labour costs and headcount
- AR system: DSO, collections, autopay
When those systems don’t agree, you find out the hard way. We had a customer whose price in our CRM didn’t match what our billing system was charging. The billing system doesn’t care what the CRM says.
You probably already have most of the data. It may just be living in five systems and one employee’s head. A spreadsheet is a perfectly good place to start pulling it together. A consistently measured imperfect number is usually more useful than a perfect number nobody can produce.
Build a financial operating rhythm
Numbers only help if you look at them on a schedule and act on them. Keep it simple:
- Compare actual results with the plan.
- Find the variances that matter. If our hiring plan ends the year off by a few grand, that’s fine. Off by thirty grand isn’t.
- Decide whether it’s temporary or structural. A one-month spike is not the same as a trend.
- Assign someone to fix it.
- Check whether the fix actually worked.
And write down why. When we make a deliberate call that moves a number the wrong way, I want it documented, so when I’m talking to the board or investors I can say, “Here’s exactly what happened, and we did it on purpose.”
Reconcile, too. Reconciling your payment clearing account against the bank every month is really the only way to catch some data entry errors before they turn into bad numbers.
Measure fewer things, understand them better
At the brewery, the fix didn’t start with a dashboard. It started with two numbers we finally stopped ignoring: the days every week spent processing payments, and the cash sitting in invoices. Once we looked at them, we knew exactly what to fix.
That’s the point of all this. Not more reports. Better decisions, sooner. Measure fewer things, understand them better, and use what they tell you.
