The Three Gaps: Why Service Businesses Earn Less Than They Should

The question nobody can answer

“I know we're leaving money on the table. I can't tell you where.”

Every service business leader has said some version of this. The numbers confirm it. In 2024, professional services EBITDA margins compressed to 9.8% — the lowest in five years — while revenue growth slowed to 4.6%. Billable utilisation fell to 68.9%, below the 75% operational threshold. On-time delivery dropped to 73.4%. Only 34% of projects stayed on budget.

These are not the numbers of an industry doing well. And yet most firms can't explain wherethe money goes. They know the gap exists. They can't point to it.

The standard response is “better reporting.” More dashboards, more KPIs, faster month-end close. But 70% of CFOs still rely on Excel for planning and forecasting. 88% of those spreadsheets contain errors. The reports arrive, but they arrive late, they measure the wrong things, and nobody has a structured process to act on what they find.

The problem is structural. It cannot be closed by working harder or hiring better people. It can only be closed by seeing it.

Why this isn't a reporting problem

There is a critical distinction between measuring accurately and measuring the right things. Most service firms have mastered the first and completely missed the second.

Your ERP knows that money was lost, but lacks the project context to explain why. Your PSA knows that a project is over budget on hours, but lacks the cost data to calculate the financial impact. Your CRM knows the pipeline, but not whether winning that deal will make or lose you money.

These systems were designed for accuracy and compliance — different objectives from decision speed. They work. They just don't worktogether, and they don't work fast enough.

The result is a three-stage failure that we call The Three Gaps.

The Signal Gap

Not knowing which economic signals matter, so reporting is never built to surface them.

The Mechanism

You cannot define signals you haven't yet learned to watch for. Every firm builds reporting based on past failures. The next failure comes through a gap the reporting wasn't designed to catch. This is self-perpetuating — each post-mortem closes one gap and leaves the next one open.

What the Data Shows

A complete service economics signal landscape contains 12–15 distinct economic signals across four clusters: revenue signals, delivery cost signals, quality and risk signals, and decision lag signals. The average firm actively monitors 4–5. That means 60–70% of the early warning system is dark.

The consequences are measurable. 40–42% of organisations experience revenue leakage. 15% of chargeable consulting work is never billed. MSPs leak roughly 10% of revenue to billing errors alone — about $68,000 per month in missed charges. Cumulative revenue leakage across billing failures and collection shortfalls reaches 22%. And 58.7% of firms cite scope creep as their top project challenge, eroding 5–20% of engagement margin.

The root cause is the same: the signals that would have caught these problems early were never defined, so the reporting systems were never built to surface them.

Meanwhile, an entire category of cost signal has emerged that most firms aren't watching at all: AI delivery costs. Teams use AI daily but the economics are invisible — buried in “software” or “overhead” line items with no attribution to specific engagements. When firms have tried to offer “unlimited AI” to clients, the result is predictable: 1–2 power users can destroy unit economics, and nobody sees it coming because AI cost isn't defined as a signal.

Why It Persists

Signal definition requires prior knowledge of what can go wrong. In a complex service business, the failure modes are numerous and not all obvious. Without a complete map of the signal landscape, firms build partial reporting systems that give false confidence. Only 28% of firms allocate indirect expenses accurately. Fewer than 13% use activity-based costing. The rest are “peanut-buttering” costs across the portfolio — and their most profitable clients are subsidising their worst relationships without anyone knowing.

The Latency Gap

Even when watching the right signals, data arrives too slowly to act.

The Mechanism

Delivery, finance, and reporting run on different organisational cadences owned by different functions. Nobody is accountable for the gap between them. The architecture was designed for accuracy and compliance, not decision speed — these are different objectives.

What the Data Shows

The median month-end close takes 6 business days. Top performers close in under 3 days. Laggards take 10–14 days. Finance teams spend 25 hours per week on manual data entry and reconciliation — time that adds latency, not insight.

But month-end close is only the beginning of the lag. After the books close, reports must be generated, reviewed, and distributed. Variances must be investigated. By the time a project manager sees that an engagement is bleeding margin, the average visibility lag is 4–6 weeks from when the problem actually formed.

The cost scales linearly with time. A problem leaking margin at the rate of a few thousand per week costs one amount when caught in week 2 — and costs three to six times more when it surfaces in a monthly report at week 6. A $15 per hour drop in realised rate for a 20-person team translates to over $100,000 lost in a single quarter. Revenue per consultant has already dropped to $199,000 in 2024. There is no margin buffer left to absorb weeks of invisible leakage.

Why It Persists

Monthly reporting cadences are embedded in accounting practice, management convention, and organisational rhythms. Changing them requires changing how three separate functions operate simultaneously. No single person owns the lag. High-performing firms review project margins weekly. Average firms do it monthly. That difference — weekly versus monthly — is the latency gap in its simplest form.

The Decision Gap

Even with complete, timely signal, no structured mechanism exists to turn signal into governed action in the same week.

The Mechanism

Service businesses have governance processes designed for planned decisions — quarterly reviews, annual pricing cycles — not reactive ones. When an engagement drifts, there is no structured process to act this week. Better data without a better decision process produces faster awareness of problems that still don't get fixed in time.

What the Data Shows

50% of executives estimate that slow decisions cost their organisations at least 4% of topline revenue annually. 49% of C-suite leaders report missing a major market opportunity in the last 12 months due to decision latency. 56% report being beaten to market by competitors who decided faster.

The data quality problem compounds this. 58% of business leaders admit they use inaccurate data for major decisions. Over 80% rely on outdated information. 89% of CFOs make decisions based on inaccurate or incomplete data. Nearly half of all business decisions are made on gut feel rather than evidence.

But decision quality has a measurable impact on financial performance. Research from Bain shows decision effectiveness correlates with financial results at a 95% confidence level. Firms in the top quintile for decision-making deliver 6 percentage points higher total shareholder return than their peers. McKinsey finds that fast-decision companies are twice as likely to outperform on profitability.

Why It Persists

Reactive governance feels like firefighting, so most organisations invest only in processes for planned decisions and manage reactive ones through slow escalation chains. Even when a firm sees a problem in week 3, without a structured decision process it takes another 2–3 weeks to convene the right people, agree on the response, and implement it. A 3-week latency becomes a 5–6 week effective lag. Real-time dashboards become 45-day dashboards in practice.

Why the order matters

This is what makes the Three Gaps a framework rather than a list of problems. The gaps have a strict sequence dependency:

The sequence is fixed: Signal → Latency → Decision.

Firms that try to fix their decision-making without fixing their signal definition first improve their decisions about the wrong things. This is the most common and most expensive mistake in service improvement programmes.

What closing all three gaps looks like

The evidence for what's possible is already in the data. Firms that have implemented professional services automation with integrated economic visibility achieve +14% project margins, +28% EBITDA, and 20% less revenue leakage than their peers.

The gap between top-performing and bottom-performing firms is not incremental — it is an order of magnitude. Level 5 maturity firms outperform Level 1 firms by 739% on revenue growth, 537% on profit margin, and 71% on utilisation. These are not firms with better people or better clients. They are firms that have closed the three gaps: they know what to watch, they see it in time, and they have a process to act on it.

In a well-instrumented firm, the journey from problem formation to action taken takes 3–5 days, not 6–8 weeks. That difference — measured across a portfolio of engagements over a fiscal year — is the difference between an industry-average 9.8% EBITDA and the 25%+ EBITDA that top performers achieve.

Where Most Firms Sit: The Four Eras

The Three Gaps map cleanly onto the four eras of service delivery. Understanding which era your firm occupies helps locate which gap is your most immediate constraint.