Which SaaS design firm
earns the brief?

Most failed agency relationships start before the first call. The client reaches out with a vague idea and expects the agency to define the scope. Agencies that are good at selling will — but their scope will reflect their strengths, not your needs.

Before you contact anyone, write a one-page brief that answers:

• What is the product, and what does it do? (Two sentences max.)
• Who are the primary users? What do they need to accomplish?
• What's broken or missing right now? Where does the user experience fail?
• What does success look like in 90 days? What metric moves?
• What's the budget range and the timeline? Be honest — agencies will find out anyway.
• What does the internal team look like? Who owns product? Who owns engineering?

Agency portfolios are curated. Every case study shows the work at its best. Your job is to read between the lines.

• Green flag: The case study explains the problem before the solution. Agencies that lead with "the client needed X" rather than "we designed Y" understand strategy.
• Green flag: They show constraints — technical, timeline, budget, stakeholder. Real projects have friction. Portfolios without it were either easy projects or sanitised.
• Red flag: Beautiful screenshots but no outcomes. If a case study ends at launch with no mention of what happened next, ask why.
• Red flag: Every project looks the same. Visual consistency across wildly different products usually means the agency applies a house style, not a problem-driven process.
• Red flag: The portfolio is mostly consumer or B2C work, but they're pitching you on enterprise SaaS. The disciplines are different. Verify specific SaaS experience.

The first call reveals how an agency thinks, not what they know. Come with questions, not a brief to present. Let them ask you things.

• Ask them to walk through a project that failed or didn't go as planned. Listen for honesty vs. spin.
• Ask who will actually work on your project — the senior person on the call, or someone else?
• Ask how they handle disagreements about design direction with a client team.
• Ask what they need from you in the first two weeks to do good work.
• Ask what they're not good at. Good agencies know their limits.

For engagements above £15k / $20k, always run a paid discovery sprint before signing the full contract. A 2–4 week discovery phase lets both sides evaluate the working relationship before committing to a longer term.

Good discovery output includes: a problem definition document, a research synthesis, a set of design principles, and a recommended scope for the next phase. If an agency resists paid discovery in favour of going straight to a long contract, treat that as a warning sign.

• IP ownership: Confirm you own all deliverables outright on final payment — source files, design tokens, component libraries, everything.
• Kill fee: If you need to pause or cancel, what do you owe? If the agency exits, what do they owe you? Both should be defined.
• Revision rounds: Contracts that specify "unlimited revisions" create perverse incentives. Negotiate a clear number of structured feedback rounds instead.
• Handoff definition: "Delivered" means different things to different agencies. Specify: Figma files with components, documented design tokens, annotated specs, and one handoff session with engineering.
• Confidentiality: You should have NDA protection from day one, including any preliminary conversations before the contract is signed.

A defined deliverable, agreed price, agreed timeline. The agency delivers X; you pay Y. On paper, the most straightforward model.

• Best for: Greenfield projects with a clear, stable brief — a new onboarding flow, a marketing site redesign, a design system from scratch.
• Risk: Scope creep. Once the brief shifts, the agency is entitled to charge more or deliver less. Fixed-scope works best when you have a complete brief before the contract is signed.
• Watch out for: Agencies that will only work fixed-scope. It often means they've been burned by open-ended engagements and have tightened the model to protect themselves — not you.

You pay for hours worked at an agreed rate, with no fixed scope. The engagement expands and contracts with your needs.

• Best for: Ongoing product work where priorities shift week to week, or early-stage discovery where the scope can't be defined in advance.
• Risk: Budget creep without strong project management on your side. You need someone tracking hours and output, not just hours.
• Watch out for: Agencies that pad T&M with process overhead — meetings about meetings, status reports that take half a day to produce. Request a weekly summary of hours by activity.

A fixed monthly fee for a set number of design hours or a defined set of outputs. You get reliable capacity; the agency gets revenue predictability.

• Best for: Teams that have a steady, ongoing flow of design work but don't need a full-time in-house designer yet. Retainers are effective for post-launch product iteration.
• Risk: Unused hours rarely roll over, and agencies may deprioritise retained clients when a larger project lands. Negotiate SLA-style commitments to response time and output.
• Watch out for: Retainers where the deliverable is "availability" rather than output. Define what a month's work should produce — otherwise you're buying peace of mind, not design.

One or more agency designers work inside your product squad, alongside your engineers, attending your standups, using your tools. Organisationally, they function like employees.

• Best for: Companies that want external design talent but need them fully integrated with engineering. Works well for Series A–B companies that are scaling too fast to hire.
• Risk: The embedded designer may over-adapt to your culture and lose the outside perspective you were paying for. Set aside time for them to bring ideas in from outside the team.
• Watch out for: Agencies that embed a junior designer and keep the senior one for account management. Confirm the seniority and continuity of whoever will be in the room.

A structured 1–2 week intensive, usually based on the Google Ventures model: problem definition, ideation, prototyping, and user testing in rapid sequence.

• Best for: Validating a new feature or product direction quickly without committing to full design and build. Useful before a fundraise or a major roadmap decision.
• Risk: Sprints create momentum and outputs that can feel conclusive but aren't. The prototype tested in a sprint is not production-ready. Be clear about what happens after.
• Watch out for: Sprint-as-sales-tool. Some agencies use sprints as a low-friction way to start a relationship, then up-sell. That's not inherently wrong, but go in knowing it.

Start with the question: how well-defined is what you need?

• Very well defined, one-time → Fixed-scope
• Evolving, ongoing product work → Time & materials or retainer
• Scaling fast, need integration → Embedded team
• Uncertain direction, need validation → Sprint first, then decide

The handoff is where most agencies under-deliver, and most clients under-specify. Before signing, agree in writing that "delivered" includes all of the following:

• Figma (or equivalent) files with named, organised layers and pages — not a single flat file.
• A component library with documented variants, states, and usage notes.
• Design tokens for colour, spacing, and typography — in a format your engineering stack can consume.
• Annotated specs for all interactive components: what happens on hover, focus, error, loading, empty state.
• A live handoff session with your engineering lead — at least two hours, recorded.
• A 30-day post-handoff support window for questions, clarifications, and minor amendments.