#0122: When it comes to failure, timing is everything
Braingasm
I have been lucky enough to fail more than a few times in my career.
I have been fortunate that those failures have not been too costly, at least in terms of sanity or family or friends. I was even able to learn from a few of them, although perhaps not quite as quickly as I would have liked.
So, here are three failure stories (out of many!) that ultimately all share a single timely lesson.
Failure 1: Admiral
Back in the mid-90s, I pitched an idea called “Admiral” to an advertising agency exec in Sydney.
He told me (paraphrasing from memory):
“That will never work because of the way that ad buyers buy ads. They do not do this online, and they never will. They do it over the phone, they always have done it over the phone, and that’s the way it will continue. There are too many ad buyers and agencies that rely on this model for it to move online. So, thanks for the chat, but no thanks, I don’t think your idea will ever work.”
The idea underpinning Admiral was to create a simple marketplace where companies could upload their banner ads, and then Admiral would work out where to place those ads onto sites that had subscribed to take the banners. Admiral would give the majority of the revenue to the site, and then take a % cut for itself. It was a pretty simple idea, but as far as I could work out, unique at the time. After mucking around with some ideas and a simple prototype, I took the advice of the ad exec at face value and threw the idea in the bin. That was a bad idea. This basic idea is what ended up being DoubleClick, Overture, and ultimately, morphed into Google AdWords.
In hindsight, this was a pretty cheap failure in that it didn’t cost me or anyone else any money at all. But the worst part about it was that I misinterpreted the outcome. Rather than seeing the mistake for what it was — one of timing — I interpreted it as a bad idea. I assumed a priori that the ad exec knew about advertising, and that I was just off on the wrong track. I was just a twenty-something kid with a crazy idea that could never work. What I should have done was asked a few more people to get some other opinions.
Most importantly, what I should have done was just backed myself. At that age, although the likelihood of failure is pretty high, the impacts of failure are pretty low. Without family and mortgage and the like, you can afford to have a very different attitude to risk. I didn’t understand that at the time, and so Admiral ended up being consigned to the dustbin.
Failure 2: TouchCoin / TouchPass
Working in and around payments for some time, I was always on the lookout for the next thing. After seeing the original iPhone and it’s touch interface and the early version of the App Store, I was experimenting with something that I could do with it. Fortuitously, I stumbled on the idea of “flicking” money from phone to phone using a simple hand gesture.
The idea of a simple, efficient, person-to-person payments system had always appealed to me. Before the iPhone, the technical barriers to building something like that seemed insurmountable. Then, along comes a device with a crazy touch interface sitting on top of what felt like a super-computer. It was just the tech platform I was waiting for to build TouchCoin.
However, as soon as I started thinking about P2P payments, it became apparent that I needed some way to authorise and authenticate P2P transactions for security purposes. Fortunately, the iPhone had something else novel: a GPS. From that, the idea for using “location as an authentication factor” occurred to me. That is, I could use the phone’s current GPS coordinates to work out if the two people exchanging money were proximate, and then use this as a factor to authorise the transaction. If the devices were proximate, the transaction was more likely to be valid, and if they were not, then the transaction was less likely to be correct. There’s a ton of other stuff that has to go on, but it turned out that this” location as an authentication factor” was a pretty novel idea.
At about this time, I was doing some work for a couple of Australian banks, and they were struggling to secure their online banking apps because of serious fraud. I worked out a way that we could use the location authentication parts of TouchCoin into an authentication mechanism for online banking transactions. The theory was that while I was trying to get TouchCoin up and running, I could sell TouchPass as a novel second-factor authentication mechanism that would help banks secure their online transactions.
There are a series of cascading failures with this story.
First up, just about the same time as I was trying to sell TouchPass to the banks, the largest bank decided to roll out SMS-based one-time-password (OTP) security for its online banking app. The landscape in Australia changed immediately for banks — and customers — for what constituted acceptable online security. Soon after, 2 of the other three big banks followed with SMS-based OTP security mechanisms. In a short period, the market went from wide open, to firmly closed. Once the banks decided to go with SMS. Once the public grokked accepting OTPs via text message, there was just no way that anyone was going to try to roll out some other mechanism, regardless of how much better it was in terms of security and user experience.
The next mistake was trying to sell something like this to a bank. Banks (at least in Oz) simply don’t buy stuff from startups in general, and they definitely do not purchase security products from startups. I should have realised that in advance and saved myself a whole lot of heartache.
The final mistake was getting way, way too excited about the technology and ignoring the product. The tech was pretty amazing, but I realised much too late that no-one cared. In the end, I had a fascinating, novel, perhaps even unique piece of technology that existed in complete ignorance of either a market or potential set of users.
Failure 3: Privacy Respecting Proximity
During the course of the TouchCoin / TouchPass project, I managed to develop a bunch of, what I thought was, patentable IP. One particular piece of IP was a thing called “privacy respecting proximity” (PRP). The idea behind PRP was to be able to compare two physical locations for proximity without the need to reveal (to anyone) the underlying physical locations. This was going to be the core part of the TouchPass authentication mechanism.
I spent a significant amount of money with a patent attorney trying to get the patent filed in the US and Australia. The Australian Patent Office granted the patent, but the US patent office dragged its feet.
What I subsequently found out is that the patent attorneys were just jockeying paperwork to a US firm and charging me for the privilege. It was impossible to communicate with the USPTO in any meaningful way via all of these hoops (yes, sometimes even over fax) and so the case dragged on and on and on until I ran out of money. At that point, I had to abandon both the Australian and the US patent application processes and just let the idea go into the public domain.
What I learned through this process is that patent law is complicated and expensive and that it is not always obvious where your interests lie or who is looking out for them. At some points, it felt like I was handing over cash for kindling. I also learned that dealing with the USPTO is an absolute nightmare, and not something that someone like me should have ever embarked on my own. Even after being an inventor named on multiple patents during my work at Nokia Research Centre, nothing had prepared me for the cost and time that it would take to file a patent, even with the help of patent lawyers.
And finally, timing. Yet again, this issue of timing raises its head. I first thought up PRP as part of TouchPass in about 2008 or so. Can you imagine how useful a patent on the idea of anonymously comparing locations for proximity (in a privacy respecting way that does not reveal an underlying physical location) would have been in the era of smartphone contact tracing paranoia during COVID19?
3 failures, 1 lesson
Across these stories, there is a single, unifying theme:
Over and over again, timing is one of the most important, if not the most important factor when trying to launch a business. You can get a lot of other stuff wrong and still succeed, but if your timing is out, you have no chance.
3 extra lessons — at no extra charge!
And now, at no extra charge, here are three extra lessons!
1. If you are going to try something genuinely different, you need to back yourself and be prepared to fail. You do not learn anything by doing nothing.
2. It is almost never about the tech, and even when you think “this is the time that it is about the tech” it still probably isn’t about the tech.
3. To the naive observer, it feels like software patents are a tool used by large organisations to create barriers to entry for small organisations, aided and abetted by very expensive patent lawyers. For the uninitiated, software patents appear to have little to do with promoting innovation and everything to do with stifling it. Unless you know what you are doing, or have people helping you who do, stay away. You have been warned.
Regards,
M@
[ED: If you’d like to sign up for this content as an email, click here to join the mailing list.]
