HN Explorer

Ask HN: Interview questions for a better Product Hypothesis?

13 points by sdrinf ↗ HN
On Venture Hack's Customer Development presentation[1], Customer Discovery starts out with a product hypothesis.

However, as they say, it's "a capital mistake to theorize without data"; and I believe if the goal is limited to "creating a viable online business", and we have a target market set -say, "professionals doing professional stuff"[2]- we can do better.

By asking well-phrased, well-framed questions to a small (but statistically significant) partition of the target market, we can pull out product hypothesizes right from the market.

So, my question is: which questions to ask, if our goal is to limit the answers to have good scaling characteristics?

Why you should care: finding an answer to this problem would contribute to massively de-risking startup creation.

[1] http://www.slideshare.net/venturehacks/customer-development-4-customer-discovery-part-1

[2] http://www.kalzumeus.com/2010/03/20/running-a-software-business-on-5-hours-a-week/

8 comments

[ 3.3 ms ] story [ 27.7 ms ] thread
jamesshamenski
At the earliest stages, you should just be identifying the problem. Interview general folks in your industry and ask, 'would you like a simpler XXX'. 'what is the most painful ZZZ'. 'what would get you to try a new service'.

From that small set, draw up your idea and re-interview those folks. if they like it, you should present the same drawings to a wider scope and see if it still has gravity.

The next step is to test the market and publish your solution through adsense . I'm unaware of another method to test an idea as well as adsense. it would be great to hear some other ways of testing an audience before building.

sdrinf
You misunderstood the question completely. simpler XXX, and painful ZZZ _presumes_ an existing problem hypothesis. However, past data points of both my own, and most startupers clearly shows, that our problem hypothesises generally suck at conceptual level.

This is what I mean by de-risking: I want to take out this unnecessary risk of blind guessing, and delegate hypothesis discovery to those _with the data_.

endtime
I didn't understand your question either. It feels very...jargon-heavy.

It sounds like your question is: if you know who you want to sell to, can you just ask them what they want rather than polling them for feedback on different ideas? Honestly, I think people often don't know what they want because they don't understand which things can be made simpler. That's why someone with a different perspective (hackers) can learn about a market, have some new insight, and come up with something disruptive.

sdrinf
you're right in the disparity of understanding between customers, and service providers.

More generally, you can't ask potential customers about "what they want", for reasons that are many.

On the other hand, professionals (target market for above)

1, do know what their own pain points are; and

2, have a firm understanding on how much it impacts their pocket;

(Note, that I'm talking about professionals here, and not "people in general")

Harvesting these informations with the added constraint of "software solutions", can lead to much better educated guesses on what would be worth for future investigations.

lnp
> learn about a market, have some new insight, and come up with something disruptive

This, how to this without the magic?

What are the (universal) questions you would ask to potential customers to come up with a problem worth solving?

sdrinf
This question constrains the possibility space by 2 factors:

-It has to tap into the "common headaches" of the target market (say, barbers, teachers, or small restaurant owners), so scalability is given

-It constrains the headaches to those, that are mallable to software solutions

Some early candidates:

"What aching problems do barbers have, that would be a computer's job?"

"In a futuristic society, what things would you except computers to do, that you're currently handling manually -but want to get rid of?"

skmurphy
In "Four Steps to the Epiphany" Steve Blank defines an earlyvangelist as "someone who has spent money to develop a solution to a problem but is still unhappy with the result." I would use this to model you to identify which professionals to have a further conversation with: "what problem(s) have spent time or money on in the last year to reduce or eliminate where you are still unhappy with your current solution?"

I would not worry about the "scaling characteristics" as much initially as finding a group of professionals who share the problem and would reference each other's decision to use your product. This does not mean that they have to know each other but that an endorsement, case study, or testimonial that describes their background and how they benefited would influence a buy decision. All innovations start in a niche, some go scale beyond that, but none scale without starting in a niche.

For an elaboration on sources of innovation see http://www.skmurphy.com/blog/2008/02/15/paul-grahams-six-pri... which includes checklists from Paul Graham, Peter Drucker, Gary Hamel, and Bob Bemer the inventor of ASCII.

lnp
From linked slides it seems, that at the beginning is listening. And for that the best questions are:

1. What is X all about? 2. What I need to know to really understand X?

It also seems, that Customer Development is not about finding the perfect product hypothesis in one shot, on the contrary - the initial hypothesis is just the starting point, things you learn on the way is what matters. Just like with dancing, it will be better after some dancing, no need for over-thinking it or looking for a perfect steps. It's more about expressing yourself.

That said optimization and automation, a heuristic or even better an algorithm, that will reliably and repeatably take some customers and output brilliant business ideas would be a nice thing to have. But it won't reduce startup risk, not massively for sure, because 'executing' alone is more risky than 'executing on a wrong idea'.

Going back to questions.

3. What do you want?

Because sometimes the problems in given industry are very well defined and understood. And even the cliché answer "a faster horse" isn't really useless, you just have to go little further asking...

4. Why?

... a couple of times. Next an obvious candidates will be a "profiling questions":

5. Where are time and money (and resources, attention, pain, fear...) in X?

Now, form a different angle:

6. What in X is worth spending money on? And what is not (because of utility/price ratio)?

I have a problem with question that will filter out problems that won't make a scalable startup, maybe because scalability is property of a solution not of a problem?

It's far from complete, but above should already give some hints about what can be done.

Learn more "startup patterns" to spot more potential chances for business.