The life and death of a startup | Hacker News Reader

The life and death of a startup

cecinestpasun.com

267 pointsj_s63 comments

I think a lot of this comes down to the sector they were going after.

Speaking from experience, software for Plumbers, Roofers, Pest Control, Cleaners and most home improvement type sectors are really hard to scale. It really is an up-hill battle getting them to spend more than a couple hundred a month and the churn was so much higher than any other sector I have worked in.

You have HomeAdvisor, Angie's List, etc. who are heavily tapped into this market and understand the challenges that come along with it. We developed a lead gen, crm and follow up system targeting this sector. So many of the Plumbers and Roofers we talked to would complain about HomeAdvisor and other similar providers. How bad the leads were, that they were shared (sent to 5+ others), etc. but were not willing to pay for a higher priced offering that fixed those things. Most that were willing to invest were horrible at even answering calls from prospects. They would call prospects back 2+ days later to discover they went with someone else who actually did answer their phone and would then say it was a bad lead.

The experience gave me a much better understanding on why the competitors focus on shared leads. Hopefully at least 1 of the 5 roofers the lead is sent to will actually respond in a timely manner.

I have a feeling that there is a reason, these small businesses stay as small businesses; They are not the most economically rationale decision makers

As someone who comes from parents who own a small construction company, I think you've actually gotten it the wrong way round. What you say could be construed as correct, but in the wrong way. From what I understand, many small businesses (especially those who are Mum and Dad run) are only looking for small business, they have no interest in going bigger, why would they? In the majority of cases, the work they do more than covers the bills, people may think that bigger is always better, but when the business supports a family, this is rarely the case, bigger can actually be far worse. Not to mention that much like was mentioned/alluded to in this article, founder rates of pay/benefits are pretty much never the same as would have to be paid to a "normal" employee. This is unscalable, I know Paul Graham has referred to this idea before, so I won't explain it worse.

TL;DR

Most small businesses owners are far better rational decision makers than you think, because most of the time, the aim of the game is not a unicorn, it's support of the family.

Coming from a similar family background I think thats why I said: 'economically rationale decision makers'.

Overall financial gain may not be the only thing they optimize for. For example their life style (being able to answer call whenever they want).

Ahh sorry if I came off to be attacking you man, wasn't my intention at all.

And yeah you got it spot on! I know that for my Dad in particular, he always said to me whenever I mentioned scaling the business up and taking on more guys that whilst it would be nice, it would likely mean more of a managerial position, keeping an eye on projects and that. He said he runs his own business to avoid being at the grindstone 24/7, he was always able to come and see school sports and participate in that sort of stuff. I always found it pretty self-controlled tbh, knowing that he could have scaled it up, but knowing the game would change if he did.

There's one kind of growth where you are investing dollars in marketing, sales, hiring and scaling the YoY percentage points; another, where you're investing in efficiency, automation, better gigs, and reduced work load. Why wouldn't anyone be interested in the latter? Small businesses aren't aware what technology can do for them and often, there's a reluctance to change.

I am not sure if businesses have become tired of the constant bombardment of promises that don't have a solid ground but I believe some level of effort can make a positive change in their lives.

Maybe good enough is just good enough for some business owners?

If you are making good profit already, which more than covers your needs, why bother looking into improved efficiency, automation, getting better gigs, etc? It's not that you can just flip 'efficiency' switch and things are more efficient.

In some cases, sticking to doing things in a familiar way and which works well enough is a rather reasonable decision in the grand scheme of things.

It's true, some people have to dragged into the future. When shopping carts first came about, stores had to hire people to go around the store demonstrating the benefit of using a shopping cart.

At the same time, if you've not read the fisherman and the businessman (1 minute read), the ending is relevant to your discussion.

http://paulocoelhoblog.com/2015/09/04/the-fisherman-and-the-...

It may not be a financially maximizing strategy, but it's often rooted in personal rationale. A friend who runs a very successful real estate business once related to me that "[he has] no desire to be another DeLeon"

All the plumbers and construction contractors I know as friends are constantly busy. They wouldn't know how to handle more incoming business. They can't really charge more for their work. So their only way to scale is to change the structure of the business (go regional rather than local for example). There are significant risks and costs in doing so, and software doesn't help much.

> They would call prospects back 2+ days later to discover they went with someone else who actually did answer their phone and would then say it was a bad lead.

That's the opportunity, but it might not be what you want. Start a very tech-savy, customer-friendly plumbing business and figure out how to franchise it. Have your own small in-house tech department that relies on outsourcing. Outcompete the local plumbers and then hire them when they go out of business. All of your franchises will be required to use your software system.

I encountered something similar when I hired some painters via a national chain. The guy who ran the local franchise saw it as an investment opportunity, and loved running his own business. One of the painters was a guy who's painting business folded and realized it was just easier being someone else's employee.

>but in our case, the purchase decision was rarely made by the person who actually had to use the software. And you have to get those people on board. If anything, they're more important, because they're the ones who are going to make the bosses life hell if the software they buy isn't doing the job - or worse - is doing the job too well.

>Over and over again, we saw internal sabotage. People would simply refuse to change processes, and would find any excuse. "Oh, the software didn't work, so I had to go back to doing it manually".

This one resonated with me, from a former workplace. Even when their jobs were nowhere near at risk, people don't want to change habits. The employees did not get paid more for being more productive, so they would find fairly pointless excuses for not using tools that made their lives easier.

Ultimately, it's a problem with management and their incentives (or lack thereof).

On the flip side, as someone who's tried to change stacks or IDEs or other software a few times, there's something to be said for already having momentum using existing solutions. Sometimes alternatives can be a little better, but not better enough to justify the time/energy/etc costs of switching. And more often than not, alternative solutions have their own flaws too. I tried three times in the past week to switch my portfolio website (https://sdegutis.com) to using Jekyll or Metalsmith or Hexo, and each time it just didn't prove any better than my own hand-rolled Node.js site generator, and in most cases, it was definitely worse in some ways, at least for the specific requirements I had. So switching tools can be good, if it's legitimately better in some way, and not worse in any other ways. But that doesn't isn't the case as often as we'd like to hope.

Currently running into this moving from Vim to Spacemacs. Even with vim bindings, it works just slightly differently enough that there's a good chunk of friction. Specifically, mashing escape no longer reliably gets you into normal mode, bash is in a separate window instead of in tmux, some of my settings are different in ways I don't want to spend the time analyzing (probably scrolling? IIRC I set things up to center themselves as best they can in the window in vim), and the tab-completion for opening a file with :e is different through trying to be way more "helpful".

Assuming Spacemacs is like regular emacs, Ctrl-g [(keyboard-quit)] is the moral equivalent to esc in vim. Mash away.

If you want esc to work that way, put in your .emacs (but I really discourage it):

  (define-key key-translation-map (kbd "ESC") (kbd "C-g"))

we were selling software

No. You were selling the solution to a painful problem that just happened to implemented in software. That's not the same as selling software.

We've been conditioned to expect that physical, tangible things are expensive - but software? That should be cheap, or better still free.

I think this is where the Django influence is strongest. People who buy enterprise software don't think that it should be free. Quite often they have strange expectations that they need to pay prices far higher than anyone would reasonably charge. $5000/month for an app that could be replaced by a cron job and a shell script isn't unheard of. I strongly suspect Tradescloud could have greatly benefitted from a mentor who had experience and success in selling B2B software.

I wish startup and industry publications spend more time with post mortem style stories of businesses that didn't work, rather than making it seem that all you need to do to print money is to start some sort of software as a service company.

Sure, it won't sell as many copies or ads for the publication, and some may say it will discourage the entrepreneurial spirit, but I think it is far worse to see a legion of young, bright, keen people quit jobs and start writing code dreaming about what colour their Ferrari is going to be.

The inevitable disappointment (for 99.9% of them) from unrealistic expectations, and the sheer, gruelling reality of just how hard it is to make a dollar in a self owned business will surely do more to crush spirits and shatter dreams.

You might be interested in http://autopsy.io/

Should watch this complementary presentation too.

(Russell's issues with depression/burn-out.)

https://www.youtube.com/watch?v=OC3v5uXR9Qc

Thank you for posting this. I'm watching this before going to bed and just started crying when he said "You're not alone". I need to get some help.

Does this have anything to do with Django whatsoever? The title is misleading.

The product was built using Django and the author is a big Django contributor.

Sure, but the title is link baitey. It suggests Django had an impact on the failure of the company when in actuality it has absolutely nothing to do with the outcome.

I dont think it had to do anything with the failure of the company either. Although Ill have to ask the author for more info privately. And yes, the title is a bit click bait, but I think it relates more to the person (a django contributor) than to the tech.

Django have an "engineering is important" philosophy so developers attracted to Django have a similar mindset. In this case they're showing how despite superior engineering and a proven market is insufficient on its own - the start up required more effort in marketing, sales, financing.

This is building a business 101. I still feel it's disingenuous to begin this long winded post with the connection to Django.

As an aside, I'd argue that any modern web framework with decent community interest also attracts the same "engineering first" individuals.

If you're betting the farm (read: assuming your progressive choice is going to do anything more than speed up dev time) on your choice of web framework you're definitely doing it wrong.

The software you create is a means to an end.

Yes. I think if we think of ourselves as Django developers while building a business it's possible we will fall into the same pattern as exhibited in the article. That's how I see it's helpful to me as a developer who works with Django - it's a warning not to identify myself with the tools I have.

interesting. I've always thought of 'engineering' focus in software as being close to the metal, using the hardware efficiently. But in this case they seem to think of it as minimizing time to market, and amount of code. Even with that, python based solutions don't really fit my idea of good engineering, because there is no reason you can't have a language and ecosystem that nice, plus good performance. But just because a thing could exist doesn't mean it will I guess.

It was a business process improvement startup, not a Django startup.

And he implied their product was on mobile devices so may not have involved Django at all.

This talk was given at DjangoCon

The fact that the author thought of his company as a "Django startup" is probably one of the major reasons for its failure. Two-sided marketplaces almost never live or die by their technology. eBay's technology and user experience is complete dogshit but it is still the only place you ever go to buy or sell a vintage Dukes of Hazzard lunchbox.

I've heard two-sided marketplaces are one of the toughest business models to get started along with advertising based businesses.

The irony of this comment being that the comment right below you at the moment is asking "what's this got to do with Django".

It's a very good article, you should have taken the time to read it.

This is a bit of a strange article. I see where they're coming from, but a number of things that come to mind as I read through this:

- Competing against Yell, Yelp, Google...etc. for local reviews and listings isn't exactly going to be easy. I'm not sure how prevalent all those local services were in 2010, but even if it had gained traction, it likely wouldn't have lasted long (or perhaps if they're lucky, been bought out).

- Did they only try cold calling? If so, that's ridiculous. Did they attempt multiple marketing methods? Did they try to educate the market before trying to get a meeting/sale? Did they run group sales events? They really didn't seem to know what they were doing at all, so how is this enlightening to people reading? Maybe learn sales and marketing before trying to actually build a business...

- Local businesses often aren't very good at running a business. It's that simple. Some survive through luck (and referrals). They're often disorganized, and so busy dealing with clients they can't even think about sales and marketing. And often even if you can outline a very clear ROI, they won't go for it since small businesses are often more of the consumer mindset (emotion driven) rather than bigger business mindset (ROI driven).

- And also, as another comment mentions, a lot of small businesses don't want to really grow. If they're earning enough to cover their bills, they're happy. The stress of managing employees is often more of a deterrent than the benefit of earning more.

- A $50k solution to a local business? That's a ridiculous price and expectation. At least have multiple service levels, and a free version is a great starting point too.

- At least they understand that there's often massive resistance to changing how they do things, especially if some members of staff aren't comfortable with technology (as many people aren't).

I'm just slightly shocked that they had to learn such fundamental truths about business, in such a hard way.

Did they consider hiring a sales/marketing consultant?

Did they test other selling approaches?

With great sales and marketing, you can sell a poor quality product. Not recommended of course, so if your product's good you have happy customers and also referrals.

They had the product-part right, but didn't seem to have a clear product-market fit, and certainly hadn't found a dependable sales/marketing approach that allowed them to grow.

I feel like another step missing or at least not fully acknowledged is knowing your competition. I just did a quick check and both Angie's List (1995) and Yelp (2004) were around years before this idea (2010). While it's certainly possible to be late to the game and win there's no visible differentiator here.

I'm not sure if you read the blog post, but they pivoted away from that idea early on.

What is the best answer to "We won't go with you because you are too small?"

Find the subset of the market who will go with you, and work your way up the market (down the riskiness scale) as you build reputation and market awareness.

It's tricky because simply being a startup is a brand negative as you try to sell to increasingly conservative customers.

There's books about it. Here's one:

https://www.amazon.co.uk/Crossing-Chasm-Marketing-Technology...

I honestly don't understand how you can develop a product without speaking to potential customers or users (in this case the plumbers). Reading this article, this makes me wonder if you really took this serious.

well they did after they developed a prototype. If you think you can build a prototype fast enough, then I can understand the decision to make the prototype first to have a clearer idea.

This is a great article. One thing I can particularly relate to is when working on a 'blue sky' project, how easy it is to rationalise away the obvious red flags and keep going until everything falls apart. When one has invested a lot of time and built up the momentum, it's very hard to admit failure and accept the big changes that must come with letting go of a project.

I think it's particularly hard for people who are used to, and enjoy being successful - is that perhaps why university dropouts often make good company founders?

> Our burn rate was closer to $22k per month.

Seems like this should have been mistake #1. That is an awfully high burn rate for a startup still trying to find product market fit.

I interpreted that as them including their salaries, and not actual on-going server, marketing, advertising costs, etc.

And that would be an incorrect way to compute burn rate. Should include all outgoing cash regardless of the purpose.

I was surprised as well. But he is talking about Australian dollars, not American ones. which makes it around $17k US. That makes $6k/month a head, plus office cost (bank fees, ads, furniture etc).

Thanks for the clarification. It's still on the high side, imo, but depending on where you live, it may be sort of a low-end floor. I would think that many people would just forgo taking money out for the first several months, but perhaps there were more than just the 2 original folks? I couldn't tell from the article.

While this is interesting for many reasons and quite frankly because of a similar situation in my life. The choice of framework had little to do with anything.

Django was on the page 11 times, 5 of those were regarding DjangoCon

My personal lesson and the take away from the article is: On the list of hats you wear founding a startup, salesperson is 1st, 2nd and 3rd, engineer is 10th.

I don't understand how this is considered an autopsy, what was he selling? what was his conversion rate? what was customer acquisition? what was retention?

Paid high salaries and burned through investment without closing sales? That's probably a little more indicative of why this one failed.

If I understand it right, you effectively acted as cheap consultants for your big accounts? Did you have a stable product that you could sell without any modifications to it?

"slow train wreck"->a boat wreck

This is a good article, well worth the read.

They missed the opportunity to title this article "13 reasons why"

what is django got to do with this?

It's a fine startup postmortem.

A small gripe: I don't get why Django has to get dragged into this. Whether it's the title, or the venue. It distracts from the broader lessons he brings to the table in hindsight as a CTO-founder.

Other than the author making fleeting references to Django, there's no technical depth or major points delivered on how the stack effected operations.

I have nothing wrong with Django criticism, but the title needs to be fixed. Maybe: "Autopsy of a slow train wreck (DjangoCon 2017)"

He was the president of Django Software Foundation while running the startup.

So naturally he was a big proponent of Django and his stack was written completely in Django.

Right. But what outcome did Django have on his startup? Seems to me they likely would have committed all the same business mistakes had they used, say, Rails or Play.

My take on this, drawing from his "Humans gonna human" is: "Engineers gonna engineer" - a mistake that's cost me dearly in the past.

The article is adapted from a speech given at Djangocon. The title includes "Django" to explain why Django developers would be interested in it. Out of context here on HN, any startup failure draws interest (no matter what technology), so the Django mention looks odd.

> A small gripe: I don't get why Django has to get dragged into this.

The fact that none of the reasons why they failed have anything to do with tech is what makes it being a Django startup relevant. The whole point is that they were focused on their technology instead of the customer acquisition channels, sales funnel, pricing strategy, etc.

Yes it's clickbaity, but it's also art.

As a Django developer working for a startup, I clicked on this expecting to read some lessons on how Django caused them to fail but didn't see anything _specific_ to Django, but more of the same general advice that I'd seen before.

It's worth nothing though that this is only the title of the presentation aimed at Django developers at DjangoCon. The title of the blog article removed the reference to Django.

The article is unclear, uses subheadings that have no connection to the content and is far too long. I am not surprised that their startup crashed and burned, if they cannot even write a clear article.

From looking at the description of their app on Appstore: "With this mobile app, field workers can receive jobs sent to them from the TradesCloud web service and then update them in the field. " their service seems to have been a way to message the staff in the field.

Reply on news.ycombinator.com