How to Improve Delivery Margins in a Professional Services Business
A services business can grow its way into worse economics. Revenue rises, headcount rises with it, and somewhere in between the gross margin on delivery quietly gives up a point or two a year, which nobody catches in any single month because every decision that caused it looked reasonable at the time. If you are a finance director, CFO or managing director asking how to improve delivery margins in a professional services business, the honest starting point is that margin rarely leaks in one place; it leaks in four, and each leak hides behind a plausible explanation.
This article covers those four leaks, the fixes that actually hold, and the metrics worth watching weekly, and it ends with a margin leak audit you can run on your three largest engagements this quarter without hiring anyone.
Where delivery margin actually leaks
Scope creep, delivered out of goodwill
The most common leak is work that was never sold. A client asks for something adjacent to the statement of work, the delivery team says yes because saying yes preserves the relationship, and the extra effort is absorbed rather than invoiced. Each individual concession is small, but across a six month engagement the concessions compound into a project that is delivering ten or fifteen per cent more work than it is billing for, and because nobody logs the concessions, nobody can see the total.
The tell is a project that finishes on time, keeps the client happy and still comes in below its planned margin, which the team explains as the estimate having been tight rather than the scope having grown.
Unpriced complexity
The second leak happens before delivery starts. The deal was priced for the straightforward version of the work, but the engagement that arrives has three more stakeholders, a legacy integration nobody mentioned in the sales cycle, and an approval process that adds a week to every milestone. The price was set by the deal; the cost is set by the delivery. When sales and delivery estimate independently, the gap between the two is paid for out of margin, and it is paid quietly, engagement by engagement.
Poor utilisation visibility
Most firms at 20 to 60 people know their utilisation monthly and in aggregate, which is precisely the resolution at which the problem becomes invisible. An average of 75 per cent can hide a senior consultant at 40 per cent doing internal work, a mid-level team running hot at 95 per cent and heading for burnout, and a new hire who has been on the bench inside a project code for six weeks. Averages hide the shape, and the shape is where the money is. A related and equally quiet leak is seniority drift, where senior people end up doing work that was priced at a junior day rate because the junior capacity was never hired or never trained.
Hero-driven delivery
The fourth leak is structural. If your margins depend on two or three people who rescue every difficult engagement, you are running a delivery model that cannot be priced, because the cost of the rescue never appears in any plan. Heroes work evenings that are not recorded, compress estimates because they know they can absorb the overrun personally, and become the constraint on how much work the firm can take. It is the delivery version of the founder dependency problem, and it caps both margin and growth at whatever those individuals can personally carry.
Rule of thumb: if you cannot state the actual margin of your three largest engagements without a spreadsheet exercise, the margin problem is a visibility problem before it is a pricing problem.
Fixes that hold
Put a threshold on free work. The fix for scope creep is not a rigid change control process that annoys clients, but a pre-agreed threshold: anything under a defined effort level, the delivery lead can absorb and must log; anything over it triggers a commercial conversation before the work starts. Logging is the crucial half, because the log turns invisible generosity into visible cost, and visible cost is negotiable at renewal even when it was given away in-flight.
Let delivery price complexity before the deal closes. Any engagement above a sensible size should get a short pre-sale review from the person who will actually run it, with delivery signing off the estimate rather than inheriting it. This costs a few hours per deal and removes the structural gap between what was sold and what must be built. Where sales resists, the resistance is usually a sign that growth has outpaced the operating model more broadly, not just in pricing.
See utilisation by person, by week. Move from a monthly average to a weekly view that separates billable work, investment work and bench, per person. You do not need expensive tooling to do this; you need a consistent definition of billable and a fifteen minute weekly review where someone with authority looks at the shape rather than the average.
Make heroics the exception. Standardise how engagements are run, with a common structure, common checkpoints and a project-level margin view that the delivery lead owns, so that a difficult project is caught at week three rather than rescued at week eleven. The aim is not to eliminate your best people's judgement but to stop the business from silently pricing it at zero.
The metrics to track weekly
| Metric | What it tells you |
|---|---|
| Margin by engagement, planned versus current forecast | Whether estimates survive contact with delivery, and which projects are drifting |
| Utilisation by person and seniority band | Where capacity is hidden, who is overloaded, and whether senior people are doing junior work |
| Logged unbilled effort per engagement | The real size of scope creep, in days rather than anecdotes |
| Write-offs and credits | Where quality or expectation problems are being settled with margin |
| Scope changes raised versus scope changes charged | Whether your change threshold is being used or bypassed |
Weekly matters. A monthly view tells you what happened; a weekly view lets you intervene while the engagement can still be corrected, and the discipline of the cadence does as much for margin as any individual number.
Run a margin leak audit this quarter
Take your three largest current engagements and, for each one, work through five questions with the delivery lead and whoever owns the number.
- Rebuild the true cost of delivery to date, including every hour worked by every person, at loaded cost rather than day rate, and including the senior time spent reviewing, rescuing and reassuring.
- Compare the margin as sold with the margin as currently delivered, and write down the gap as a number rather than an impression.
- List every piece of work delivered that was not in the statement of work, however small, and total the days.
- Compare the planned seniority mix with the actual one, because a project staffed one level heavier than it was priced loses margin even when it runs perfectly.
- Ask who the engagement could not survive without, and what that dependency is costing in unrecorded effort and in work you cannot take on.
The gaps will cluster. In most firms two of the four leaks account for most of the erosion, and the audit tells you which two, which means your fix list is short: not a 40-point transformation plan but three or four changes that, if they hold, materially improve the margin on everything you sell from here.
Price is set once, at the start; margin is set every week, by how the work is run.
Where Vitori fits
Everything above can be done internally, and in a firm with a strong finance function and a delivery leader with capacity, it should be. The difficulty is rarely knowing what to do; it is holding the changes in place while the business keeps selling and delivering, which is where good intentions usually give way to the next busy quarter.
Vitori works with founder-led technology services businesses whose growth has outpaced their operating model, using the Operational Scale Framework to assess where Delivery and Operations are leaking margin and, through the Operator model, embedding to implement the fixes rather than leaving a report behind. Engagements are fixed-term and outcome-based, and we stay accountable until the changes hold. Whether you do this with Vitori or anyone else, the goal is the same: delivery economics that are visible weekly, priced honestly and independent of heroics, in a business that runs, and scales, without the founder in every decision.