The usual configuration in a project company: tasks in a tracker, money in accounting, and management reporting in a spreadsheet the owner maintains in the evenings. Each part works. Together they fail to answer the central question of the business: is this project making money or taking it?

The cause isn’t the quality of the tools — it’s that the question sits precisely on the seam. The work is in the tracker, the money is in accounting, and profit is born where they meet. If they never meet, profit stays a feeling.

The four gaps

1. The work is done, but what it cost is unknown

The tracker knows the task is closed. It doesn’t know what that cost the company, because it doesn’t connect hours to rates. Accounting knows about salaries but not which project they went to.

So project cost is computed by hand, once a quarter, approximately. And pricing decisions are made by eye.

2. The money arrived, but what for is unclear

Accounting shows a client payment. Which stage it closes, what’s still unpaid, whether there’s overdue receivable on this project — all of that has to be worked out separately.

It’s especially painful with staged payments: a deposit, an interim sign-off, a final settlement. Without a link to tasks, it’s unclear whether you’ve even reached the stage that’s already been paid for.

3. The project goes into the red and you find out after delivery

The most expensive gap. Hours burn as the project runs, but the comparison against budget happens afterwards — when nothing can be done.

If the number appeared at the moment 60% of the hour budget was spent at 40% completion, there’d be a choice: cut scope, negotiate an increase, move work to a cheaper resource. After delivery there’s no choice.

4. Decisions are made blind

Questions a project business answers by instinct rather than by number:

  • Which clients are genuinely profitable?
  • Should we take this job at this price?
  • Who should we hire next?
  • Which service should we push?
  • Whose price should go up, and who should we part ways with?

All five need the chain “work × cost × revenue”. No tracker and no accounting system answers them alone.

What needs connecting

The minimum loop that makes the picture complete:

Project → hours → rates → cost. Every task belongs to a project, every person has an hourly cost. Cost then computes itself.

Project → contract → payments → receivable. The contract amount, what’s been collected, what’s still outstanding. The rule is simple: collected + receivable = contract. If that equation doesn’t balance, a stage got lost somewhere.

Project → direct costs. Contractors, purchases, fees — anything that wouldn’t exist without the project.

Result: margin = contract − labour cost − direct costs. The method in detail is in how to calculate project margin.

Note what’s absent here: full accounting. The management loop and the accounting loop are different things. Accounting answers to the tax authority; management reporting answers to you. The second belongs in the system where the work lives; leave the first with your accountant.

Why integration rarely solves it

The logical idea is to wire the tracker to the accounting system with a connector. In practice such links live badly:

  • The entities don’t match. The tracker has projects; accounting has contracts and invoices. One-to-one correspondence is rare.
  • There’s nowhere to get hours from. Accounting doesn’t record who worked how long on what. Rates are a management figure too, not an accounting one.
  • The connector needs maintaining. It acquires an owner, a cost and a failure mode. A team under 20 usually has no such person.
  • The data drifts. Once there are two sources, reconciliation begins — and it consumes more time than the manual calculation it was meant to replace.

The practical conclusion: for a small company it’s cheaper when management reporting lives in the same system as the work, rather than being wired to it.

Pricing stops being guesswork. When you can see the cost of similar past projects, an estimate is built from fact rather than feeling.

The conversation about raising a client’s price becomes possible. “Our costs went up” is a weak argument. “This project takes 40% more hours than last year, here’s the breakdown” is a strong one.

You can see which clients feed you. Usually 20% deliver 80% of margin. Nobody knows that before the calculation — and afterwards the team’s priorities shift.

An early signal appears. A project burning budget faster than it progresses becomes visible before delivery.

Hiring stops being frightening. You can tell whether another person pays for themselves, because revenue per head and utilisation are known.

Where to start if everything is separate today

You don’t need to build it all at once. The order is:

1. Set hourly rates for your people. Loaded ones: salary + payroll taxes + equipment, divided by billable hours (60–75% of calendar), not by 160.

2. Start recording hours per project. A weekly estimate is enough — 10–20% accuracy suffices for management decisions. How to do that without resistance is in tracking time without micromanagement.

3. Record the contract and payments on the project. The contract amount and what’s collected. Receivable appears by itself as the difference.

4. Compute your last two completed projects. They’ll show whether the method works on your data.

5. Only then automate. If the loop holds together in a spreadsheet, it moves into a system easily. If it doesn’t, automation merely speeds up the delivery of wrong numbers.

Bottom line

A tracker without money answers “what are we doing”. Accounting answers “how much money moved”. Neither answers “are we making anything on this” — and that’s the question that decides which projects to take, whose price to raise and who to hire.

For a project business, connecting tasks and money isn’t a convenience — it’s a precondition for being in control. And it’s simpler when they start out in one system than when they’re stitched together by an integration.

In Gosudarynya tasks, clients and per-project finance share one database: hours attach to tasks, rates to people, and margin computes as you go. 7 days free, no card.