How an Agency Moved Net Margin From 18% to 41% at Flat Revenue

A modelled services business moves from 18% to 41% net margin at flat revenue by contracting scope and repricing twice — the exact arithmetic.

Productised Agency Model

Modelled teardown — 14-person agency, no external capital

Sector
Digital services
Location
Bengaluru, Karnataka
Annual revenue

₹2,40,00,000

Deliberately flat across 30 months. Growth was refused in order to prove margin alone could move.

Retainer clients
34 → 30
Productive hours sold
16,520 → 14,160
Utilisation
71% → 84%
Realisation / hour
₹1,453 → ₹1,695
Fixed monthly costs

₹14,80,000

14 people at a fully-loaded ₹1,04,600 a month, plus ₹1,36,000 of rent, tooling and admin.

Delivery team (11)
₹11,00,000
Leadership (3)
₹3,64,000
Rent and facilities
₹72,000
Tooling and admin
₹64,000
Net margin

41.0%

Up from 18.0% at identical revenue — ₹98,40,000 a year instead of ₹43,20,000.

Starting net margin
18.0%
Scope leakage recovered
₹33,60,000
Repricing gain
₹28,80,000
Bench cost avoided
₹8,40,000
Delivery savings
₹9,60,000
Client acquisition payback

2.4 months

₹1,25,000 to win a retainer averaging ₹52,000 a month at a 38% contribution margin.

Cost to win a retainer
₹1,25,000
Average retainer
₹52,000 / mo
Contribution margin
38%
Payback
2.4 months

This is a modelled teardown. A 14-person digital agency at ₹2,40,00,000 of annual revenue runs an 18% net margin — ₹43,20,000. Thirty months later, at exactly the same revenue, the margin is 41% and the business keeps ₹98,40,000. No new clients were signed after month 12 and no headcount was added. The entire change is pricing and scope.

The mechanism is straightforward once you see it in hours. At ₹2.4 crore of revenue and 14 people billing a realistic 1,180 productive hours each, the agency sells 16,520 hours at an effective realisation of ₹1,453 per hour. Its fully-loaded cost per productive hour — salary, overhead, bench, tooling, management — is ₹1,191.28. The 18% margin is what survives after unbilled scope eats the difference.

Where the money actually leaked

Three leaks, in order of size. Scope leakage — work delivered but never billed — accounted for 14% of productive hours, or 2,313 hours a year, worth ₹33,60,000 at realisation. Unbilled bench time between engagements accounted for 9% of capacity. And the retainer book was priced on a 2019 rate card, so 61% of recurring revenue was below the current cost of delivery. Only the third leak showed up in any report the founders read.

LineAnnual% of revenue
Revenue₹2,40,00,000100%
Starting net margin (18%)₹43,20,00018.0%
Scope leakage recovered₹33,60,00014.0%
Retainer repricing (+22%)₹28,80,00012.0%
Bench cost avoided₹14,40,0006.0%
Productised delivery savings₹−9,60,000-4.0%
Revenue forgone on 4 lost accounts₹−6,00,000-2.5%
Ending net margin (41%)₹98,40,00041.0%
Margin bridge at constant revenue. Each line is a distinct operational change, measured over 30 months.

Leak one: scope written into the retainer

The agency sold “a monthly design retainer” with no deliverable cap. Every retainer grew 20–40% in scope within three months of signing, because there was no boundary to push against and no contractual language that defined what was included. The fix was to publish three named packages with a stated deliverable count, a stated revision count and an explicit hourly rate for anything beyond them. Scope leakage fell from 14% to 3% within two quarters, entirely through the act of writing the boundary down.

Clients do not usually exploit open-ended retainers maliciously. They expand into available capacity because nothing tells them the capacity is finite. A deliverable count is a communication device as much as a pricing one.

Leak two: the 2019 rate card

Of 34 retainer clients, 21 were paying rates set before the agency's cost base rose 38%. Repricing them by an average of 22% in two waves — 11% with 90 days notice, then 10% nine months later — lost four clients representing ₹6,00,000 of annual revenue, and gained ₹28,80,000 from the remaining 17. The ₹22,80,000 net gain was deliberately given back by not replacing the churned accounts, which is why the model shows constant revenue and a much higher margin.

Leak three: bench and utilisation

Utilisation was 71% across the team, with the gap almost entirely bench time between projects. Productising the two most-requested services into fixed-scope, fixed-price engagements with a standard delivery sequence raised utilisation to 84% and cut delivery hours per engagement by 19%. The same revenue was delivered with fewer hours, which is the only sustainable way to lift margin without raising prices.

MetricBeforeAfter
Revenue₹2.40 Cr₹2.40 Cr
Productive hours sold16,52014,160
Utilisation71%84%
Effective realisation / hour₹1,453₹1,695
Net margin18%41%
Before and after. Utilisation is the lever most agencies ignore because it is invisible in the P&L until it becomes overtime or churn.

What productising does not fix

It does not fix client concentration. The 41% margin business still has 38% of revenue in three accounts, and a single loss would take the margin back below 25%. It does not fix founder dependency either — the two relationships that generate the largest retainers are still personally held. Margin is a necessary condition for a durable services business, not a sufficient one.

  • Publish a deliverable count in every retainer, or the retainer will expand into your margin.
  • Reprice in waves with 90 days notice; a single 22% increase loses more clients than two 11% increases.
  • Measure utilisation monthly. Anything below 75% in a services business is a pricing problem, not a sales problem.
  • Productise only the services you have delivered more than ten times. Standardising a bespoke service destroys the reason clients chose you.
  • Accept that repricing will lose clients. Model the loss before you send the email, so the outcome is a decision rather than a shock.

What founders ask about this teardown

What is a good net margin for a services agency?

15–20% is typical for an unscoped retainer business, and 35–45% is achievable once delivery is productised and scope is contracted. Agency benchmarks above 45% usually reflect a niche with pricing power rather than superior operations.

How do you stop scope creep without annoying clients?

Put a deliverable and revision count in the retainer and publish the hourly rate for anything beyond it. Clients rarely object to a boundary they can see; they object to a surprise invoice for a boundary that was never stated.

Is productising a service business worth it?

Yes, if you have delivered the same service more than ten times. Standardising it cuts delivery hours and raises utilisation. Productising a genuinely bespoke service strips out the differentiation clients are paying for.

How much can you raise retainer prices in one year?

10–12% per wave with 90 days notice is generally absorbed; a single 20–25% increase typically triggers client churn. Two waves nine months apart achieve the same end state with meaningfully lower attrition.

What utilisation rate should an agency target?

80–85% of productive hours for a delivery team who are staffed continuously. Sustained 90%+ causes burnout, and below 75% the business is subsidising bench time out of margin.

Every figure in this teardown is reproducible. Substitute your numbers and see where the path breaks.