[encoderer]

I have to say, I don't agree with this part at all:

"But most of the time, customers don’t really want the the features they are asking for. At least not very badly."

Customer feedback drives an absurd amount of our roadmap at Cronitor. We have a good idea of the many shortcomings of our product and are constrained primarily by resources in developing it faster. When a customer -- especially somebody on a trial -- puts their thumb on the scale of a specific flaw or deficiency, we look at it as an opportunity to seriously delight that user and at the same time level-up the product for all users after. We don't build everything asked for, but I would say "most of the time, customers know exactly what they need, and we try to give it to them within our ability."

A specific example for us would be Etsy, who uses Cronitor on a part of their business and during evaluation asked for a couple API endpoints to expose more advanced functionality.

[joeemison]

For evolutionary features, you're right. But I think the point in the article is that if you're trying to get to product-market fit, your customers' feature suggestions are rarely the ways to get there; you generally have to watch, think, synthesize, hypothesize, and iterate again and again to get to the solution that actually does what you need done.

[encoderer]

That's a little too hand-wavy for me. I think you solve a problem, produce an acceptably bad solution, and as you get traction you improve it. You do it evolutionary because you're not Steve Jobs and improving something continuously is a repeatable, adaptable technique in a way that spontaneous invention isn't. If you never get traction, and can't crack it, then you invested a minimal amount and you can move on. A portfolio approach.

I think this is true for both big ideas and little ones. I guess there may be some nuggets of truth deep down in the cw that "users don't know what they want until you give it to them" but I think it's mostly an over-used canard.

[ben_jones]

I think product-market fit is more then just making sales. It's having a sustainable company that can "service" those sales in a sane way. If your company is completely back logged and getting by by cutting corner after corner, you're not succeeding.. you're teetering on an edge.

There's an argument for the mental health of the founders as well as the soundness of the company.

[mpbm]

I think that's repeatable and adaptable for two reasons:

1) technology keeps changing and it's usually easier to just write new software to take advantage of new infrastructure or design assumptions or whatever

2) almost nobody who has the expertise to do so gets paid to research pre-existing software, test it out, and find something that already exists that does the job. However, there is a large financial incentive to sell people on why they should use your new thing

So, there is never-ending opportunity to sell something new that solves the same old problem. It's a marketing-driven approach. You're basically just exploring the frontier of what people are willing to pay for, rather than the frontier of what's technically possible. It's "there's a sucker born every minute" style repeatable.

Not that there's anything wrong with that. People would have had to pay for research/adaptation to find an existing solution, so paying for development of a brand new solution is the same money. And marketing is one of those skills that's necessary no matter what your approach, so laying a foundation on top of it makes sense.

They're just two different approaches. The big idea approach isn't hand-wavy, it's just less common than the marketing driven approach. I think maybe it seems hand-wavy when people try to wear the big idea's skin like a mask cuz they don't actually have a big idea and everybody's like "grandma, what big teeth you have".

[rdtsc]

> But most of the time, customers don’t really want the the features they are asking for. At least not very badly."

It is easy to agree or disagree with that because it is rather ambiguous and everyone has seen it go either way.

At least, I've seen it go both ways. It really depends on the context and most of all who the customer is. Are they paying for the product, will they be more likely to upgrade or pay if the feature is there, do they know that feature contradicts with other feature they already requested and so on?

Even silly things like customer is trying to be nice to you can interfere with the process. Because they don't want upset you and they want to be polite, they might ask for something trivial and hide something more annoying or bothersome.

I have seen this happen with me and design. I know a crappy design when I see it. It is annoying and bothers me. But, at the same time I wouldn't know how to tell the designer to fix it. If they press me I might make up something like "yeah, change color to beige" or something silly like that. I can imagine that happens with any domain where customer might not have as much expert knowledge.

[vonklaus]

the horribly over used Ford quote:

> if i'dve asked people what they wanted, they'd have said faster horses.

is applicable here, much more so than in the battlecries of every company making something no one wants for some hypothetical market that doesn't exist.

While I don't understand the author's biz/space enough to comment situationally, this is a key insight. People almost always miss this. Users want to get to places much faster, and much more confortably in the Ford example. They speak in solutions and it is an entrepreneurs job to translate this into problems, rank them, and provide actionable responses.

The flaw I believe in your example, could be a few things. possibly if you are paid by a customer directly & are more a consultant/contractor than a SaaS provider, well, that's how you get paid. However, generalizing about a feature, especially without behavioural feedback is dangerous. if it is very easy to do, of course do it. However, the risk of a feature is some subset of:

* value to customer

* conviction/data in that value prop. & feedback

* difficulty to achieve

* is it replicable. how valuable is this to all my customers? is it very valuable to a few, or semi-valuable to many/all.

This is again, highest level, even in the last point above you can see the thread fan out.

So, again, i am bot super familiar with the authors company but I agree wholeheartedly with his process of decision making. Consider it is possible that either etsy was a big client, thus worth retaining for markwt goodwill & rev, or that it was quite easy to open an endpoint to data you already had, and thus doesn't discount this methodology

[stephengillie]

The somewhat underused 3M quote (paraphrased):

> When someone is shopping for a drill bit, they don't want you to sell them a drill bit. They want you to sell them a hole in their wall.

If you're wondering, most people will put a nail or hook in the hole in their wall, and hang a picture or something. 3M would go on to sell them tape or other wall adhesives instead of a hole in their wall.

Rephrased to correlate to this thread:

> When someone is shopping for document signing software, they don't want you to sell them software. They want you to sell them a set of agreements already negotiated with their business partners.

Don't confuse the tool with the goal.

[vonklaus]

I agree with this and I think it compliments Ford, or at least my interpretation.

I've put a fair share of holes in walls and also fastened plenty of material together. No one would buy a drill for one hole in a wall but there is a massive amount of leverage in being able to put thousands in quickly, ect.

> don't confuse the tool with the goal.

I think this is the lens/acid test. For your correlation above, it sounds like it was harder to sell them automated agreements than it was for them to manually produce them. Or at least, the perceived opportunity cost was higher and the cost to educate them otherwise was higher than the authors profit marg.

edit: just looked at your profile; very nice array of quotes you have.

[rhizome]

The weird thing is, isn't it obvious that the first thing to do is to replicate all of the functionality of the customers' current package? It's nice to get a laundry list of possible future features, but when testing indicates that people are only using it for the minority of their contracts which fit the software, that's a signal that besides the laundry list, the software is still incomplete. Did they really shut down due to an inability to prioritize enhancements? Kinda sounds that way, but runway is runway.

[Spooky23]

I was struck by that too. Most people actually know how to do their jobs, and what they need to run them.

I worked with some guys a few years ago that do vertical market software like this. They have like $12M in sales and 6 employees. Most sales is via referral, trade shows and partnerships with other vendors.

Not sexy stuff, but lucrative.

[gmarx]

If paying customers are asking for features pay attention. If non customers suggest features, you should be extremely skeptical that implementing those features will result in sales

[jcoffland]

Maybe you're wrong. It's worth considering.