Thomas Edison famously once said “I have not failed. I’ve just found 10,000 ways that won’t work”, and I can relate. With the exception of my freelance developer days, BugMuncher is my first venture that I would consider successful (even though it’s still not quite profitable). The truth is, I’ve been trying (and failing) to get some kind of business/side project/get-rich-eventually scheme off the ground for over 12 years.
I was thinking about this recently, and decided it would be fun* to look at some of my previous projects and half-baked ideas, specifically why they failed:
* fun for me at least, and hopefully not mind-numbingly tedious for you
Plug and play LCD readouts for computers
I almost forgot this venture until I was nearly finished writing this post. In 2004 I was just 18 years old and taking a gap year between college and uni. This wasn’t so I could travel or do anything cool like that, it was simply because I didn’t know what I wanted to do at uni. I actually spent the year doing telemarketing for a pest control company (you can’t make this stuff up).
At the time I was quite into the modified computer scene, and one popular mod was to put an LCD readout on the case that could display things like internal temperatures, memory usage, custom messages, etc. To do this you could either buy an expensive USB unit, or buy a cheap bare LCD screen and wire it up to the computer’s parallel port. (Hey, remember when computers had parallel ports? No, me neither)
With my A-level in electronics fresh in my mind, I decided I would buy the bare LCD displays, and manufacture circuit boards that would allow them them simply plug into the parallel port without needing any extra wiring. No one else was making plug and play parallel displays, there seemed to be a gap in the market.
I used to post a lot on a forum for computer modifying, so I created a post on there to gauge interest. The post took off, a lot of people were interested and wanting to order from me, I thought I was gonna be rich.
This was straight forward (or so I thought), buy the parts, put them together, add a profit margin, sell.
The main problem I had was sourcing the base LCD units at a decent price, with out having to order thousands. I went through quite a few suppliers but they all seemed to be too expensive, or fell through before I could sort out a deal. Added to that, I found etching the circuit boards had quite a high failure rate, and there was still the time required to put the parts together. In the end, it just wasn’t viable.
Startups are hard, hardware startups are harder, bootstrapped hardware startups are hardest. If I were to try this again knowing what I know now, I’d make sure I had enough capital to order a big batch of LCD units, as well as have the circuit boards professionally printed rather than trying to etch them myself in the garden.
Mobile Recording Studio
Now 20 years old, I was in University, and dreamed of being a rock star. I’d been recording my own music for a few years, and thought I did a pretty good job of it (I didn’t). I’d also recently passed my driving test, and now had a car. A plan formed.
I blew a large chunk of my student loan on some new, more portable, recording equipment, with the plan of being a low cost alternative to a recording studio for local bands looking to cut a demo.
As is the story of my life, there was no marketing plan. I made a basic logo in photoshop, and set up a MySpace page. (Hey, remember MySpace? No, me neither)
I had a total of two customers, neither of whom paid for my services: A band in which I was the bass player, and my younger brother’s band. The idea being I’d record these for free so I’d have a half-decent portfolio of work to show potential paying customers. Unfortunately the paying customers never came. But then, I never went looking for them.
I’d have to say lack of marketing efforts, but had I done any marketing, I’m sure the shitty quality of my recordings would have put off any potential customers. I don’t really remember deciding to shut Nomad Recordings down. I think that the lack of customers, plus the growing work load of my third year of Uni, caused me to lose interest.
If you build it, they probably wont come. It’s a lesson I’ve learned many times, and I’m still learning with BugMuncher. If you don’t actively try to get customers, they’re unlikely to come find you.
Easily edit Static Websites
Fresh out of University, and with no interest in using my computer games development degree to actually develop computer games, I started freelancing as a designer and developer of websites. One of my first clients was a local gym, who initially wanted some changes made to their website.
Their website was 100% static, in the old-school way, ie: No databases, no build scripts, just hand-coded HTML. Back in ‘07, this kind of set up was fairly common. (Hey remember when websites were hand coded HTML? What you mean this is getting old?)
This gave me an idea - what if I could build something that could drop into a site like this, and allow you to edit it? xStatic was born. With youth comes naivety, so I didn’t do any market research, or even think about how I’d make money from this, I just dove straight into code.
My initial demo worked pretty well, but I never got as far as adding security and launching it. I’m confused by my own actions here, as I can’t for the life of me remember why I stopped working on xStatic.
Back in 2007 this could have been a pretty awesome tool - a database-free CMS for existing websites. In 2012 EllisLab (the guys behind Expression Engine) launched MojoMotor, which did roughly the same thing. I was 5 years ahead of them!
I never thought as far ahead as monetisation, but I suspect it would have been a one-off cost to buy the software, as there was no need for it to be centrally hosted or ‘phone home’, so a subscription would have been disingenuous.
It just kind of, stopped. Maybe there was some bug I just couldn’t fix, maybe I just lost interest, I don’t remember. I don’t even have the source code any more, but one way or another, xStatic never saw the light of day.
Don’t give up so easily on something that could have actually been a pretty cool product.
Journey Cost Calculator
Ah TripCost, this one was about as close to success that I ever came. Now 23 years old, I’d already been freelancing for a year, and I loved the idea of having side projects to bring in some extra cash. I don’t remember exactly how I came up with the idea, but I know I’d been experimenting with the Google Maps API.
Anyway, TripCost was basically a mash up between Google Maps and the UK Government’s car fuel efficiency data. It was simple, you choose your car from a list, enter where you’re going from and to, and TripCost would tell you roughly how much it would cost in fuel. I used it quite a lot, as did my friends and family, and it turned out to be surprisingly accurate.
Initially I relied on my standard marketing plan of no marketing plan, but then I had the idea to email a TV show on BBC two called Working Lunch, as during December they were doing a feature where they’d open a door on an advent calender to reveal a Christmas treat or Christmas turkey. I managed to get TripCost to be a featured as a Christmas treat, sending a huge traffic spike my way.
I was just using adsense to try to monetise the site. Traffic was steady, but in the low thousands each month, so my total earnings from TripCost was about £100.
I stopped actively working on TripCost after about a year, but I kept it online for three years or so, until I shut down the server on which it was hosted. My plan was to rebuild it, add some new features, and re-launch on a new server, but that never happened. To make things worse I accidentally let the domain expire. A domain squatter got it and held it to ransom, at which point I abandoned the whole idea, especially as I’d recently launched BugMuncher.
This year I discovered that the domain squatter themselves had given up on selling the domain, and let it expire, so I bought it back for £10, and now plan on rebuilding and relaunching this one. It’s not (quite) dead yet.
Unless you have a lot of traffic, ads are never going to make you much money, especially these days. I should have tried some other monetisation strategies, and put more effort in to generating traffic.
Shared Hosting Reseller
As a freelance developer, I also offered my clients hosting, which I provided via a reseller account on Heart Internet. I decided to try selling hosting to the general public as well. At this point I hadn’t heard of the Minimum Viable Product (MVP) approach, so I spent months building a custom hosting control panel, integrating with Heart’s reseller API, I even built a support ticket system (which no one ever used).
Once it was finally finished, I adopted my tried and tested marketing strategy (none) and launched. I did have a USP though, instead of just offering predefined hosting packages, I had an option to build a completely custom hosting packing, with fancy jQuery UI sliders for things like disk space, bandwidth and databases. It was later I learned about the paradox of choice, and why companies tend to offer predefined packages.
Despite my marketing plan I did get some customers, most of whom found me through the same forum on which I posted Pre-wired Parallel LCDs. I don’t know how much I made in total, but including freelance clients for whom I was also providing hosting, I’d guess about £1500. Because Heart Internet charged about £40 / month for an unlimited reseller account, the profit margins where at least high.
A few factors caused me to shut down ClearVector, but the main reason was 2010 marked the year I (temporarily) gave up all hope of running my own business and went and got a real job. I shut down my freelance business and hosting business. I don’t think ClearVector would have lasted even if I had continued freelancing though, as shared hosting is such a crowed market, I never had many customers, and support took up a lot of my time.
Shared hosting is a race to the bottom. I used to wonder how Heart could offer unlimited reseller hosting for £40 a month, surely they’d lose money? That was until I realised that one of the biggest costs in hosting is time spent on customer support. Reseller hosting is great for Heart (and other hosts that offer it) as it allows them to offload the customer support onto poor unsuspecting re-sellers. Seriously, shared hosting is a suckers game, don’t do it.
Also, don’t give people too many choices, most companies offer packages for a reason.
Find a Co-Founder for a Startup
In 2010 I discovered Hacker News, and learned about things like minimum viable product and test driven development. I subsequently spent far too much time hoovering up all the knowledge and information I could from Hacker News, its users, and its creator Paul Graham.
A recurring theme on Hacker News was how single person startups don’t work, Paul Graham even listed “Single Founder” at the top of his list of 18 things that kill startups. It turns out this line of thinking is complete bullshit, but at the time I bought it into it.
I decided to build Founder Finder partly because I wanted to try test driven development, and building an MVP. But also because I didn’t know anyone who would fit the ‘business guy’ role in a startup, so I set out to build what was essentially online dating, but for entrepreneurs to find each other with a view to starting a company.
At the time I couldn’t find any similar services, although now there are quite a few, so in the early mornings before my 9-5 job, I’d spend an hour or two building a very basic app where people could create a profile, listing their skills and location, and find potential co-founders with complimentary skills.
Although I built a fully working (and tested) MVP, I never actually launched it.
At the time I was into buzzwords and bandwagons, so of course I was planning to make it Freemium. Had I actually launched this, I probably wouldn’t have made any money anyway.
A couple of factors caused me to abandon this one before launch. Partly the chicken-and-egg problem (can’t get users until I have lots of users), and also because I started to wake up from Silicon Valley echo-chamber and realised that single founders can work out just fine, and that freemium often doesn’t work.
Don’t get sucked into “hot right now” bandwagon chasing. Test Driven Development and Minimum Viable Product are both valid methodologies though, which served me well with BugMuncher.
Visual Website Feedback
You probably know the story of BugMuncher, if not you can read all about here. I don’t consider BugMuncher to be a failure, but I’m including it in this timeline for reference, and as it wasn’t really a success until I went full time on it in 2015.
Find Camp Sites suitable for Motorcycles
In the summer of 2012 my fiancée Sophie* and I spent six weeks travelling Europe on a motorcycle. It was an amazing time, and something we’d both love to do again. One thing we learned while travelling is most campsites are not really suitable for those travelling by motorbike, and many really only cater to caravans and campervans, with tent pitches tacked on as an afterthought.
So while travelling we hatched a plan, start a company that would certify campsites as “Biker Friendly”. This was probably the most well thought out business idea I’ve ever had, and I suspect that is because I couldn’t start working on it while we were travelling, so we had lots of time to discuss options and come up with a business plan.
* Sophie was just my girlfriend back then, we got engaged in July this year! 😄
The plan was to do a soft launch, ie: have a website which would capture email addresses of bikers who would be interested. Then, we’d tour UK campsites including any that we considered suitable in the directory.
Once we had a decent number of certified biker campsites, we’d launch the directory, emailing everyone who’d registered their interest. Assuming this worked, we’d then expand into Europe and the rest of the world.
Campsites would pay to be inspected, and potentially certified, as we could bring them new customers. Bikers looking for campsites would have free access to the list.
This was another non-starter, I still think it’s a good idea, but once we got back to England, I had loads of work to catch up on, and we had to find a new house - basically life got in the way. Shame really. Another issue was the time and budget required to physically go and check all the campsites. Testing the campsites would be a pretty awesome job though.
A good idea isn’t worth anything if you don’t actually execute it.
Easily track business use of a personal vehicle
In 2014 I built MileDiary, a simple web app for tracking business miles travelled in the user’s personal vehicle. In the UK, you can claim up to £0.45 per mile as an expense when using your personal vehicle for business. I had a bad habit of forgetting to log my business miles, so I built MileDiary.
MileDiary is a simple web app that can track business miles using your phone’s GPS and google maps, as well as allow you to easily repeat common journeys, and give a report of how much money you need to claim back.
My launch plan was a little bit better then usual, I was going to try get in with accountants on a referral scheme - they could then get their clients to sign up.
It was going to be a subscription model, probably around £5 / month.
I never actually launched MileDiary, as while beta-testing it myself, I still kept forgetting to use it, so it didn’t actually solve the core problem. Also my client base changed to a less local one, so I found myself rarely driving for business anyway.
Solve the right problem. I was forgetting to log my business miles, a fancy app didn’t change that, it just made it slightly easier when I did remember.
Find your next restoration project
Even after going full-time on BugMuncher, I can’t resist the allure of the side project. The idea here was to create a site that would import listings from eBay, Auto Trader, Gumtree and a bunch of other places where people sell cars, but only import those that could be projects, eg: non-runners, mot-failures, classics.
That was phase one. If the site gained traction I’d stop the import and allow people to list on the site directly, either free or for a small fee, and with the option of promoting their listing at an extra cost.
This is another that never launched, to be honest I only ever spent a couple of days on it, playing around with the eBay API, and figuring out how best to filter so it only imports project car listings.
Had I launched, revenue would have initially come from adverts, but once the site started having its own listings, revenue would come from those as well.
BugMuncher was (and still is) too time consuming. I’d still like to take this one further, but I’ve not found any more time.
A lot of my ideas involve cars in some way, which isn’t a huge surprise as I’m a massive petrol head. I think my next proper attempt at a side project should be car related.
Simple tool for sending and managing cold-email outreach campaigns
This one doesn’t have a real name, as it originally wasn’t built so much as a side project, but a tool for me to use when marketing BugMuncher.
Cold Email Thingy allowed me to very quickly add new contacts for an email campaign, intelligently guessing their name and business name from their email address, and then send simple emails to them, tracking opens and clicks.
Once I’d added a load of prospects and sent my first campaign, I realised I’d built something pretty cool, and decided it might be worth turning into a service others could use. I never got much further than that though (see cause of death)
This would have been a subscription model, probably based on the number of emails sent each month.
While adding prospects and setting up campaigns was really easy, the problem was performance. I got pretty good open and click though rates, but no one ever signed up to BugMuncher from one of these campaigns. The issue was volume, I’d probably need to send thousands of emails before anyone signed up, and tens of thousands to get a new paying customer.
I had deliberately chosen to manually find and add email addresses, as I didn’t want to do anything that could be considered spam, so buying lists was a no-go. Unfortunately, even with my fancy interface, the time required to add prospects was too prohibitive.
While I learned that manually adding emails takes too long to be effective for cold email outreach, I wouldn’t do anything differently. I only spent a day building the basic tool, tried it on a small sample, and found it wasn’t viable. On to the next one.
So there’s a selection of some of my failed business ideas, there have been others, as well as countless open-source libraries and abandoned git hub repos.
I stand by everyone of them, some of them weren’t total failures, a couple I may even try to resurrect one day, but all of them taught me valuable lessons.