Scope Creep at 22% Overrun: What It Costs Per Year

Scope creep is the silent tax on every solo consulting business. The 22% overrun number is so consistent across industries and project types that it might as well be a constant. This article runs a typical solo AI consulting project through the Scope Creep Cost calculator and shows where the $19,440 annual loss comes from and how to convert it back into revenue.

1. The numbers: $19,440 annual creep tax

The scenario: solo AI consultant, $14,400 project, quoted at 80 hours at $180/hour, actually delivered in 98 hours. The engine returns:

# scope-creep-cost-calculator (computed live from /engines/scope-creep-cost-calculator.js)
Engine input
  project_quote         = 14400
  billable_rate         = 180
  quoted_hours          = 80
  actual_hours          = 98
  projects_per_year     = 6

Engine output
  effectiveHourlyRate   = 146.94
  rateCollapsePercent   = 18.37
  unpaidHours           = 18
  unpaidHoursValue      = 3240
  annualCreepLoss       = 19440
  scopeCreepTaxPercent  = 22.5
  quotedTotal           = 14400
  actualTotal           = 17640
  overrunPercent        = 22.5

The Standish Group's CHAOS Report places average scope creep at 23-30% across software projects, with median around the same band^[1]. PMI's Pulse of the Profession reports project overruns ranging from 15% to 40% depending on industry and project type^[2]. The 22.5% in this scenario sits at the median — typical, not exceptional.

2. From $180/hr to $147/hr — the rate collapse

The most diagnostic number is the rate collapse. The consultant thinks they bill at $180/hour because that's the rate on the proposal. Actual realised rate is $146.94 — an 18% reduction. Across the consultant's full year, this means every "$180/hour" project is silently a "$147/hour" project. Six projects' worth of effort produces five projects' worth of revenue.

The number compounds in two directions. First, the consultant prices the next project at the same nominal rate, which underestimates time and reproduces the overrun. Second, the consultant compares their realised hourly rate (the $147 they actually earn) against the nominal rate of competitors and concludes they should raise prices when in fact they should fix scope first.

3. Where the 18 unpaid hours come from

Scope creep on solo projects breaks into four predictable categories:

• Out-of-scope requests treated as "small." Client asks for one more dashboard, one more report, one more integration. Each one takes 2-4 hours and feels too small to bill separately. Across 6-10 small requests per project: 10-30 hours of unbilled time.
• Underestimated complexity in the original SOW. The consultant quoted 80 hours assuming the data was clean, the API was stable, and the requirements would hold. Half the time, one of those assumptions breaks and adds 5-15 hours.
• Revision cycles beyond the original commitment. Most contracts allow "up to 2 rounds of revisions" and most clients actually need 3-5. Each extra round is 2-6 hours. The consultant doesn't push back because the relationship matters.
• Post-delivery support not in the contract. Bug reports, "quick questions," reauth on the API the consultant set up six months ago. Each item is small; the year total is 30-80 hours.

The 18 hours in this scenario maps roughly to 6-8 small out-of-scope requests plus one underestimated complexity event plus one extra revision round. Multiplied by 6 projects per year: 108 hours of unbilled time annually. At $180/hour that's $19,440, which the engine reports.

4. Pricing for the 22% overrun upfront

The simplest fix is to price for the overrun before it happens. If the consultant's gut estimate is 80 hours, and the historical overrun is 22%, the right quote is 80 × 1.22 = 98 hours, or $17,640. Two things this does. First, it produces accurate revenue. Second, it gives the consultant slack on the project — if it lands at 80 hours, they have time for the next project; if it lands at 98 hours, they were paid for what they delivered.

The objection is competitive pricing. Quoting 22% higher than the consultant's gut estimate makes them seem more expensive than competitors who quote at gut. The realistic answer is that the competitors also overrun by 22% and absorb it — they're either burnt out, exiting consulting, or running their business at the same rate-collapse. Pricing for the overrun is what makes consulting sustainable past year three.

Upwork's 2024 Freelance Forward data shows that freelancers who explicitly include change-order clauses and pad estimates report 30-50% higher effective hourly rates than those who don't^[3]. The pricing discipline is the largest single income lever in freelance consulting.

5. Contract clauses that earn back $19k

Three clauses that materially reduce the creep tax:

• Explicit scope definition with named deliverables. Not "build the dashboard" but "build a dashboard with the following five widgets, on this data source, with the following filters." Anything not named in the SOW is a change order. Reduces "small request" creep by 40-60%.
• Change order rate and process. "Any work outside the original SOW is billed at $180/hour, in 4-hour minimum blocks, on approval of a written change order." Forces the conversation about scope before the work happens. Converts the 18 unbilled hours into 18 billed hours.
• Limit on revision rounds. "Includes 2 rounds of revisions. Additional rounds billed at the change-order rate." Pushes the client to consolidate feedback. Cuts revision-cycle creep by 50-70%.

These clauses do not require a lawyer to draft. A two-paragraph addendum to the standard SOW handles all three. The hard part is enforcing them — most consultants write them and then waive them when the client pushes back. The clauses are only worth what the consultant is willing to enforce.

6. Client types and their creep profiles

Creep is not evenly distributed across client types. Three patterns:

• Enterprise clients (Fortune 500). Highest creep rates (30-40%) but also highest tolerance for change orders. They expect formal scope management. Push back hard on creep; they'll respect it. Project economics typically rescue themselves through change orders.
• SMB and solo-founder clients. Moderate creep (15-25%) but resistance to change orders — they perceive it as "nickel and diming." The right pricing posture is generous upfront SOW (build in the 22% buffer) and rare change orders for genuinely large additions.
• Agency intermediaries. Creep is structurally driven by the agency's underbidding to win the original engagement. Avoid these projects unless the rate clears 1.5x the consultant's direct-client rate to compensate for the creep tax.

7. The discipline that stops creep at the source

Four habits that reduce creep without changing contract terms:

• Log time per project, not per day. Toggl, Harvest, or even a spreadsheet. The act of tracking time per project surfaces creep in real-time; the consultant sees the hour count climbing past the SOW before the bill arrives.
• Send weekly status emails with hours used and hours remaining. Forces the conversation about the budget before it's blown. Sets the client's expectation that hours are finite.
• Refuse "quick questions" outside the SOW. "Happy to help — I'll send a change order for the 2 hours this will take." Either the client agrees (revenue captured) or the request quietly disappears (time recovered).
• Run the calculator after every project. Compare quoted vs actual hours. If overrun trends above 25%, the gut estimate is broken and needs structural correction (longer quotes, more buffer, better SOW templates).

One pattern worth pre-empting: the "client of record" trap. Solo consultants who anchor on a small handful of regular clients tolerate higher creep rates because the relationship matters more than any single project's economics. The right framing is that the relationship is precisely why the change-order discipline should be tighter, not looser. Clients who push back on hourly change orders for a long-time consultant are signalling that they would prefer the consultant absorb the cost — which they will, until they quietly stop accepting the next engagement. Long-term clients deserve transparent pricing more than transactional clients, not less.

The $19,440 annual loss is roughly one month of unbilled capacity. Recovering it pays for a vacation, a new tool subscription, or six months of growth investment in the consulting business. The freelance rate capacity planner handles the broader question of how many billable hours are realistic at the target income. The project pricing calculator handles fixed-fee vs hourly trade-offs for specific projects. See the methodology for the full derivation^[4].