Manual MVP
Concierge MVP Explained — Deliver Value by Hand First
A concierge MVP is not a half-built product. It is you (or a tiny team) personally delivering the result the product would eventually automate — spreadsheets, emails, calls, and all — so you learn whether the outcome matters before you invest in software.
Last updated: September 6, 2026
Direct answer
What is a concierge MVP?
In lean practice, a concierge MVP means the founder manually performs the service end-to-end for early customers. The customer gets the outcome; you get high-resolution learning about steps, edge cases, and willingness to continue — without pretending the system is already automated.
Key takeaways
What to remember
- Concierge answers “does this outcome create value?” — not “can we scale it yet?”
- Be explicit that parts are manual. Hidden labor that looks like a product creates distrust later.
- Cap the number of customers. Concierge that never ends is consulting, not validation.
- Write down every manual step. Those notes become the automation backlog — in priority order from real friction.
- Charge something when you can. Free concierge often attracts curiosity instead of buyers.
Not the same as
Concierge vs fake door vs prototype
Pick the instrument that matches the risk you are testing.
Fake door / landing test
Measures whether people reach for a promise. Little or no delivery. Use when demand is the unknown.
Concierge MVP
Delivers the outcome by hand. Use when you need to learn the workflow and whether people keep coming back.
Narrow technical prototype
Automates the hardest technical slice. Use when feasibility is the unknown, not the customer journey.
Wizard of Oz
Looks automated to the user while humans operate behind the curtain. Overlaps with concierge; still require honesty about timelines.
How to run it
Six steps to run a concierge MVP
Keep the promise narrow enough that you can deliver it this week.
- 1
Name the single outcome
One sentence: “For [who], we will produce [result] within [time].” If the sentence needs three “and”s, cut scope.
- 2
Recruit a handful of right customers
Three to five ICP-matched people who already feel the pain. Prefer people who would eventually pay.
- 3
Deliver end-to-end manually
Use whatever tools you have. Time yourself. Note every place you improvise — those are product requirements in disguise.
- 4
Ask for a next cycle
After delivery, ask whether they want it again next week and under what price or condition. Repeat behaviour beats one thank-you.
- 5
Separate delight from dependency
If they only love it because you personally babysit every edge case, you have not found a product yet — you found a service.
- 6
Decide what to automate first
Automate the step that is both painful for you and frequent for them. Do not automate the rare custom flourish.
What you should learn
Evidence a concierge run should produce
If you finish a concierge cycle and only have “they seemed happy,” the test failed — even if the customer smiled.
- Which steps the customer insists on vs which they ignore
- Where they get stuck without you
- Whether they initiate a second cycle unprompted
- What they compare you to (current workaround and its cost)
Checklist
Copy-ready concierge checklist
- Outcome sentence written and shared with the customer
- Manual nature disclosed up front
- Customer count capped (and written down)
- Timebox for the first delivery cycle
- Log of steps, minutes, and surprises after each run
- Ask for a repeat / pay condition after delivery
- Stop rule if nobody wants a second cycle
- Automation candidate list ranked by frequency × pain
Common mistakes
What usually goes wrong
Building tooling “just for the concierge”
You accidentally start the product build. Stay on email and spreadsheets until the outcome is proven.
Serving anyone who asks
Off-ICP customers teach the wrong workflow. Say no early.
Never naming a price
Free forever hides whether the outcome is valuable enough to fund the eventual product.
Automating before the second cycle
One successful delivery is anecdote. Two or three cycles start to show a pattern.
Hiding that it is manual
Customers feel cheated when the “app” later gets worse. Honesty preserves trust.
Continue or stop
After a few cycles
Keep going when
Customers request another cycle, describe the outcome in their own words, and tolerate rough edges. Then automate the highest-friction repeated step.
Pause or pivot when
Nobody wants a second cycle, or every delivery requires a unique custom path you cannot describe in five steps. The outcome may not be a product — or not for this ICP.
Sources
Where these ideas come from
- The Lean Startup principles (Eric Ries) — Concierge MVP as a learning vehicle before automation.
- Steve Blank — Customer Development — Get out of the building; learn the customer’s process before scaling a solution.
With Startup MRI
When the report points here
If build complexity or founder-fit risk is high, Startup MRI often recommends a narrow or manual path first. Use the report’s primary risk to decide whether concierge, fake door, or interviews should come first — then log evidence on the same report.
Analyze my idea →FAQ
Questions founders actually ask
Isn’t concierge just consulting?
It becomes consulting when there is no timebox, no learning log, and no plan to productize. Concierge is temporary on purpose.
How many customers do I need?
Start with a handful you can serve well. Depth beats a long waitlist you cannot deliver for. There is no universal “statistically valid” number for this stage.
Do I need to charge?
Not always on day one, but a price conversation should happen early. Payment — even small — separates politeness from priority.
When do I start coding?
When the same painful step repeats across customers and you can describe the happy path without improvising every time.
How is this different from an MVP app?
An MVP app ships software. A concierge MVP ships the outcome with people. Both are valid; they test different risks.
How does Startup MRI help?
It surfaces whether your weakest risk is demand, delivery, or distribution. Concierge fits delivery and learning risks; pair it with the experiment workspace on your report.
Find the risk that should come before code
Startup MRI helps you see whether you should talk, test demand, or deliver by hand first — then bring evidence back into the same report.
Generate a Startup MRI report →