How to Write an Engineering Proposal That Wins Projects

Most engineering firms write proposals the same way. A principal pulls a project manager off billable work for two weeks, the team reuses the approach section from the last pursuit, and the package goes out on the deadline day. The debrief, if there is one, says evaluators found the firm qualified but scored another team higher on approach. The hours are gone and nothing in the firm’s records explains why.

The 47th Deltek Clarity A&E Industry Study surveyed 896 firms in 2026. It puts the median win rate at 49% and the capture rate at 44%, the lowest since 2019. Proposals submitted rose 32% over the same period. The prior edition broke win rates out by size, and A&E firms with 0–50 employees won 40.2% of pursuits against 50.0% for firms of 51–250. Zweig Group’s 2026 marketing benchmarks put the median proposal win rate at 45%. Small firms sit below every one of those medians.

The gap closes when you treat an engineering proposal as a response to a published scoring rubric rather than a brochure. Qualifications-Based Selection, or QBS, has the agency score your people, your record and your approach, and the agency negotiates the fee only after it has ranked you first. Specifics are what the technical approach is scored on. The fee is not scored at all until the agency has ranked you first.

Why qualifications matter more than price

The Brooks Act removes fee from the selection decision. Congress enacted it as Public Law 92–582 in 1972, and it is codified at 40 U.S.C. §§ 1101–1104. Federal agencies must announce architect-engineer requirements publicly, evaluate firms on demonstrated competence and qualifications, hold discussions with at least three firms, and rank at least three in order of preference. The contracting officer then negotiates fair and reasonable compensation with the top-ranked firm. The agency terminates failed negotiations and moves to the second firm.

FAR Subpart 36.6 implements the act, and 36.601-3 applies it in place of Parts 13, 14 and 15 whenever the work substantially requires a licensed architect or engineer. Price doesn’t enter the selection stage. The same rule reaches state and local work paid with federal highway money: 23 CFR 172.7 bars cost proposals, salary rates and indirect rates from evaluation, ranking and selection, and requires the agency to rank at least three consultants. State law extends it further. NSPE counts 46 states with some form of QBS law, and the ACEC Research Institute’s 2022 analysis reports a mandate in all but four states, naming Indiana and Vermont among those without one.

Federal agencies changed what their solicitations tell you under the 2025 FAR overhaul. Model deviation text dated July 2025 applies to solicitations issued on or after October 1, 2025, restructures Part 36 into pre-solicitation, evaluation and award, and post-award subparts, and removes the specific evaluation criteria formerly at 36.602-1 and 36.603. The Brooks Act, evaluation boards, discussions with three firms and ranked selection all survive. No final rule has been issued, so the deviation stays in effect agency by agency. Read each announcement’s own stated criteria and order of importance before you outline anything.

The six criteria at legacy FAR 36.602-1 still describe what evaluators read for, and 23 CFR 172.7 lists nearly the same factors for federal-aid work:

• Professional qualifications: Licenses, registrations and credentials of the people who’ll do the work.
• Specialized experience and technical competence: The firm’s record on the type of work solicited, including energy conservation and recovered materials.
• Capacity: The ability to complete the work in the required time.
• Past performance: Cost control, quality and schedule on prior contracts.
• Location and knowledge of the locality: Presence in the general geographic area and familiarity with local conditions.
• Other appropriate criteria: Anything the agency adds in the announcement.

The six criteria don’t carry equal weight. DFARS PGI 236.602-1 adds that secondary factors such as geographic proximity and equitable distribution of work can’t outweigh qualifications and past performance. Put your strongest evidence behind your people and your record, and treat location as supporting detail.

Qualifications, past performance and approach decide your ranking

Winning under QBS depends on how clearly you prove the qualifications of your people and the results behind your past performance. Your technical approach then shows evaluators how that team will address the specific project. These three criteria carry the most weight, so give each one your strongest evidence. Evaluators score each criterion separately against the announcement’s stated order of importance.

Professional qualifications with substance

Evaluators score the person, so a resume that describes the project instead of the individual’s role earns nothing. Ruth Lesser’s 2024 SAME presentation on the SF330 lists the recurring failures: missing dates on resume projects, resumes that describe the overall project rather than the person’s own work, and inconsistencies between Section E resumes, Section F projects and the Section G matrix. USACE’s EP 715-1-7 states that boards won’t assume qualifications that aren’t clearly stated and that missing, conflicting or obsolete information ranks a firm low or disqualifies it. A PM who ran the structural scope on a similar bridge has to say so, with the year, the owner and the task she led.

The SF330 gives you the structure. Section E holds resumes of key personnel with up to five projects each, and Section G shows which of those people worked on the Section F example projects. Agencies cap the length. USACE New York District limits resumes to one page, Caltrans allows two pages, and a NAVFAC announcement from April 2026 caps the submission at 24 resumes considered collectively. Those limits push proposal teams toward one page per person that names role, dates and the decision that person made on the comparable job.

Key personnel carry the largest share in some municipal rubrics and tie for it in others. Seattle City Light’s RFQ SCL-24257 gives key personnel 50 of 100 points, Seattle’s 20-107-S RFQ gives key team qualifications 35%, and Austin’s QBS matrix gives key personnel experience 20 points, the same as project approach. Owners also score whether the team you propose is the team they get. An SMPS owners survey lists a strong team pitched and a weaker team assigned as a recurring frustration, and 70% of respondents in a 2015 SMPS Foundation study cited client expectations to meet project staff as the main reason for the seller-doer model. Name the PM who’ll run the job and state core-team availability in hours per month, which WSDOT’s 2024 Southworth program asked for directly.

Past performance that proves results

Agencies verify past performance from their own records, so every example project you choose needs a traceable evaluation behind it. FAR 42.15 makes CPARS the official source, requires evaluations for A&E contracts of $45,000 or more, and lets agencies use them for six years after completion on A&E work. USACE Honolulu’s 2025 announcement required a CPARS or past performance questionnaire for every Section F project and treated firms with no record as Unknown, scored as Acceptable. NAVFAC’s CLEAN VII solicitation states that awards and commendation letters aren’t considered. A 2023 Sacramento notice excludes them too. A client reference who’ll answer a questionnaire well is worth more than any plaque.

Section F allows ten example projects unless the agency specifies otherwise, and the specification varies widely. USACE Omaha’s July 2025 announcement capped Section F at ten projects, two pages each, and excluded anything completed before June 2020. Honolulu allowed five projects completed within seven years. NAVFAC’s POL solicitation wanted a minimum of four and maximum of six projects completed within five years, while CLEAN VII accepted ten projects completed or at least 85% complete within ten years. Pick projects for the same client or a similar one, inside the geographic footprint, and date every one.

Past performance under legacy FAR 36.602-1 means cost control, quality and schedule, and most rubrics weight it less than firms expect. MDOT’s best-value scheme assigns 20 of 130 points to it, Caltrans District 3 gives it 2 of 10, and WSDOT’s manual requires references without scoring them. A poor evaluation can still eliminate a firm. A 2026 TxDOT pre-RFP presentation lists a past-performance score of 5 to 15 points. The record only helps when you can state the cost and schedule outcome for each project, and that means the firm has to have tracked fee burn and milestones while the project was open.

Technical approach that shows understanding

The technical approach is where the published weights give a smaller firm its best chance, and it’s the section most firms fill with boilerplate. PSMJ describes also-ran proposals as too generic, padded with irrelevant statistics and boilerplate.  ACEC’s 2026 career-center guidance on SOQs says generic claims score nothing because every competitor writes them.

The weight justifies the effort. Phoenix’s 3rd Street Connector solicitation gave project understanding and approach 425 of 1,000 points, 42.5% of the SOQ score. FDOT’s procurement manual scores awareness of project issues 0–30 and proposed approach 0–30, and ACEC Oklahoma’s model scoring gives 20 points to grasp of requirements and 20 to design approach. FDOT’s Chapter 7 adds a caution that explains a lot of lost pursuits: proposals should be limited to project awareness, approach and staffing, and performing actual design in the proposal should be discouraged. Evaluators want evidence that you’ve identified the site’s problems. The design work belongs in the contract.

Agency requirements converge on the same content. TxDOT’s RFQ template asks for scope understanding and the top project risks and how you’d address each, NCDOT’s US 19W RFQ asks for design and construction challenges, risk mitigation and collaboration ideas, and MnDOT’s RFP Part A asks for a project understanding statement and a detailed work plan with tasks, deliverable dates and personnel by task. A Sacramento Section H went further and required a work management plan, a quality management plan covering version control, comment tracking and independent reviews, communication protocols, and a capacity plan to run at least three $500,000 task orders at once. Build the section from those five elements:

• Project understanding: The site, the owner’s constraints, the stakeholders, and the issues the RFP doesn’t say out loud.
• Risks and mitigation: The three to five technical or schedule risks you’d flag in a kickoff meeting, each paired with the step you’d take.
• Work plan: Tasks by phase, deliverable dates, and the named person responsible for each task.
• Quality management: Who checks what, when independent review happens, and how comments are tracked to closure.
• Work management and communication: Meeting cadence, submittal routing, and how you’ll report progress against fee and schedule.

Length comes from the solicitation. Section H caps in recent federal notices run from 5 pages at USACE Huntsville to 15 at USACE Seattle, NAVFAC Hawaii and USACE Galveston, with 10 at NAVFAC Pacific, and Caltrans allows the prime 4 pages. PSMJ’s advice fits inside those caps: follow the RFP’s order, use its terminology, and allocate pages by the relative weight of each criterion. SMPS recommends two fresh reviewers before submission, one for content and one for proofreading.

Show evaluators how your disciplines will coordinate

An engineering proposal for a multi-discipline project gets scored on how the disciplines will work together as well as on who’s on the chart. USACE New England’s 2026 solicitation lists work management and teaming third among its primary criteria, Seattle’s 20-107-S gives the organization chart and roles 15%, Denver’s 2026 on-call evaluation gives staffing approach 10 points and communication 5, and the Michigan QBS Coalition interview sheet gives consultants 10 of 100. SF330 Section C lists the proposed team and Section D holds the organizational chart, and FAR 9.603 requires team arrangements to be disclosed in the offer. FAR 9.604 keeps the prime fully responsible for performance, so the evaluator wants to know how you’ll manage what you’ve committed to.

The org chart shows reporting lines. The narrative has to show the review cycles. Name the person who’ll be the written point of contact for each subconsultant, the cadence at which discipline models and drawings get exchanged, how interdisciplinary comments are logged and closed, and who signs off before each submittal. Name one engineer as the written point of contact with authority to coordinate the other disciplines, and give that person the job of flagging defects in a subconsultant’s work as soon as they appear. Describe the responsible-charge structure as well. NSPE’s Code of Ethics lets the coordinating engineer seal the whole project, provided each technical segment is sealed only by the qualified engineer who prepared it.

Get the teaming agreement signed before the proposal goes out. Foley’s guidance on government teaming agreements covers scope allocation, proposal responsibilities and cost sharing, and it recommends a commitment to award the subcontract rather than a promise to negotiate in good faith. Wiley notes that agreements to agree are too vague to enforce under Virginia law and recommends attaching a draft subcontract.

Standard forms exist for both stages. AIA C102–2015 and EJCDC E-580 cover the pursuit, and AIA C401–2017 and EJCDC E-570 cover the engagement with flow-down of the prime agreement. They also include a standard of care without warranty and indemnity limited to the consultant’s negligence. NSPE’s Board of Ethical Review has found it unethical to join more than one competing team without disclosing it to all of them.

Victor/CNA’s risk advisory warns that the prime carries vicarious liability for subconsultants, so capping a subconsultant’s liability leaves the prime holding the excess unless the owner agreement contains a matching limit, and the same insurer says subconsultant coverage must at least satisfy the prime agreement. Settle limitation-of-liability, indemnity and insurance terms with each subconsultant before you put their name in Section C, since several states, including California, Florida and Texas, restrict indemnity clauses for design professionals.

Non-compliance stops a proposal before it reaches scoring

Non-compliance is the most consistently documented reason an engineering proposal never reaches scoring. TxDOT’s 2024 PEPS session on disqualifications lists extra pages and inconsistent firm names across forms as grounds, citing intro letters, tables of contents and blank pages as extra pages, and its procurement manual rejects any proposal missing an RFP requirement as non-responsive. Caltrans rejects missing forms, late submittals, the wrong delivery location and an inadequate number of copies. Mississippi DOT’s consultant services manual, revised July 1, 2025, adds conditional proposals and incomplete information to the list. The ACEC/PTAB owners manual describes the sequence every committee follows: check conformance to mandatory requirements first, then score against the published criteria.

Format rules vary enough that last year’s template will fail this year’s solicitation. Recent solicitations show how widely the requirements differ:

• USACE Huntsville: Part I was capped at 50 pages in 12-point type, with covers and tabs counted.
• VA: 120 pages with half-inch margins, each sheet counted as two.
• NAVFAC Hawaii: 11-point Times New Roman throughout.
• USACE New York District: Arial 9-point, single-sided.
• USACE Galveston: PDF text must not be locked.

Those variations make the current solicitation, rather than last year’s template, the controlling document. The current SF330 is Rev. 7/2021 under OMB control number 9000-0157, and both the Omaha and Honolulu 2025 announcements require that edition by name. Firms without a Part II on file or attached aren’t considered at all under EP 715-1-7.

A compliance matrix and a maintained master file keep these requirements organized. Build the matrix from the solicitation on day one: every page cap, font, section order, form and deadline becomes a row with an owner. ACEC’s career-center guidance calls those items pass-fail gates, and APMP’s foundation study guide defines compliance as meeting requirements to the letter.

Keep Part II for each branch office. Maintain one-page resumes and dated project sheets in standard documentation packages you refresh on a schedule, because FAR 36.603 has agencies review qualification files annually and encourages annual Part II updates. OpenAsset’s 2024 vendor-published survey of 535 marketing professionals at architecture, engineering and construction firms found 64% submit on the deadline day. Schedule the consistency check across resumes and project sheets two days before that, then compare both with the Section G matrix.

A discounted fee won’t win a QBS pursuit

The agency negotiates the fee only after it has ranked you first, so a discounted number in the proposal buys nothing. FAR 36.606 starts negotiations with the top-ranked firm under Part 15, and the Brooks Act’s § 1104 requires the agency to terminate failed negotiations and move to the next firm. 23 CFR 172.7 bars cost proposals, salary rates and indirect rates from every step of evaluation on federal-aid projects.

Three state agencies set different weights on qualifications versus price. NCDOT stays qualification-based, while MDOT and Colorado’s OSA use best-value schemes that score cost:

• NCDOT: Its 2024 policy requires qualification-based criteria to total at least 90% of the score.
• MDOT: Its best-value scheme for service contracts gives price 35 of 130 points.
• Colorado OSA: The office uses a 70-30 qualifications-to-cost split for integrated-project-delivery interviews.

Confirm whether the solicitation is pure QBS or best value before you decide what fee information to include.

Fee caps shape the negotiation you’re preparing for. FAR 15.404-4(c)(4)(i)(B) limits A&E fees for public works or utilities to 6 percent of estimated construction cost. GAO reports that a DFARS final rule implementing the FY 2024 NDAA raised that statutory limit to 10 percent for Army, Navy and Air Force work, and a proposed rule published September 18, 2026 would confine the cap to cost-type contracts. It has not been adopted. The fee you’ll defend has to be built from hours by discipline and phase, with subconsultant costs and overhead rates you can show, because the negotiator will ask for all three.

Your own pursuit cost is the number proposal teams rarely track. PSMJ’s 2020 excerpt from Frank Stasiowski puts a proposal at $12,000 to $15,000, with three of ten bringing in work. SMPS practitioner guidance puts proposal cost at 5% to 10% of anticipated revenue. Zweig Group’s 2026 benchmarks found project managers account for 59% of total BD effort against 12% for principals, even though 89% of firms name principals as primarily responsible. ENR’s marketing column reports that labor is the number-one sales and marketing expense and advises tracking proposal expenses on major pursuits. Track pursuit hours against each proposal the same way you track hours against a phase, and you’ll know your cost per win by year end.

Carry your proposal’s phase plan into delivery with Monograph

The phase plan you wrote for the technical approach shouldn’t die in a PDF the day you win. It does at most firms. The PM rebuilds the budget in a spreadsheet from the fee letter, the hour allocations by task drift from what the proposal promised, and the past-performance claim about cost control depends on records nobody kept. The next SF330 Section F asks for the project’s cost and schedule outcome, and the answer is a reconstruction.

Monograph is practice management built exclusively for A&E firms with 5 to 50 employees, and 16,000+ architects and engineers across 2,000+ firms work on its phase structure. Its phase-based data model follows the Architect’s Handbook of Professional Practice. Engineering firms set up a project by phase with fee, hours and team assignments, which is the same structure a work plan uses.

Pipeline management holds each pursuit with a proposed budget and a win probability, multiplies the two, and spreads the result across phase timelines so you see probability-weighted revenue instead of a lead list. Capacity forecasting weights expected lead hours by win probability against compensation hours per employee, which answers the capacity question evaluators ask and the hiring question you ask. A signed proposal converts directly into a live project without re-entering the plan.

MoneyGantt™ turns delivery into the past-performance evidence Section F asks for. It combines a Gantt-style timeline with budget-to-cash progression in one display, where a traditional Gantt chart shows dates alone. MoneyGantt™ flags a phase that has used 80% of its budget at 50% completion while you can still correct the fee through staffing or a scope conversation. You can see fee burn against schedule by phase at close-out. That record is the cost-control and schedule story Section F and a CPARS questionnaire ask for. Garrison Architects reports a billing cycle twice as fast with Monograph, a customer-reported outcome rather than an independent benchmark.

The fee you defend in negotiation should come from what similar work actually cost your firm. Monograph reads your historical project data and recommends a budget for a new pursuit, so the number you bring to the table is grounded in closed phases rather than a markup on last year’s letter.

Monograph’s AI covers six workflows. Contract-to-project setup and proposal budget recommendations are the two that bear on proposal work directly:

• Contract-to-project setup: Upload a signed contract and Monograph generates the phase budgets and schedule from it.
• Spreadsheet-to-project import: Upload a milestone or deliverable spreadsheet to build the project structure.
• Proposal budget recommendations: Monograph uses historical project data to recommend the budget for a new pursuit, so the fee you negotiate rests on what similar work cost your firm.
• Staffing recommendations: Monograph uses team availability to recommend who can take the work.
• Calendar-based time suggestions: Monograph surfaces calendar activity as candidate entries in the web timesheet, which the person confirms before they count.
• Natural-language time entry on mobile: Monograph converts a description of the work into a structured phase-level entry after confirmation.

Your team’s phase-level time entries give Monograph the history it uses to make the next proposal’s budget recommendation reliable. Spreadsheets and disconnected time trackers hide the relationship between pursuit hours, phase budgets and the past-performance record you’ll need in the next Section F. Monograph connects pipeline, phase budgets, staffing, timesheets and billing in one place. Request a demo to see how Monograph helps engineering firms deliver on their promises.

Frequently asked questions

How do I know if my engineering firm is ready for QBS-based proposal writing?

You’re ready when you can hand an evaluator a current SF330 Part II for every office, dated project sheets with a reference contact for each, and a go/no-go record from the last year. Most firms aren’t there. The 46th Deltek Clarity study found 49% of firms report a formal BD process, and the top quarter of firms won 65.5% of pursuits against 34.4% for the bottom quarter. PSMJ’s BD research found firms where most staff get formal BD training reported the highest hit rates, yet 66% had no training program. Start with the Part II and the project sheets, since boards won’t consider a firm without a Part II.

What’s the biggest mistake engineering firms make in proposal writing?

Submitting something that fails a mandatory requirement, because it never gets scored. The next biggest is writing about your firm instead of the client’s project. SMPS loss-debrief interviews from 2017 found principals who talked about themselves more than the client’s needs, rushed preparation, and tunnel vision on the RFP text that missed the hot-button issues. A 2013 Hinge study of 522 buyers and sellers of design and construction services found 22% of buyers avoid firms that broke promises and 13% avoid firms indistinguishable from their competitors. Write the approach for the person who owns the problem, and keep the team you named.

How can I improve my firm’s Go/No-Go decision process?

Score it on paper before anyone argues, and make the scoring hard to game. Zweig Group notes that subjective forms get cooked into a go by whoever wants the project. PSMJ’s Kennedy/Jenks form asks for three reasons the client would pick you, three weaknesses and a one-sentence win strategy, scores six criteria 0–5 where 22–30 means go, and leans toward no-go when three or more of its true/false screening questions come back false. SMPS suggests win-probability bands of 75–100% high, 25–75% medium and 0–25% low, and one firm reported a 68% hit rate after a year of consistent use. Tie the form to capacity too, because a two-person team chasing six RFPs at once produces six mediocre proposals.

Should I invest time in proposals for projects outside my firm’s core expertise?

No, unless you’re joining a team led by a prime that has the experience. An SMPS owners survey lists firm size as one factor that eliminates firms before scoring. Lack of experience with the job type or geography also eliminates firms. ACEC’s Engineering Inc. argues that firms which learn to say no more often win more, and FMI calculates that at a 30% capture rate a firm must pursue $1.2 billion of work to land $350 million. The exception is a teaming arrangement under AIA C102–2015 or EJCDC E-580 where your discipline fills a gap on a prime’s team and your people appear on their comparable projects.

How do I handle consultant coordination requirements in engineering proposals?

Treat the subconsultants as part of your team from the first draft. Disclose the arrangement in the offer as FAR 9.603 requires, put each firm in SF330 Section C and the org chart in Section D, and have the teaming agreement signed with scope, cost sharing and a subcontract commitment before you submit. Require proof of compliant insurance before any payment, which BBG’s client advisory recommends. Track each consultant’s budget by phase and pull their bills into your invoices, so the coordination you promised in Section C shows up in the billing record the next evaluator asks about.