Most agency proposals are beautiful PDFs engineered to get a signature. 92 of 100 had the same 8 structural flaws. Here is what to find before you sign.
Ninety-two of the last 100 agency proposals we reviewed had at least four of the same eight structural problems. Not variations on a theme. The same problems, repeated by agencies of every size. The proposal is not a service agreement. It is a sales document engineered to produce a signature. Once you know the eight patterns, you will not be able to unsee them.
A clean proposal takes ten minutes to evaluate. Below is the teardown, the eight patterns, and what a proposal looks like when it is actually written to serve the client. We will tell on our entire industry. The buyers this post saves are worth more to us than the deals it costs us.
We reviewed 100 proposals sent to founder-led businesses between $500K and $10M in annual revenue. The proposals came from agencies we competed against, agencies our clients left before hiring us, and agencies that prospects had received before calling us. These are the eight patterns.
Seventy-eight of 100 proposals quoted hours or sprints instead of named deliverables. "160 hours of development" is not a scope. It is a blank check written to the agency. If the deliverable takes 80 hours, you paid double. If it takes 240 hours, you get a change-order invoice.
A deliverable-based scope says: "Landing page — 5 sections, mobile-responsive, integrated with HubSpot, delivered by May 31." An hour-based scope says: "Landing page — 20 hours design, 40 hours development." Spot the difference in 15 seconds.
We tracked 16 projects over 18 months where the client contracted on hours. The average overrun was 47%. The average client absorbed it without a formal dispute because they had no contractual basis to object. They paid for hours. They received hours.
Sixty-one proposals had deliverables with no definition of done. "Brand refresh" is not a deliverable. "Brand guidelines: logo in 3 formats, 2 color palettes, 2 typefaces, usage rules PDF, delivered by June 15" is a deliverable.
Without a definition of done, the agency decides when the work is complete. They decide it is complete when they run out of hours. You disagree. The next invoice appears.
Fifty-four proposals included a paid discovery phase before any tangible deliverable. Prices ranged from $4,500 to $18,000. The typical output: a strategy brief and a competitive audit. Both exist to justify the next engagement phase. Neither is an asset you can use anywhere else.
In 2026, with public-by-default companies and AI-assisted research, an agency that needs $12,000 to understand your market has not built a scalable intake process. Real discovery is two calls and a shared document. Productized agencies do not charge for it.
The tell: ask the agency what you receive at the end of discovery and whether you keep it if you do not proceed. Agencies running discovery-as-revenue will hesitate on the second question. Agencies with real discovery processes answer immediately.
Forty-nine proposals defined success as impressions, sessions, followers, open rates, or "brand awareness." Zero of those forty-nine included a dollar figure, a lead count, a close rate, or a revenue target.
Output is activity. Outcomes are dollars. If the agency's success criteria cannot be tied to revenue, the agency is not accountable for revenue. You are buying momentum that looks good in a monthly deck and leaves the revenue line flat.
Forty-seven proposals buried cancellation terms in footnotes or appendices. The specific language: 90-day notice periods, hours forfeiting at termination, work-product withheld until "all outstanding invoices are cleared." One proposal required 120 days notice to cancel a six-month engagement.
The cancellation clause is the most important clause in any agency contract. It is the only clause that matters when the relationship breaks down. If you cannot find it in the first three pages, find it before reading anything else. The agency buried it deliberately. That tells you something about the agency's intentions.
Forty-three proposals did not disclose who would actually do the work. The proposal featured senior partners and the founding team. The engagement would be staffed by junior associates and offshore contractors. This is standard practice: sell with the senior team, deliver with the junior team.
One question solves this: "Who specifically will work on my account, and can I speak with them before signing?" Agencies with strong delivery teams name the people. Agencies running the bait-and-switch say "we have a dedicated team of experts."
Thirty-eight proposals led with case studies that did not match the buyer's scope, industry, or budget. Enterprise results for SMB buyers. SaaS metrics for e-commerce companies. The case study exists to manufacture credibility, not demonstrate relevant experience.
A useful case study has four elements: client industry (even anonymized), scope in named deliverables, timeline, and one specific before-and-after metric. Everything else is decoration. If the agency cannot show you one case study that resembles your situation, ask why before scheduling the next call.
Thirty-one proposals submitted pricing in a format impossible to compare to competing proposals. All-in monthly retainers. Bundled phase totals. Project fees with no line-item breakdown. The obfuscation is deliberate. It prevents you from calculating the cost per deliverable.
Ask for a line-item breakdown: one row per deliverable, price per row. Agencies that refuse are protecting margins. Agencies that comply have a productized practice with the internal discipline to put numbers next to names.
Two structural incentives explain all eight patterns.
First, most agencies bill by the hour. Every unclear scope item is a future billable hour. Clear proposals reduce future billable hours. The incentive points away from clarity.
Second, proposals are evaluated on aesthetics before contents. The polished deck with custom iconography and five case study slides wins more deals than the two-page deliverable list with fixed prices. Buyers are trained by the market to equate presentation quality with service quality. They are not the same thing.
The eight patterns are not accidents. They are optimizations for a buyer behavior the industry has learned to exploit. The fix is to evaluate proposals in the wrong order: cancellation clause first, scope second, pricing third, case studies last. Most buyers do the opposite. That is why 92 of 100 proposals get signed.
Here is the side-by-side comparison of a weak and a strong proposal for the same scope — a 10-page marketing site plus a three-month paid ads launch.
The weak proposal was signed 3-to-1 over the strong proposal in our sample. It had better design and more slides. It was also almost impossible to hold accountable. That is not a coincidence. That is the business model.
Use this sequence on every proposal you receive.
Score 5 of 7 or higher: worth a second call. Score 3 or lower: you are reading a sales deck. Do not sign a sales deck.
See what named-deliverable scope looks like in production at Striveloom's services page. That is the baseline we apply to every client.
Your next proposal will look like a $100 bill. Beautiful design. Impressive case studies. A confident pitch. And buried in appendix B: a 90-day cancellation clause, expiring hours, scope defined in sprint blocks.
Read the cancellation clause first. Count the deliverables by name. Ask for the line-item breakdown. Then look at the case studies.
The proposal is the first piece of work product the agency delivers. If it fails your 10-minute audit, the invoice will fail you too. The pattern that produces a sloppy proposal produces a sloppy engagement. The pattern that produces a clean proposal produces clean delivery.
See what fixed-deliverable pricing looks like at Striveloom's pricing page and compare it to what you received. The gap will be obvious.
The cancellation clause. It is the only clause that matters after the relationship goes sideways. Look for the notice period (30 days is fair, 60-90 is a trap), what happens to unused hours at termination, and whether work product is released before the final invoice is paid. If you cannot find the cancellation clause in the first three pages, ask the agency to move it to page two before you read anything else in the document.
Yes, in two cases: when the technical complexity is genuinely unknown and when the discovery output is a standalone asset you keep regardless of whether you proceed. A $5,000 technical audit you own is legitimate. A $10,000 strategy brief that exists only to justify the next phase is an entry ticket. Ask: what do I receive at the end of discovery, and can I take it to another agency if I choose not to proceed? The agency that hesitates on the second question is running discovery-as-revenue.
Ask the agency to convert each line item from hours to a named deliverable with a definition of done and a fixed price. Most productized agencies will comply because they already work this way internally. Traditional agencies will push back, which tells you their delivery is not systemized. If they resist but you still want to proceed, accept hours only with a per-line-item cap and a clause requiring written approval for any work exceeding that cap before billing.
Always. Ask for two current clients and one former client. The current clients tell you what delivery looks like now. The former client tells you how the agency behaves when the relationship ends — asset handoffs, final invoices, cancellation process. Agencies that refuse references are telling you the references would not be favorable. Agencies that provide current clients only have probably not maintained good relationships after exits.
At least one specific, measurable business outcome tied to a date. Examples: 250 qualified leads in 90 days, 3% conversion rate on the primary landing page, 40% email open rate on campaign one. Metrics like impressions, sessions, and brand awareness are output metrics, not outcome metrics. A good proposal lists both: output metrics the agency controls directly, and at least one business outcome metric they are committing to contribute toward.
Two to three, each closely matching your industry or scope. Quality beats quantity. One case study with a named client, a specific scope list, the exact timeline, and a before-and-after revenue or lead count is worth more than ten case studies with vague descriptions and stock photos. If the agency does not have a case study that resembles your situation, ask why before proceeding. A good agency knows the answer immediately.
Founder & CEO of Striveloom. Software engineer and Harvard graduate student researching software engineering, e-commerce platforms, and customer experience. Builds the agency that ships like software — one team, one pipeline, one platform. Writes on AI agencies, web development, paid advertising, and conversion optimization.
Most agencies bill hours so they make more money when projects take longer. Here is the exact pattern, the four contract clauses to delete, and a checklist that has saved buyers $40K on average.
Retainers were a discount mechanism. They became Stockholm syndrome. Here is the math, the four clauses to renegotiate, and the exit playbook for $120K of sunk-cost retainers.
Most agencies sell hours and activity. This is how to find the rare agency that sells outcomes and how to write a contract that actually enforces it.
Book a free 30-minute call to scope your project. Fixed pricing, transparent timelines.
| Element | Weak proposal | Strong proposal |
|---|
| Scope | "Website design and development" | "10 pages: home, about, 3 service, pricing, 3 case study, contact. Next.js, mobile-responsive, HubSpot integrated." |
| Success criteria | "Increased online presence" | "250 MQLs in 90 days, 3% form-conversion on primary landing page" |
| Pricing | "$42,000 project fee" | "$12,000 site build fixed; $3,500/mo ads flat; $2,000 per landing page fixed" |
| Milestone dates | "8–10 weeks" | "Design comps: Day 14. Dev complete: Day 35. Launch: Day 42." |
| Cancellation | Appendix B, 90 days | Page 2, 30 days, unused hours roll forward |
| Team | "Our team" | Three named operators, LinkedIn links, roles |
| Discovery | $8,000 paid phase | 2 onboarding calls, no charge |
