A UK CEO we spoke to last month had shortlisted three agencies for a £280k custom platform build. His CTO was uneasy about all three. She could name specific things that felt wrong (case studies looked like wireframes, proposals were vague on IP ownership, one agency refused to name the engineers) but could not put a framework around it. She ended up recommending walking from all three and restarting the search. Her CEO wanted to trust her instinct but needed a defensible framework.
That is the agency evaluation conversation across UK and US businesses in 2026. Sales polish across the agency market is high. Real capability varies widely. Buyer intuition often catches problems before the framework does. A structured red-flag checklist turns intuition into defensible decision-making.
This article is a candid checklist for CEOs, CTOs, and heads of digital evaluating software agencies. Fifteen warning signs across pitch, proposal, and delivery stages. What each sign actually means. What to do when you spot multiple. Written from the buyer's side, though we are ourselves a software agency.
The 15 Warning Signs (Across Pitch, Proposal, Delivery)
Fifteen signs grouped by evaluation stage.
Pitch stage red flags.
1. Cannot show comparable production work. Agency pitches capability but portfolio has nothing similar to your project. Comparable domain, comparable technology, comparable scale required for meaningful reference. Below this, they are guessing.
2. All AI/blockchain/Web3 buzzwords with no depth. Every slide mentions the trend of the day. Technical questions return vague answers. Signals shallow capability disguised as broad expertise.
3. Refuses to name specific engineers before contract. Sales team dominates every conversation. Engineers appear briefly at pitch or not at all. Cannot get commitment on which engineers will actually work on your project. Predicts you will get whoever is available on kickoff day.
4. Case studies show wireframes not shipped work. Portfolio images are design mockups or product screenshots without live URLs. Cannot demonstrate live products they built. Sometimes indicates they contributed one small part of a larger project.
5. Pushes fixed-price on unclear or evolving scope. Scope is exploratory but agency insists on fixed-price with detailed breakdown. Either optimistic (planning to lose money) or planning to fund the gap through change requests. See our earlier post on fixed-price vs T&M.
Proposal stage red flags.
6. Cost breakdown missing or vague. Proposal shows total cost without breakdown by workstream, engineer type, or phase. Cannot compare like-for-like across agencies. Prevents meaningful negotiation.
7. No mention of who owns IP. Contract silent on IP ownership. Should be explicit: work-for-hire, full IP transfer on payment, defined open source policy. Silence creates legal risk downstream.
8. Non-negotiable payment terms. Agency will not adjust payment schedule, milestone structure, or termination clauses. Signals inflexibility that will surface again during delivery.
9. Timeline unrealistically short. Agency quotes 4 months for work that comparable projects take 8-10 months. Either underscoped (change requests coming) or planning to under-deliver.
10. No scope of what's excluded. Proposal describes what's included but never what's excluded. Every unspecified item becomes a change request. Explicit exclusions demonstrate thoughtful scoping.
Contract and delivery stage red flags.
11. Escalation path unclear. When things go wrong, who gets involved from leadership. Agency without documented escalation path escalates through chaos. Ask specifically.
12. Skip technical discovery before contract. Agency signs contract without technical discovery meeting or documented understanding of your specific systems. Predicts assumptions that will surface as change requests.
13. Change requests handled vaguely ("we'll figure it out"). Ask how change requests are quoted, approved, and priced. Vague answers predict fights during delivery.
14. Communication only through sales, not engineering. Weekly calls attended by sales lead and account manager. Engineers never on call. Predicts insulation between you and the people doing the work.
15. Onshore team promised, offshore team delivered. Sales pitch emphasises local team. Contract signed. Delivery team is entirely offshore. Common bait-and-switch across some UK and US agencies. Ask specifically who will be on your project by name.
The GOV.UK procurement policy notes publish public sector supplier evaluation guidance that reinforces most of these signals as universal warning patterns. The Gartner IT services research publishes ongoing analysis showing these exact patterns predict failed engagements across sectors.
What These Signs Actually Mean
Red flags translate to specific downstream risks.
Capability red flags (signs 1-5) mean the agency likely cannot deliver what they claim. Downstream: project delivered wrong or delayed 3-9 months.
Commercial red flags (signs 6-10) mean the commercial relationship will become contentious. Downstream: cost overrun 30-60 percent, change request wars, potential legal dispute.
Execution red flags (signs 11-15) mean the working relationship will erode. Downstream: communication breakdown, delivery slippage, potential relationship termination.
Any red flag category signals real risk. Multiple categories together compound risk multiplicatively, not additively.
What to Do When You Spot Multiple Red Flags
Response depends on number and severity of red flags.
5+ red flags: walk immediately. Agency demonstrates fundamental problems in capability, commercials, or execution. No amount of negotiation fixes fundamentals. Restart your search.
3-4 red flags: negotiate specific concerns. Raise each red flag explicitly with the agency. Ask them to address it in writing. Watch how they respond. Constructive response (they acknowledge, explain, adjust) signals partnership potential. Defensive response (they dismiss, deflect, or promise vaguely) signals continued problems. Walk if response is not constructive.
1-2 red flags with otherwise strong signals: consider small trial. Some red flags are individually manageable if the rest of the evaluation is strong. Consider a small trial engagement (single-milestone fixed-price, £30k-£60k scope) to test working relationship before committing to major engagement.
Any single red flag on non-negotiable contract terms or onshore-promised-offshore-delivered. Enough on its own to walk. These specific patterns predict trust-destroying execution failures.
Never proceed with major engagement on unresolved red flags. Trust the framework over the sunk cost of evaluation time. Restarting the search costs weeks; wrong agency engagement costs months.
What We Learned Across Client Engagements
Across 50-plus client engagements at WhiteStone, three patterns recur.
Buyers who walk from red flags find better fits within 4-8 weeks. The perceived cost of restarting the search feels high but rarely proves so. Better-fit agencies exist; disciplined evaluation surfaces them.
Buyers who proceed despite red flags almost always regret it within 6 months. Delivery problems surface. Commercial disputes emerge. Working relationship deteriorates. The red flags predicted exactly what happened.
Structured red-flag evaluation protects both parties. Buyers avoid bad engagements. Good agencies get selected over polish-heavy competitors. We prefer buyers who evaluate systematically because it surfaces our actual strengths against pitch-heavy alternatives. See our earlier post on how to choose a software development partner for the positive-framing companion checklist.
You can see our shipped work at our portfolio. If you want structured vetting of your agency shortlist, book a vetting session with WhiteStone.
Common Failure Modes
Three failure modes we see repeatedly.
Trusting agency intuition over checklist. Something feels off but buyer cannot articulate. Proceeds anyway. Feeling was correct; six months later the engagement fails.
Discounting single red flag as coincidence. Red flag noticed, dismissed. Turns out to be systemic. Six months later, same issue at scale.
Committing before addressing red flags in writing. Verbal reassurances given at pitch stage; nothing changes in contract. Contract terms govern the actual relationship; pitch conversation does not.
Frequently Asked Questions
Is "AI powered" in an agency pitch a red flag?
Only if the agency cannot demonstrate production AI capability beyond chatbots. Real AI shops ship RAG systems, agents with tool use, and fine-tuned models. Wrapper factories describe capability without production reference. See how to vet an AI development company for specific 10-question framework.
How do I check if an agency actually built its case studies?
Ask for live product URLs. Talk to the client contact who commissioned the project. Ask what the agency delivered vs the client team or other vendors. Big projects involve many parties; be sure the agency's contribution matches their claim.
What proposal terms signal a bad agency?
Cost breakdown missing, no IP ownership clause, non-negotiable payment terms, unrealistically short timeline, no explicit exclusions from scope. Any three together suggest fundamental proposal problems. Any single non-negotiable clause is enough to negotiate hard or walk.
How do I test agency communication style before signing?
Insist engineers attend at least one call before contract. Ask specific technical questions and observe how they respond. Ask about change request handling in detail. Ask who gets involved when things go wrong. Vague answers at pitch stage predict vague accountability during delivery.
Should I proceed if I spot 1-2 red flags but the agency otherwise looks strong?
Consider a small trial engagement (single-milestone fixed-price, £30k-£60k scope) to test working relationship before committing to major engagement. Any red flag on non-negotiable contract terms or onshore-promised-offshore-delivered is enough to walk on its own, regardless of other signals.
The One Thing to Remember
Fifteen specific warning signs across pitch, proposal, and delivery stages predict failed agency engagements with high reliability. Buyers who use structured red-flag evaluation avoid bad engagements. Buyers who proceed despite red flags almost always regret it within 6 months. The perceived cost of restarting the agency search feels high but rarely proves so; the actual cost of the wrong agency engagement is materially higher.
If you want structured vetting of your agency shortlist, browse our custom software development services or come to the vetting session.
