How to evaluate a productized agency: 12 questions to ask

If you're a B2B SaaS founder evaluating a productized agency for an MVP build, the marketing language across agencies looks similar. "Fixed price, fixed timeline, productized scope" appears on most agency websites in 2026. The actual operational discipline behind that language varies dramatically.

This piece is the evaluation framework: 12 specific questions to ask on the scoping call that distinguish real productized shops from traditional agencies in a productized hat. Real productized agencies answer these questions consistently and clearly. Agencies that hedge on multiple questions are signaling how the engagement will actually run.

#1. Where are your prices published?

Real productized agencies publish prices on the website. Specific tier names, specific dollar amounts, specific scope per tier. "Standard tier: $14,800 for 6 weeks, includes auth, billing, six core flows, custom design system, admin panel, analytics."

Traditional agencies in a productized hat say "starting at $X" or "we'll send a custom quote after the discovery call." That's not productized pricing; that's traditional pricing with a productized marketing veneer.

The signal: if you have to wait for a custom quote to know the price, the engagement is structurally a traditional one. The price negotiation is the first scope negotiation.

For the broader cost framework, see How much does a SaaS MVP cost in 2026.

#2. What's the actual timeline, and what happens if it slips?

Real productized agencies quote a specific timeline (usually 4, 6, or 8 weeks) and have a clear escalation path if the build risks slipping. The escalation usually involves the agency absorbing the slip cost (eating margin to ship on time) rather than passing the cost to the founder.

Traditional agencies in a productized hat quote "approximately 6 weeks, depending on scope finalization." The "depending" is the tell. The timeline isn't fixed; it's a starting estimate that drifts.

The follow-up question to ask: "What happens specifically if you're behind in week 4?" A real productized agency has a documented escalation process (additional engineer added, scope re-confirmed with founder, weekend work absorbed by the team). A traditional agency says "we'd talk to you about it." The difference is delivery discipline.

#3. How do you handle scope creep mid-build?

The single highest-signal question. Real productized agencies have specific language for refusing scope creep and pushing requests to v1.1. They have a v1.1 list that captures requests so founders see their input is being heard, not ignored.

Traditional agencies in a productized hat say "we're flexible" or "we can absorb small changes." Both are anti-signals. Flexibility on scope is what turns a 6-week build into a 12-week one.

Ask the agency to walk through a specific scope-creep scenario. "If in week 3 I realize I need a seventh flow, what happens?" A real productized agency will quote the cost transparently ("that adds 4 days, $3,000, pushes ship to week 7; want to do it?"). A traditional agency will say "we'd see what we can do."

For the patterns themselves, see SaaS MVP scope creep: 8 patterns to refuse on the scoping call.

#4. What stack do you use, and why?

Real productized agencies have a vetted stack they reuse across engagements. They can articulate why each piece is in the stack and what the alternatives are. The stack is boring on purpose: well-documented, large hiring pool, predictable failure modes.

Traditional agencies say "we pick the right stack for each project" or "we're stack-agnostic." Both are anti-signals for a productized engagement. Stack flexibility means week 1 of every engagement is spent re-deciding architecture, which is week 1 not spent shipping.

The 2026 default productized stack is well-established: Next.js + TypeScript + Tailwind + Drizzle + Postgres + Stripe + Clerk + Vercel. For the deep dive on each choice, see The SaaS MVP tech stack for 2026. An agency that ships on a meaningfully different stack should be able to explain why.

#5. Who owns the code at the end?

Real productized agencies transfer the code to your GitHub organization at the end of the build. No "we host it for you," no "it lives in our repo," no "you license access." The repo is yours.

Traditional agencies in a productized hat sometimes hedge here: "we transfer ownership when we finish the project" (without specifying when "finish" means), or "you have access to the code but we maintain the deploy infrastructure." Both are vendor-lock patterns that compound over time.

The follow-up question: "What happens if we want to bring in another engineer to work on the codebase in month 3?" A real productized agency says "great, here's the runbook and the architectural docs; we're happy to onboard them in a 60-minute session." A traditional agency says "we'd need to discuss licensing terms."

#6. What's included in the post-launch coverage?

Real productized agencies ship with a defined post-launch fix coverage period (typically 30 days) for bugs in the v1 scope. The coverage isn't unlimited; it's specifically for issues with what was built, not for new features or scope additions.

Traditional agencies in a productized hat are vaguer here: "we offer maintenance retainers starting at $X/month." That's not post-launch coverage; that's an upsell.

The follow-up question: "If a bug surfaces in day 15 in code you shipped, what's the response?" A real productized agency says "we fix it within X business days at no charge if it's in the v1 scope." A traditional agency says "we'd quote a fix on a time-and-materials basis."

#7. What's NOT included that I might assume is included?

The clarity question. Real productized agencies have a specific "not included" list they walk through proactively. Common items: app store deployment, marketing site (if separate from the product), domain registration, third-party tool subscriptions, ongoing prompt iteration for AI features, content writing, real branding work beyond the visual design.

Traditional agencies in a productized hat are vague here, which means the founder discovers exclusions mid-build. Each discovery is a small re-negotiation that erodes the productized promise.

Ask explicitly: "Walk me through what's not included in the standard scope so I'm not surprised in week 4." A real productized agency has this rehearsed; a traditional agency improvises.

#8. How do you handle billing edge cases?

The technical depth question. Real productized agencies ship hardened webhook handlers for Stripe with idempotency, retry logic, dead-letter queues, audit logging. They've debugged the same five edge cases (card expired, charge disputed, plan switch mid-cycle, customer cancels and re-subscribes, signup-not-completed) 50 times before.

Traditional agencies in a productized hat ship with the basics. The webhook handler does signature verification and basic event handling; the edge cases get discovered in production by paying customers.

Ask: "Walk me through what your default webhook handler does for invoice.payment_failed." A real productized agency has a rehearsed answer that includes retry, dunning email, eventual downgrade or cancellation, audit logging. A traditional agency says "we'd handle it case-by-case."

For the deep dive on auth and billing edge cases, see Auth + billing for SaaS MVPs: Clerk vs Auth.js + Stripe in 2026.

#9. What does your handoff documentation look like?

The seriousness check. Real productized agencies ship with structured handoff docs: README in the repo, architectural overview, runbook for common operations (deploy, rollback, schema migration, common bug patterns), credentials and access transfer document, post-launch fix workflow.

Traditional agencies in a productized hat ship with whatever documentation the engineer happened to write during the build. Quality varies wildly.

Ask to see a sample handoff doc from a previous engagement (with the client name redacted). If the agency hesitates or can't produce one, the documentation discipline isn't there.

#10. Can I talk to a recent client?

The reference check. Real productized agencies have at least 3 to 5 clients willing to take a 15-minute reference call. The references will tell you whether the timeline held, whether scope discipline worked, whether the handoff was clean.

Traditional agencies in a productized hat have testimonials on the website but resist setting up live reference calls. The friction is the signal.

The questions to ask the reference: Did the build ship on the original timeline? Did the price match the quote? How did the agency handle scope-creep requests? Was the handoff documentation usable? How would they describe the agency to another founder considering the engagement?

#11. What's your escape clause if the engagement isn't working?

The risk question. Real productized agencies have a clear early-termination clause in the contract: defined milestones (end of week 1, end of week 2), pro-rated refund if either party wants to walk, code transferred up to that point.

Traditional agencies in a productized hat don't have this. The contract is a one-way ratchet; once you've paid the deposit, getting out is messy.

Ask explicitly: "What happens if at the end of week 2, we both agree the engagement isn't working?" A real productized agency has a documented answer. A traditional agency says "that's never happened."

#12. What kind of project would you turn down?

The scoping discipline test. Real productized agencies have a clear list of projects they don't take: scope outside the playbook (e.g., 12 flows in v1, native mobile, regulated workloads), founders without product-market-fit signal who really need a no-code prototype, projects with budget under the floor of the agency's tier structure.

Traditional agencies in a productized hat say "we'd consider any well-scoped project." That's not productized scoping; that's traditional sales-team scoping.

Real productized agencies turn down 50 to 70% of inquiries because they fit a different bucket. Ask the agency to describe a recent project they declined and why. The clarity of the answer is the signal.

#What the answers tell you in aggregate

After 12 questions, you'll have a clear pattern. Real productized agencies answer consistently: published prices, fixed timelines with specific escalation, refused scope discipline with v1.1 framing, vetted stack with reasoned alternatives, clean ownership transfer, defined post-launch coverage, specific not-included list, hardened billing edge cases, structured handoff docs, willing references, defined escape clauses, clear scoping discipline.

Traditional agencies in a productized hat hedge on multiple of these. Each hedge predicts how the engagement will actually run. The agency that hedges on scope discipline will hedge during the build; the agency that hedges on timeline will slip the timeline; the agency that hedges on ownership will create vendor lock.

One or two hedges is fine; productized agencies aren't dogmatic. Three or more hedges means the agency is a traditional shop with productized marketing language. The structural model isn't there even if the website implies it is.

#Bottom line

Productized agencies and traditional agencies in a productized hat look similar on the website and diverge sharply during the actual engagement. The 12 questions above surface the difference on the 20-minute scoping call before you've signed a contract or paid a deposit.

Real productized agencies welcome these questions because they confirm the productized model the agency has built around. Traditional agencies in a productized hat resist them because the answers expose the gap between marketing language and operational reality.

If you're evaluating SolvSpot for a B2B SaaS MVP build, ask all 12 questions on the scoping call. We'll answer each one specifically and quickly because the productized model requires it. For the broader productized vs freelance comparison, see Productized agency vs freelancer for a SaaS MVP.