HN Explorer

Ask HN: Is an MVP the first thing a startup should build?

6 points by MaknMoreGtnLess ↗ HN
It seems to be widely believed that:

1. entrepreneurs shouldn't waste any time getting their ideas to collide with reality

2. Building an MVP, that helps visualize their ideas, should be the first thing a startup should build

i. Does this hold up to your experience?

ii. What are the pros and cons of this (MVP first ASAP) approach?

iii. Are there better approaches that have worked better for you?

16 comments

[ 3.6 ms ] story [ 46.4 ms ] thread
MaknMoreGtnLess
Happy to expand on this, but unless, the founders fit the exact beachead market the MVP is testing to help solve problems of, finding alternative ways to discover the beachead market than through an MVP, is a way more effective way to learn.

Definition of MVP: an actual product being built to "validate the product" with the beachead customers needs.

It's important to call out and define what an MVP means because other people might mean "cheapest market discovery tool" instead of "cheapest product validation tool" - very different things

gashmol
IMHO its better to start with a UI prototype. It's much faster and cheaper while still being effective.
MaknMoreGtnLess
> IMHO its better to start with a UI prototype

Right - I am assuming you are comparing to vs. starting with a backend heavy prototype where it could get very expensive before exposing to customers?

> It's much faster and cheaper while still being effective

Effective as in effective in getting customer feedback as to whether the product makes sense to them?

but what if there were faster, cheaper and effective ways to get customer feedback than a UI prototype for a fresh startup just a few days old?

karmakaze
This looks like a trap to me. An MVP whether it's functional or a UI prototype (i.e. has no or fake back-end) should be built as quickly and crudely as possible while maintaining a semblance of what it is used for.

When I see scope reduced from a functional minimal-product to a non-functional one. I usually see accompanied with that more time spent on producing a higher-fidelity UI-'prototype' which is clearly not a better use of time/resources.

MaknMoreGtnLess
I see your point and want to follow up:

> while maintaining a semblance of what it is used for.

and how do you know how it will be used for and by who without actually talking to users and understanding if, why and how they will actually use it?

said differently, would you agree that market risk (there's no one who cares about the product at all) is higher than the product risk (we can't build the product)?

karmakaze
I think describing what I mean (or think people in general often mean) by the terms will help explain this.

The main goal is to find Product-Market fit as a first priority. This is done through prototyping, mockups, user-testing, etc, etc. Often this will be considered/called 'making an MVP'. This is an ongoing iterative process and the phase ends when you know what the core function or value proposition of the product is and who it is for.

At that point you move to the phase of building a higher quality (hopefully still very feature limited) v1.0 product.

MaknMoreGtnLess
> The main goal is to find Product-Market fit as a first priority

Agreed.

I want to see whether you realize that in the "Product-Market fit", the only thing you control is Product - you can manage/change/affect it as a builder.

However, you do not control the Market.

So, would it stand to logic that reducing the thing that you do not control - Market Risk (there's no one who cares about the product, we are going to build, at all) should be much higher priority than the Product Risk (we can't build the product)?

karmakaze
> However, you do not control the Market.

Not entirely true. Part of finding Product-Market fit isn't only to shape your product, it's also to possibly redefine your target user as well. Often you will discover a use-case that is stronger than the one you set out with at the start with only minor changes to the original product idea.

gashmol
Overworking the prototype is a real risk, that's why experienced devs should work on it. Sometimes its enough to do it on paper. Whats important is to capture the product in an objective form as quickly as possible and then show it to users.

Don't you think that iterating on a software MVP is more wasteful than iterating on a prototype? By the time you finish to setup the dev environment for your MVP you could've finished the prototype. And you don't have to carry any technical debt going forward.

gashmol
I think there's should be a difference between customer feedback and user feedback. UI prototype is effective only in finding what users want. To get customer feedback you probably need a full fledged MVP and ask for money or other kind of commitment.

> but what if there were faster, cheaper and effective ways to get customer feedback than a UI prototype for a fresh startup just a few days old?

Like what?

nivertech
You need to find the fastest and cheapest way to de-risk your startup idea.

That can be MVP or it can be something else. For example if you're a FinTech lending startup, then you better prove that your loan algos/model are profitable, rather than building UI wireframes and mockups. You can do this by backtesting your models or building a simulation.

"The Lean Startup" book by Eric Ries is already 10-11 y/o, so it's not up-to-date and with lots of gaps.

Also, ignore most of the data-driven methods, as they require lots of data which will not be available for pre-launch product. Instead you need to build a theory about your potential users' needs, and focus on shipping something first.

I suggest to get familiarized with the following methodologies:

  - JTBD (Jobs To Be Done)
  - 5 Whys
  - GIST (Goals->Ideas->Steps->Tasks) method (partialy data-driven)
  - Google Ventures Research Sprint
  - Google Ventures Design Sprint
If you still decide to ship an MVP - focus on a single most important feature. Use nocode/lowcode tools if possible. Or use some addon/plugin/widget for an existing platform: i.e. a chatbot, Slack/Teams app, twitter bot, Chrome extension, etc.
MaknMoreGtnLess
> You need to find the fastest and cheapest way to de-risk your startup idea.

Agreed - but, if I were to break out startup risks into the following, which one is the highest?

- Market

- Product

- Channel

- Business model/unit economics

> "The Lean Startup" book by Eric Ries is already 10-11 y/o, so it's not up-to-date and with lots of gaps.

Can you mention just a few of the gaps?

> Instead you need to build a theory about your potential user's needs, and focus on shipping something first

What's the most effective way to discover:

- who these potential users are?

- what their needs are?

nivertech
1. Market, then Business model/unit economics. The latter is less problem for VC-funded startups - they assume it can be optimized/fixed later.

2. In many places it's just hand-wavy. Many case studies, but no formal methodology how to do Build->Measure->Learn feedback loop.

3. The needs and desires at a particular time/situations. Personas don't work. For example the same person may want a luxury meal in one situation and a hotdog in another. See JTBD.

MaknMoreGtnLess
> Market, then Business model/unit economics

Agreed and I recently mentioned it elsewhere: https://news.ycombinator.com/item?id=30318531

> For example the same person may want a luxury meal in one situation and a hotdog in another

Sure - but would it be wrong to say the same person may have different Personas depending on the context?

> In many places it's just hand-wavy

Do you feel there can/should be a better way to discover customers?

nivertech
> Sure - but would it be wrong to say the same person may have different Personas depending on the context?

Persona is Persona, Job is a Job (as in JTBD).

Products are answering Jobs, and the same Persona/Consumer may have multiple Jobs, depending on the context.

  Consumer->Job->Product
Naive solution is:

  Consumer->Product
but the JTBD method makes implicit—explicit and makes invisible—visible.
MaknMoreGtnLess
Fair. I just define Persona as "a specific way/why the consumer wants to use the Product"

So that strongly aligns with your Consumer->Job->Product

Here, you mention: https://news.ycombinator.com/item?id=30319105

> "The Lean Startup" book by Eric Ries is already 10-11 y/o, so it's not up-to-date and with lots of gaps.

Can you mention just a few of the gaps?

Interested to hear your thoughts.