Where the Fee Money Actually Sits

Most people writing an acivil engineering business plan start by quoting the size of the construction market. That is the wrong denominator. A civil engineering practice does not sell concrete or asphalt; it sells stamped drawings, calculations, permit submissions and professional judgement. The number that matters is the size of the engineering services fee pool, and it is a fraction of the construction spend it sits on top of.

In the United States, the engineering services industry is worth $386.7 billion in 2026, up from $350.8 billion in 2025, having grown at a 3.1% compound annual rate between 2021 and 2026 according to IBISWorld, 2026. That is the pool your firm bills against. Within it, the largest consultancies are consolidating: the ENR Top 500 Design Firms, 2026 list reports combined revenue of $158.7 billion, up 7.4% year on year, which means roughly 41% of all US engineering services fees flow through 500 firms.

The UK side of the same question

Total UK construction output reached £225.5 billion in 2025, split into £132.9 billion of new work and £92.5 billion of repair and maintenance, with new work making up around 59% of the total, per ICAEW analysis of ONS data, 2025. London alone accounted for roughly a quarter of new work in the final quarter of the year, which is why so many small consultancies cluster in the South East and then struggle to differentiate.

The sub-sector that most directly drives civil consultancy demand is less comfortable reading. IBISWorld UK, 2025 puts civil engineering project construction at £41.2 billion, a 6.5% contraction in 2025. A plan that claims a rising tide in UK civils will be contradicted by the first reviewer who checks. The stronger argument is counter-cyclical: when programme volumes compress, clients cut in-house technical capacity first and buy it back as consultancy, and repair and maintenance work at £92.5 billion is far more stable than new build.

Demand drivers to write into the plan

• Asset renewal, not new build. Culverts, retaining walls, bridge decks and drainage systems installed in the 1960s and 1970s are reaching the end of design life simultaneously. Condition assessment and remediation design is recurring work that does not depend on a development cycle.
• Stormwater and flood compliance. Detention design, SuDS schemes, flood risk assessments and discharge consents are triggered by regulation rather than by client appetite, which makes the fee defensible in a downturn.
• Small-municipality capacity gaps. Towns and districts that cannot justify a full-time staff engineer buy on-call engineering by retainer. These are unglamorous, low-competition, high-renewal accounts.
• Private land development. Site grading, utility layout, access design and entitlement support. The highest fee velocity available to a small practice, and the first thing to disappear when interest rates rise.
• Specialist technical review. Peer review, expert witness work and constructability review bill at principal rates with almost no drawing production cost attached.

Choose two of those five and build the plan around them. A services section that claims civil, structural, mechanical, electrical, environmental and geotechnical capability inside a two-person firm reads as inexperience, and both lenders and sophisticated clients treat it that way.

If your practice leans toward one discipline, the plan should say so explicitly and the supporting material should match. We maintain separate guides for a civil engineering consulting business plan, a structural engineering consulting firm plan, a geotechnical engineering consulting firm plan and a surveying firm plan, each with its own cost base and fee behaviour.

SBA Lending & Wage Benchmarks for Engineering Practices

Two datasets do more work in a civil engineering plan than any amount of narrative. The first tells you what a lender has historically been willing to write against a firm like yours. The second tells you what your payroll will actually cost, which is the input every other number in the forecast depends on.

What SBA lenders have funded under NAICS 541330

Engineering services is classified under NAICS 541330, and it is a well-travelled SBA category rather than an exotic one. Across all SBA programmes, 12,075 loans have been approved to businesses in this code, deploying $4.0 billion of capital. Of those, 10,205 (85%) were 7(a) loans. The average engineering services loan is $329,000, about 3% below the $340,000 national SBA average, with a typical term of 125 months, per PeerSense SBA industry data, NAICS 541330.

The trend inside that data is worth quoting in your plan. Windsor Advantage, 2020 tracked 3,248 SBA loans to engineering firms between October 2009 and September 2019 totalling roughly $1.15 billion, and noted that the average loan size rose from $194,000 in FY2010 to $507,000 in FY2018, a 161% increase. Lenders have become steadily more comfortable writing larger facilities into this sector, which means a $150,000 to $300,000 ask for a new practice is unremarkable rather than ambitious.

What the underwriter will test is not whether civil engineering is a good business. It is whether your projected cash flow services the debt with a cushion, and whether your receivables assumption is honest. That second point is where new engineering practices fail credit committee, and we cover it in detail in the cost section below.

In the UK the equivalent entry route is the government-backed Start Up Loans scheme, which lends up to £25,000 per founder at 6% fixed with free mentoring. For a two-director practice that is up to £50,000, which covers software, insurance and a modest working capital buffer but not a survey fleet. Beyond that, UK consultancies typically fund growth from invoice finance against certified applications rather than from term debt, because the asset base is too thin to secure against.

What your payroll will cost

The US Bureau of Labor Statistics Occupational Outlook Handbook reports median pay of $99,590 a year ($47.88 an hour) for civil engineers (SOC 17-2051) in 2024. The bottom ten percent earned under $68,240 and the top ten percent above $163,220. Employment stood at 368,900 jobs in 2024, projected to grow 5% through 2034 with about 23,600 openings a year over the decade.

Read that distribution as a hiring plan rather than a statistic. The $68,240 band is where entry-level engineers-in-training sit, and that is the labour you need for production drawing work. The $163,220 band is senior PE and principal territory, and you cannot bill enough principal hours to pay two of them in year one. A practice that hires at the top of the distribution before it has a backlog burns its working capital in under two quarters.

What It Costs to Open the Doors

Starting a civil engineering practice takes $44,000 to $239,000 (£34,000 to £188,000) depending on whether you subcontract survey, whether you take a lease, and how many states or frameworks you register in. The spread is wide because almost none of it is fixed: a solo PE can be invoicing within six weeks for under $50,000, while a six-person studio with its own field kit and multi-state licensure needs the top of the range before the first invoice clears.

Line-by-line, with named vendors and real prices

• Design software. An Autodesk Civil 3D seat runs about $305 a month or $2,430 a year. Bentley Civil WorkSuite starts at $6,161 for a 12-month Virtuoso subscription. Pick one toolchain and standardise; running both because two founders trained differently costs a small practice $8,500 a year for zero extra capability.
• Project accounting. Deltek Vantagepoint, BQE Core or Unanet are the A&E-specific options. They matter because generic bookkeeping software cannot report utilisation or net labour multiplier, and those are the two metrics your whole plan rests on. Budget $60 to $120 per user per month.
• Professional liability / professional indemnity. Insureon reports engineers paying an average of $168 a month, roughly $2,010 a year, with civil engineers in 1 to 5 person firms typically at $1,000 to $2,500 a year. UK practices should budget £1,500 to £9,000 depending on the limit of indemnity their client contracts demand.
• Firm licensure. Certificate of Authorization applications, PE registration renewals and multi-state registration: $1,500 to $12,000 depending on how many jurisdictions you intend to stamp in. This is the line most plans omit entirely.
• Survey and field equipment. A robotic total station, GNSS rover, levels and PPE run $8,000 to $70,000. For the first year, subcontracting survey to a local firm and marking it up as a pass-through is almost always the better cash decision.
• Workstations, plotter and cloud storage. $6,000 to $28,000. Civil 3D is unforgiving on under-specified hardware, and a draughtsperson waiting on a surface rebuild is unbilled time.
• Working capital. $20,000 to $90,000, and the real answer sits at the top of that band if your client mix is public sector.

The receivables gap is the real startup cost

Payroll runs weekly or fortnightly. Municipal, county and DOT clients pay on 60 to 120 day cycles, and the clock usually starts at invoice approval rather than invoice submission. A practice billing $55,000 a month into public-sector clients needs something close to $110,000 to $165,000 of standing working capital just to stay solvent at steady state, before it funds a single piece of growth.

This is also what your plan should lead with when it reaches the funding section. An underwriter looking at an engineering practice is not worried about whether you can do the work. They are worried about the gap between when you pay your people and when your clients pay you. Put a 13-week rolling cash view in the appendix, show the gap explicitly, and state what facility closes it. That single exhibit changes the tone of the conversation.

Our Research + Content package builds that cash view from your actual client mix and payment terms rather than from a generic template, and the bespoke plan extends it to a full five-year forecast with debt service coverage.

Three Ways to Build the Practice

"Civil engineering firm" covers at least three businesses with different cost bases, different cash profiles and different exit values. Picking one and saying so is the clearest signal of seriousness a plan can send. Trying to run all three at once is why small practices stall at four people.

The competitive reality is worth stating plainly, because it is where most plans get vague. At the top of the market, ENR's 2026 Top 500 puts Jacobs Solutions at number one, AECOM at two and WSP at four. In the eastern US alone, ENR's 2025 regional list had AECOM at $1.67 billion, WSP USA at $1.39 billion and Stantec at $640 million. You are not competing with those firms for prime contracts in year one. You are competing to be their subconsultant, and separately competing against the 15 to 40 person regional practices that actually hold the small-municipality retainers in your county.

The practical implication for the plan: name the three to five local firms you will take work from, say what each is weak at, and show the specific first five pursuits. Reviewers can verify a named pursuit list. They cannot verify "we will differentiate through responsiveness and quality".

Fee Structures, Rates & Unit Economics

An engineering practice converts salaried hours into billed hours. Everything else is detail. Two ratios describe that conversion, and if your plan does not contain both, it is not a credible engineering plan.

Utilisation rate is billable hours divided by total paid hours. Net labour multiplier is net revenue divided by direct labour cost: how many dollars of fee each dollar of billable salary produces. The 46th annual Deltek Clarity Architecture & Engineering study, covering nearly 700 firms across the US and Canada, reported median utilisation of 58.9% in fiscal 2025, down 2.2 points from 61.1% and the lowest since 2021, alongside a net labour multiplier of 3.15, the highest in a decade, and operating profit on net revenue of 21.4%. Firms Deltek classifies as high performers, defined as a multiplier of 3.0 or better and operating profit of 15% or better, reached 31% profitability and made up 17% of participants.

A staff rate schedule you can defend

Licensed civil engineers bill roughly $100 to $250 an hour in 2026. Billing from a single blended rate is the most expensive habit a young practice can form, because it sells technician production at engineer cost and gives away principal judgement. Use a ladder:

Rate bands from LaunchAdvisor engineering firm pricing guide, 2026 and LandPro Civil fee guide, 2026. Utilisation targets are Avvale planning assumptions anchored to the Deltek median of 58.9%.

Worked example: a four-person practice

One principal-in-charge, one project engineer with a PE, one engineer-in-training, one CAD technician. At 1,900 paid hours each, that is 7,600 paid hours a year. Apply the Deltek median utilisation of 58.9% and the practice books roughly 4,478 billable hours. At a blended realised rate of $142, after the discounts and write-offs every practice takes, gross revenue is about $636,000.

Strip out subconsultant and reimbursable pass-through at roughly 11% and net revenue is about $566,000. At the 3.15 net labour multiplier, that net revenue corresponds to roughly $180,000 of direct billable labour cost. Apply the 21.4% median operating profit on net revenue and the practice clears about $121,000 after the principal's salary sits in overhead.

Now move one number. Lift utilisation from 58.9% to 65% and the same four people book about 464 more billable hours, roughly $66,000 of additional fee. Those people were already paid, the software was already licensed, the rent was already due, so almost all of that $66,000 drops to the bottom line and operating profit moves from roughly $121,000 to $187,000. That single sensitivity is the most persuasive exhibit you can put in front of a lender, because it shows you understand that your business is an hours-conversion machine rather than a drawing factory.

Revenue mix worth building in

• Fixed-fee design phases. Highest margin when scope control is good, worst exposure when it is not. Phase the fee and bill on milestone, never on a single completion date.
• Hourly time-and-materials. Right for unknowable scope: condition assessment, forensic work, permit negotiation. Caps are negotiable; abandoning the hourly basis is not.
• On-call retainers. A $4,000 to $12,000 monthly municipal retainer buys baseline cash that covers fixed overhead, which is what lets you refuse underpriced development work.
• Construction phase services. Submittal review, RFI response and site observation. Low design intensity, long duration, and it keeps senior staff billable between design pursuits.
• Subconsultant pass-through with markup. Survey, geotechnical and environmental routed through you at 5% to 15%. Record it as pass-through, not revenue, or your margin reporting becomes meaningless.
• Expert and peer review. Principal rates with no production cost. One expert engagement can carry a quarter of overhead.

Across small and mid-size practices, net margins commonly land in the 10% to 20% band, with the Deltek median at 21.4% of net revenue and high performers at 31%. If your forecast shows 35% in year one, a reviewer will assume you have mislabelled gross revenue as net, which is the single most common arithmetic error in engineering plans we are asked to fix.

Firm Licensure & Statutory Duties

There are two separate permissions in engineering, and almost every plan we review covers only the first. The individual needs a licence to practise. The firm needs a separate authorisation to offer services to the public and to invoice for them. Miss the second and you have a company that cannot legally take the work it just won.

United States

All 50 states and the District of Columbia require a licensed Professional Engineer to offer or supervise engineering services to the public. The route is an ABET-accredited degree, the Fundamentals of Engineering exam, four years of supervised experience, then the Principles and Practice of Engineering exam. Exam fees are $175 for the FE and $375 for the PE, plus state application and licence fees that commonly run $100 to $400.

At firm level, Harbor Compliance's 50-state guide records that 34 states require a Certificate of Authorization or equivalent firm licence from the state engineering board before a firm may offer services, with 8 further states imposing other registration requirements. The recurring conditions are worth planning around:

• An engineer in responsible charge. You must assign engineering decisions to a named PE (or someone under their supervision) before the certificate issues. If that person leaves, you must notify the board and appoint a replacement.
• Ownership and management tests. Several states require majority ownership or control by licensed engineers, which directly constrains how you can structure outside investment.
• Firm naming rules. Some states require the word "engineering" in the registered name, and some restrict the use of individual surnames. Check before you print letterhead.
• Secretary of State registration. Entity formation and good standing is a prerequisite to the board application, not a parallel task.
• Per-state repetition. Every state you stamp in is its own application, its own renewal cycle and its own resident-PE question. A plan claiming regional coverage should show the licensure cost of that coverage.

United Kingdom

The UK has no equivalent firm licence, which founders often read as a lighter regime. It is not; the burden simply moves from licensing to statutory duty-holder roles and insurance.

• CDM 2015 principal designer. Under the Construction (Design and Management) Regulations 2015, a client appoints a principal designer on any project involving more than one contractor. The role, as the Institution of Civil Engineers sets out, is to plan, manage, monitor and coordinate health and safety through the pre-construction phase, and the appointee must hold sufficient knowledge, experience and ability plus genuine construction experience.
• Building Regulations principal designer. England now has a second, distinct principal designer role created under the Building Safety Act regime, with its own duties and its own consequences for non-compliance. Being appointed CDM principal designer does not discharge it. Your plan should state which role or roles the practice will accept, and what competence evidence supports that.
• Chartered status. CEng registration through ICE under the Engineering Council is not a legal licence to practise, but it functions as a commercial gate: Tier 1 contractors and public-sector frameworks routinely require a chartered engineer in a named technical role.
• Professional indemnity. Required by essentially every client contract, at a limit the contract specifies, in force before signature. Budget £1,500 to £9,000 a year for a small practice and expect the limit demanded to rise with project value.
• Prequalification schemes. Constructionline, CHAS and SafeContractor registration is a practical prerequisite for public and Tier 1 work and takes weeks, not days. Treat it as a launch task, not an afterthought.

Ontario, Canada

Canada regulates provincially, and Ontario is the clearest illustration of the firm-level requirement. Professional Engineers Ontario issues a Certificate of Authorization to individuals and business entities that offer professional engineering services to the public, explicitly distinct from the individual P.Eng. licence. Applicants must designate a PEO licence holder to assume professional responsibility, supply that engineer's resume, provide a Master Business Licence if trading under a different name, and comply with the professional liability insurance regulations under the Professional Engineers Act.

One detail belongs in any plan contemplating Ontario work: if the only P.Eng. named on a firm's Certificate of Authorization resigns, the certificate is suspended for 90 days and revoked if no suitable replacement is found. A single-PE practice therefore carries a hard key-person risk that an underwriter will price. Naming a second licensed engineer, even part-time, removes it.

Write the licensure section as a dated checklist with owners and costs, not as prose. Reviewers skim it for omissions, and the omission they look for is the firm-level authorisation.

Where These Plans Get Rejected

We have redrafted enough engineering practice plans to know which six errors account for most of the rejections. None of them is about technical competence.

1. The everything-discipline services page

A services section claiming civil, structural, mechanical, electrical, environmental and geotechnical capability inside a two-person firm has no credibility with a lender or with a sophisticated client. It reads as an attempt to look larger. Name two disciplines, evidence them with specific project experience, and list the rest as subconsultant-delivered.

2. Forecasting revenue from 2,080 hours per head

Headcount multiplied by annual hours multiplied by billing rate is the fastest way to build a forecast nobody believes. At the Deltek median of 58.9% utilisation, that method overstates fee capacity by roughly 40%. Forecast per role, at role-specific utilisation, and show the arithmetic.

3. Budgeting for the PE licence and not the firm licence

In 34 states the firm itself needs a Certificate of Authorization before it can offer services. Plans that budget the $375 PE exam and nothing else have a company that cannot invoice in its own name on day one. Add the state-by-state filing and renewal cost to the startup table.

4. Monthly cash flow against public-sector payment terms

Payroll is weekly. Municipal and DOT payment runs at 60 to 120 days from approval. A monthly cash flow statement hides the gap entirely. Put a 13-week rolling view in the appendix, show the trough, and name the facility that covers it. This is the single most common reason an otherwise sound engineering plan fails credit committee.

5. One blended billing rate

A single rate sells CAD production at project engineer cost and gives away principal judgement for free. It also makes it impossible to diagnose a bad project, because you cannot see which labour grade overran. Publish a staff rate schedule and bill against it.

6. Counting subconsultant pass-through as revenue

Routing $180,000 of survey and geotechnical work through the practice at a 10% markup adds $18,000 of margin and $180,000 of noise. Report net revenue, state the pass-through separately, and your margin percentage becomes a number a reviewer can actually use. Every benchmark in this guide, including the 21.4% Deltek operating profit figure, is measured on net revenue.

A seventh habit is not an error so much as a missed opportunity: almost no small-practice plan includes a named pursuit list. Five specific upcoming projects, the client, the likely prime, the fee range and your role, does more to establish credibility than ten pages of market narrative. You can see how we structure that evidence in our published client plans.

Sample Business Plan Preview

Here is an extract from a civil engineering practice plan written by our team, so you can see the level of specificity we work to:

What's in the Template

Every Avvale business plan template includes these sections, pre-structured for your industry, with the engineering-specific prompts noted below:

• Executive Summary — the practice in one page, with the funding ask and the use of proceeds stated up front
• Company Overview — entity type, ownership against state engineer-ownership rules, office location and founding story
• Services & Disciplines — the two disciplines you are evidencing, and what is subconsultant-delivered
• Industry Analysis — engineering services fee pool rather than construction spend, with the demand drivers that apply to your region
• Client Analysis — segment by payment behaviour as well as by sector, because that is what sets your working capital
• Competitor Analysis — named local firms, the primes you will sub to, and your first five pursuits
• Marketing Plan — prequalification registrations, SOQ library, chapter and procurement-portal presence
• Operations Plan — staff rate schedule, role-level utilisation targets, QA and stamping workflow, software stack
• Licensure & Compliance — PE registrations, Certificate of Authorization by state, insurance limits, duty-holder roles
• Management Team — founder bios with stamp authority, and the second-PE appointment that removes key-person risk

The optional Financial Forecast add-on (included in our $300/£250 and $1,000/£800 packages) provides a five-year Excel model with income statement, cash flow, balance sheet, break-even analysis and startup capital requirements, built with utilisation and net labour multiplier as the driving inputs and a 13-week rolling cash view for the receivables gap. If you would rather have the whole thing written, see our business plan writing service or start from the free template library.

Frequently Asked Questions