How we were building the wrong product | Hacker News Reader

How we were building the wrong product

kilometer.io

58 pointstzipis27 comments

The fact that they feel the need to have mixpanel, segment.io and hotjar on their webpage doesn't exactly show confidence in the product :p

Analytics is such a saturated market at this point. It will be interesting to see how things shake out in terms of consolidation / niches / price ranges.

It is also interesting to me that they decided to invent yet another format in which customer data must be exposed. Just like the previously mentioned companies. https://www.w3.org/2013/12/ceddl-201312.pdf (shameless plug: https://github.com/mkohlmyr/ddl-json-schema)

I saw it mentioned while doing searches a while back and ended up on Quora where a Kilometer rep says in his post:

> our users often call us "Mixpanel on steroids" [1]

I then click through to find ironically, that Kilometer themselves use Mixpanel..

[1]: Screenshot: http://thumbsnap.com/i/kFfmeZEx.png?0310 (from https://www.quora.com/What-are-the-best-Mixpanel-alternative...)

Will their next post be "How we were building something that already existed for the last 10 months"? Because this a lot like https://www.geckoboard.com/

And as far as I can tell, there are around 20 companies offering exactly the same thing in this space - most of which have been around for a couple of years already.

Not sure it's a good idea to pivot into this space.

Ideas don't have to be original to succeed. Not at all. Having an original idea doesn't even give you a leg up. It's all about execution and having an idea that is at least desirable.

That's what Alex is trying to do. He's trying to find that mix.

Overall I think analytics aren't as good a market at they sound.

The basic problem is that they are something you bolt onto your business after the fact rather than something that is core to running a business. They are a "nice to have" rather than "have to have" and when something new comes along it is too easy to replace them.

Quite often it depends if the analytics are actionable e.g. they give a good understanding of what exactly needs to be done. If that's not the case, many analytics are just a nice to have to look at.

At Weekdone (https://weekdone.com/) we have some dashboards for your team and employees. What works is if something that you see trending up or down is actionable towards specific people. If it's average across company or department - not so valuable.

A good example is employee satisfaction. It's most valuable looking at trends for a single person or a small team. Whole company or large department? Not so much, nice to have.

It's the difference between a vitamin and a pain-killer.

You want to make pain-killers, because people need them.

Vitamins can be helpful. But people don't usually need them.

I would have put money on this example being structured the other way around. People can live with pain; they can't live without vitamins. They only don't need vitamins in the same sense that people don't need oxygen.

Pretty sure that the post you're responding to meant "vitamin supplements."

That's it, thanks.

Not being familiar with this product to begin with, I'm wondering if he gave up on his product too quickly!? In my experience customers are typically wrong and misinformed - the whole "faster horse" concept at play. Instead of acquiescing to his customers whims (I want what I want when I want it), could he have had a viable, differentiated product in the market if he would have worked on selling, onboarding and educating about his product's POV. More often SaaS products are selling an idea combined with a methodology (Slack is a good example of this). You need to sell the idea that your organization will benefit from doing it "our way" rather than "your way".

"5 basic events through the Kilometer.io API (website visit, user signup, app visit, user cancel and user billed) so that we can automatically calculate 25 charts and metrics we believe every SaaS company should have."

25 metrics from 5 data sources seems... overkill?

SaaS for Saas companies seems to be a hard market, because the SaaS company is likely to just build what they need themself, no?

From a similar experience I have found this to be true. Specifically:

  - one size fits all is very difficult to do.  most companies have a critical feature that is unique to them

  - the CEO is likely to view this kind of product as 'trivial', without really considering how much engineering effort it will take / they really need that custom feature.  so yes they build it themselves.

  - very small companies are not comfortable providing 3rd party access to their db

  - larger companies want more than this

  - the selling cost to companies tends to be high enough that going the low cost / high volume route is difficult, quickly driving you into consulting level prices / a more fully featured & tailored product

Another problem is that for startups, these metrics change all the time. Often you start measuring something to diagnose a problem.

Once you have done that, you don't need to diagnose it any more.

Also, I'm not sure how a few pretty charts help me here.

The pitch in their email is "plug & play SaaS analytics tool".

This pitch tells me nearly nothing. It doesn't say what I would use it for, what benefits it offers, how it stands compared to other solutions on the market, or even any details about what platform(s) it supports. So basically unless I happen to currently have no solution and am actively looking for one, then I don't care. Which I would guess is about 99.999% of the people who received their email.

The way I understood it is that the people receiving such an email were already on the beta-test invite mailing list. So they knew about the product already.

Yep, users don’t see the value in the product (as shown in the email from the user). “plug&play analytics” does nothing to explain what’s new and why it’s worth my time.

Based on the info from the post, there was no specific need and ideal customer profile in mind for creating this product.

It’s not like an unknown problem, but it’s crazy how often startups have fell in the “do stuff no one needs” and wasted months or years on it. It seems to be one of these things where you need to experience it yourself to understand it.

Great, open reflection on the product. Thanks for sharing!

Off topic.. Your logo arrow (and favicon) looks a bit like a body part to me.

> It took us almost 10 months to put our product in the hands of real users.

Almost wondering if I should read the rest of the article at all...

How did they expect to build the right product without collaborating with their target customers? Is this a company staffed by psychics?

TLDR; Prototype with customers before you build anything.

You make it sound easy and obvious, but the problem is that customers doesn't actually know what they need. They know what they think they want, but they usually want something very much tailored to their business, using their ideas, and their workflows, but at a price of $10/month because it's a SaaS app. Building something that works for many customers means you actually have to come up with a very astute compromise between what the customer thinks they need and what they actually need, and sell the product to them by persuading them they don't actually need the things they just want.

On top of that there's an added problem that customers are often wrong. They tell you what you want to hear, or worse, they tell you what they want to hear. And you have to work out what's really needed based on those conversations.

My startup failed for this exact reason. We listened to our potential customers, we talked to them in focused test groups, we built what they said they needed, we showed them prototypes, but at the end of the day they realised what they really wanted was something else, and in many cases the pain we were solving (managing project requirements) wasn't even that important to them. By then we'd run out of funds. The end.

Customers know their pain. Don't expect them to innovate and create the solution for you, but get them to describe the problem. You can then go back in other rounds offering potential solutions through prototypes. Try to keep moving towards getting them to commit (i.e. pay). Ask what it would take to make them buy, what they'd pay, etc.

It may not be easy to do, and I agree that people don't know what they want, but it took these guys 10 months to do any research. They shouldn't have built a thing, but should have put prototypes in the hands of customers ages ago. Then they might have learnt the things 10 months ago they're only just learning now.

There's the old adage from Ford "If I'd have asked customers what they wanted, they'd have said a faster horse". But that's asking them for a solution. It's your job to come up with that. But the customer is the one who knows the pain. And if they can't describe it, it's not big enough to solve anyway.

Nail it then scale it (www.amazon.co.uk/Nail-then-Scale-Entrepreneurs-Breakthrough-ebook/dp/B0055D7O1U/) describes a practical product development process that puts research at the forefront.

We took similar actions at my recently failed startup, which headed on a slow, downward path. I identified an apparent pain in our target market based on past conversations with friends working in that space. My team and I felt we validated the pain because we kept hearing about it in our face-to-face interviews with potential users & buyers. One interviewee even went into mild theatrics half way through his interview, "Stop! This keeps us up at night because we ..." That was just the beginning of his 10 minute proclamation that the problem was worth solving, and that they'd pay a decent amount for the solution. Exactly how he said it.

We developed what we thought and what our potential customers claimed was a solution to an unmitigated problem in a very sensitive financial process. We followed startup advice and made it clear that we would charge for the solution. Fast forward a few months: usage dropped from supercharged to minimal then to nothing. Most of the customers we worked so hard to acquire went back to their old ways. When we did followup calls, their newer recruits were none the wiser that we even built something specifically for them. Clearly, the problem wasn't that important in the grand scheme of things.

Please don't be snarkily dismissive on HN.

If you know so much that others' mistakes are trivial to you, why not share something informative and non-obvious? Don't just put others down; teach us something.

The pop-up without a close button was a bit of a deal breaker.

The close button is there in the upper right hand corner, just not visible when reading in landscape orientation. Rotating to portrait allowed me to close it. Very annoying.

Reply on news.ycombinator.com