What Is a Concierge MVP? Definition + Example
A concierge MVP tests a business idea by delivering the service to early customers manually, by hand, before any software exists. The customer knows the process is personal. The goal is to learn exactly what the solution must do before you spend a line of code automating it.
Quick Answer: A concierge MVP is a validation experiment where you personally perform the service for a handful of real customers — no product built yet — and they know it is being done by hand. It tests whether you can create value and reveals the workflow to automate later.
How a Concierge MVP Works: Deliver the Service by Hand
A concierge MVP works by making you the product. Instead of building software, you sit alongside a few early customers and deliver the outcome yourself, step by step, the way a hotel concierge personally handles each guest's request. There is no automation behind the scenes and no pretense that there is.
The customer knows it is manual. This is the defining trait. You are transparent that you are doing the work by hand — often in person, over email, or on a call. That openness is a feature, not a compromise: it lets you ask questions, watch reactions, and adjust in real time.
The point is learning, not scale. Eric Ries frames the concierge MVP in The Lean Startup as a way to gather validated learning cheaply. By serving customers one at a time, you discover what the solution genuinely needs to do — which steps matter, which you imagined but nobody wants — before committing engineering effort.
You serve a tiny number of customers. A handful is enough. The manual approach cannot scale, and that is the point: it keeps effort low while the questions are still open. When patterns repeat across customers, you have found the workflow worth automating.
Because it delivers real value to real users, the concierge MVP is one of the delivery-focused options in the broader menu of MVP types — the ones that test whether you can create value rather than whether people merely want it.
A Concrete Concierge MVP Example: Food on the Table
The clearest example comes from The Lean Startup itself: a company called Food on the Table, which planned to build software that generated weekly meal plans and shopping lists tuned to what was on sale at a shopper's local grocery store.
They started with one customer, served entirely by hand. Rather than build the recommendation engine first, the founder personally visited a single customer each week. He learned her family's food preferences, walked her local store's aisles to check prices, and hand-built her meal plan and shopping list himself.
No software existed yet. Every task the app would eventually automate — checking sales, matching recipes, assembling the list — was done manually, and the customer knew it. That intimacy surfaced needs the team would never have guessed from a spec document.
Learning came before automation. Only after repeating this with more customers, and seeing which steps mattered, did the team encode what they had learned into actual product features. The manual phase was the research; the software followed the evidence.
The lesson is transferable. Whether your idea is meal planning, financial advice, or resume editing, you can often deliver the promised outcome by hand first — see our guide on how to deliver a concierge MVP manually.
Concierge MVP vs Wizard of Oz MVP: The Key Difference
Both approaches deliver value manually with no product built, so they are easy to confuse. The difference is entirely about what the customer knows. The table below contrasts the two on the traits that matter most.
| Dimension | Concierge MVP | Wizard of Oz MVP |
|---|---|---|
| Who does the work | You, by hand | You, by hand |
| Does the customer know | Yes — the process is openly manual | No — it looks like automated software |
| Front end shown | A person, not a product | A polished, product-like interface |
| Primary question | What does the solution need to do? | Do people want the automated experience? |
| Relationship | Personal, consultative | Arm's length, as with real software |
| Best for | Learning an unknown workflow | Testing demand for a known workflow |
Takeaway: A concierge MVP is honest about being manual; a Wizard of Oz MVP hides the manual work behind an automated-looking front end. Choose concierge when you are still learning what to build, and reach for a Wizard of Oz MVP when you already know the experience and want to test whether people will use it as if it were finished software.
When a Concierge MVP Is the Right Test
A concierge MVP fits best when the biggest unknown is not whether people want the outcome, but what delivering that outcome actually involves. Reach for it in these situations:
- The workflow is unclear. You know the destination but not the steps. Serving customers by hand maps the real process before you hard-code assumptions into software.
- The service is high-touch or complex. When value depends on judgment, personalization, or nuance, doing it manually teaches you where that nuance lives.
- You have few customers and time to serve them. The method does not scale, so it suits early-stage founders who can afford to go deep with a small group.
- Automation would be expensive to guess at. If building the wrong engine would cost months, a manual phase de-risks that investment cheaply.
It is a poor fit when your open question is demand. If you mainly need to know whether anyone wants the product at all, a cheaper demand test — a landing page or smoke test — answers that faster without your hands-on time.
Key Takeaways
- A concierge MVP delivers the service manually, by hand, with no product built yet. You are the engine while you learn what the software will need to do.
- The customer knows it is manual. Transparency is the defining feature and the source of rich, real-time learning.
- Its job is to test value delivery and reveal the workflow. The goal is validated learning about what to build, not scale.
- Food on the Table is the canonical example. Eric Ries describes serving one customer entirely by hand before writing any recommendation software.
- It differs from Wizard of Oz by disclosure. Concierge is openly manual; Wizard of Oz hides the manual work behind an automated-looking interface.
- Use it when the workflow is the unknown. Choose a cheaper demand test instead when you only need to know whether people want the outcome.
Frequently Asked Questions
Is a concierge MVP the same as a Wizard of Oz MVP?
No. Both deliver value manually with no finished product, but they differ in what the customer knows. In a concierge MVP, the customer is aware the service is being done by hand and personally. In a Wizard of Oz MVP, the manual work is hidden behind a front end that looks like automated software, so the customer believes it is a real product.
Do you need to write any code for a concierge MVP?
No. A concierge MVP requires no software at all — that is the entire point. You deliver the outcome using whatever manual tools work: conversations, email, spreadsheets, or in-person visits. Code comes later, only after serving customers by hand has shown you which steps genuinely need automating and how the workflow should behave.
How many customers do you need for a concierge MVP?
Just a few. Because the work is manual and cannot scale, a concierge MVP typically serves a small handful of customers — sometimes starting with a single person, as Food on the Table did. The aim is depth, not breadth: you want enough repetition to spot patterns in the workflow, not a large sample for statistical confidence.
What does a concierge MVP actually test?
A concierge MVP tests whether you can create real value for a customer and, crucially, what delivering that value requires. It answers questions about the workflow — which steps matter, what personalization is needed, where the effort concentrates — so you learn what the eventual product must do before you spend time and money building it.