Once upon a time

So many of the things we believe about what it takes to create and grow a successful business are wrong.

This is the fairytale:

Our hero, a lone genius slightly lacking in social skills, has an amazing new idea for a website or app. They lock themselves away in their bedroom and stay up all night coding, powered by energy drinks. They launch the next morning and are immediately inundated with customers. Shortly after that, they convince a slumbering corporate to buy their young company for squillions and then retire to spend the rest of their life happily sipping mojitos on the beach of a remote tropical island.

So many of these things that we think we know about what it takes to create and grow a successful business are wrong. It’s very distracting for anybody working on a startup, or considering it.

I count at least eight separate myths contained in that simple description.

Let’s start with three…


There was a time when I imagined myself as the central character in this kind of story.

In 1999 I was living in Sydney and working as an IT consultant on a large Y2K project that was slowly destroying my soul. The dot-com bubble was fully inflated. I had a nagging fear that I was missing out.

As a software engineer I was curious about the emerging web development tools. I started to get my hands dirty on evenings and weekends. But I had no idea what to build and as a result spent quite a long time going around in circles.

My motivation wasn’t to start a company, it was to experiment with technology and learn what was possible. I was looking for the sweet spot between “challenging to build” and “achievable with limited time and skills”. I didn’t really think about the potential business model at all. I was much more interested in the data model.

The idea I eventually settled on came less from a moment of inspiration and much more as a result of discontent.

My fiancée was living in Wellington, so I was a very frequent flyer and motivated to move back to New Zealand. Back then classifieds were not available online and finding a good flat meant getting up really early on a Wednesday or Saturday, buying a newspaper, and hitting the phones. By lunch time all the good ones were taken. That wasn’t very convenient for me in a different timezone.

So I put those two things together and came up with the idea for what I’d eventually call Flathunt – a website where anyone with a flat or room to rent could advertise their properties, connecting with potential tenants or flatmates.

It’s difficult to pinpoint a single “Eureka!” moment. The website was built like a jigsaw, in small seemingly insignificant steps, over the course of several months. One evening I worked out how to let people search and filter listings; one lunch time I sketched out the user interface, which defined the scope of what I needed to build; waiting at Sydney airport on one of my trips home I found that flathunt.co.nz was available, so I had a domain name; one weekend I found a local web host that supported databases. They were a “cloud” service provider years ahead of their time, eliminating the need for expensive server hardware.

When we make something new it’s a natural instinct to hide and wait for the perfect opportunity to reveal the completed masterpiece. It’s hard to avoid the tempatation to keep adding just one more feature. These days I spot this behaviour straight away. Inventors ask me to sign non-disclosure agreements before they are prepared to say anything, whereas entrepreneurs will insist on talking, even when they don’t have much to show yet.

The best thing to do is to ask for considered feedback from anybody who will give us an honest opinion. That is easy advice to give, but much harder to do in practice. It takes a lot of courage to tell others about our half-formed ideas, because chances are they won’t be anywhere near as excited about them as we are.

Thankfully I managed to avoid this pothole with Flathunt. Partly because I never considered what I was building to be especially original or interesting. But mostly because I had lots of questions and there were people who I hoped might have some of the answers. Even all these years later I’m still grateful to all of them, because with their help what I was building got better, little by little, and as it did my confidence and impatience increased. Eventually that combination got me over the line.

So now I can look back and retrospectively invent a story about the mythical magical moment when the idea popped into my head, perfect and fully formed.

Let’s pretend!

The firehose

Building something new is great, but working out what something is for is where things really get interesting. To create a valuable startup we need to execute a good idea.

Lots of effort is wasted by founders who are worried about being inundated at launch. Unfortunately, despite all the work done in preparation, the anticipated deluge of customers rarely eventuates.

Trade Me took over four months to get 1,000 registered members, despite being completely free to use. Xero listed on the NZX with only 100 paying customers, and those of us working there at the time accounted for a good portion of those. Vend had just six paying subscribers when I first invested in 2010. After 12 months of hard work we had about 200.

I was extremely naïve about how hard it would be to get anybody to use Flathunt. I’d hoped it would be quick and easy. It was neither. I tried everything I could think of with my limited budget.

I printed out advertisements and pinned them to notice boards at Victoria University where students advertised available rooms. Instead of including a phone number, the tear-off tabs had the website address. The limited number of tabs ripped off reflected the modest level of interest.

That was humbling, but at least resulted in the first few listings from people that I didn’t personally know. I was still building the website, so when somebody placed an advertisement it only generated an email to me. I would manually enter the details into the database to create an illusion of automation.

This is sometimes called “Flintstoning” (imagine the car from the old cartoon which is actually just powered by the feet of the passengers). It’s true that automation is critical to scale. But it’s also true that we can automate a process too soon - before we really understand the exceptions, how the process is used by customers and how to implement it efficiently. As it turned out, the number of emails I got wasn’t overwhelming at all.

I also bought the newspaper, and called everybody who had placed a traditional classified to ask if I could add their listing to Flathunt. I figured they would have nothing to lose, and some of them were willing. But the more interesting lesson was how many of the classifieds were placed by the same small number of property managers. One of them got quite annoyed when I called them for the third time in one morning.

That revealed a business model and generated my first significant sale, to a real estate agent in Palmerston North. I scrambled to build the “pro” listing features I’d promised.

Over the summer I hit the road and door-knocked every real estate agency and property manager I could find. They were mostly disinterested. I got a much better response at university campuses where there were more internet-literate people in need of a room or a flatmate who were also poor enough to be reluctant to shell out $30 for a newspaper classified and so willing to give something new a go.

This is a repeated pattern. In the beginning, the biggest category on Trade Me was computer parts. Later it was womenswear. There isn’t a lot of overlap in those categories. Likewise for Xero, many of the early customers were independent contractors, who didn’t need the more complex accounting features we were still busy building, and so were happy to use it even though the product was incomplete. Often the first customers are very different from those that come later.

David Randal with the homemade Flathunt banner, Lake Tekapo, 2000
David Randal with the homemade Flathunt banner, Lake Tekapo, 2000

A few weeks after Flathunt launched I got some media coverage, when a friend introduced me to Tom Pullar-Strecker. He published a short story in the “InfoTech Weekly”, which included this slightly-cringy-in-hindsight quote:

There are a few other sites out there charging $1 a day to advertise. Even that small amount is enough to turn landlords and agents away. The internet works best when it’s free.

He asked me what my job title was. I thought it was an odd question as I was the whole company. I said “Managing Director”. After the article ran I got a phone call from somebody asking to speak to the “Head of Marketing” and had to explain that I was that person too. That was an embarrassing reminder of the downsides of pretending to be something I wasn’t.

I was desperate for listings. I had a chicken-and-egg problem: advertisers didn’t see value in listing on the site because there were no listings on the site; meanwhile anybody searching on the site didn’t find much there, so they had no reason to come back.

But I knew it worked. Soon after I moved back to New Zealand, we found a flat for ourselves - it was listed at 2pm one afternoon, we went and saw it at 4pm and signed the lease later that evening. Which was great, except that our new landlord immediately deleted the listing, so that was self-defeating!

Screenshot of Flathunt website, late 1999
Screenshot of Flathunt website, late 1999

Two critical lessons from this time stand out from the rest:

Firstly, I learned that improving the website made little difference (despite significant room for improvement). The much bigger constraint was that nobody knew it existed - and I wasn’t very good at sales or marketing. Growing the business wasn’t about making myself more efficient, it required skills I didn’t have and couldn’t afford to hire.

It’s very tempting, especially for technical people like me, to focus too soon on optimisation. In practice it nearly always ends up being premature. We have more time than we think, and adjusting when needed is not as difficult as we imagine. At that point we don’t start with a blank sheet but with all the lessons about what works in our particular circumstances, which bits of the application are popular and where the biggest bottlenecks are.

At Trade Me we used the Team NZ mantra to remind ourselves to focus on critical things:

Does it make the boat go faster?

I still think that’s great advice for companies that are growing, but before we can worry about that we need momentum. More relevant advice might be: 1

You cannot steer a boat that’s not moving.

We only have to expand our operations once we have operations. Before we have customers there are definitely more important things to work on… like getting customers.

Secondly, I learned that people only tell their friends about things that are remarkable. Unless there is something genuinely different (or terrible!) about our idea it probably won’t be news, so we’ll need to find some other way to spread the word.

At Trade Me we were fortunate to have an idea that was inherently viral. Once a new user sold some stuff that they assumed was worthless, they couldn’t help but tell everybody they know about it.

The idea that Trade Me didn’t spend any money on traditional marketing, like billboards, in the early days is also a myth. We were mostly lucky that we didn’t have much money at that time, or we might well have wasted more of it on things like that!

It wasn’t so obvious at Xero in the beginning, but focusing on accountants and bookkeepers has proven to be an amazing way to spread the word and get in front of millions of potential small business customers. Co-founder and CEO Rod Drury was frequently featured in the media talking about broadband and specifically international fibre connections, and was able to leverage that profile to promote Xero to a wider audience. That was a deliberate strategy.

At Vend we discovered that irreverent viral videos shared on YouTube were a great way to reach our target market of boutique retailers. We complimented that with Google AdWords, which in those days were still relatively cheap, to generate in-bound sales leads, and worked hard to make the sales process as low-touch as we could.

A Vend billboard concept from 2010 - with the benefit of hindsight it’s probably best this never saw the light of day, other than on social media
A Vend billboard concept from 2010 - with the benefit of hindsight it’s probably best this never saw the light of day, other than on social media

Timely had to work harder for coverage and attention. One of the unique things about the business was the team was distributed and fully-remote. In a lot of the early media coverage and commentary that was the hook. We took what we could get and made the most of it.

However, before all of that, I discovered the hard way that there was nothing inherently viral about Flathunt. I wasn’t blessed with a lucrative distribution channel. And in early 2000 Google AdWords hadn’t even been invented yet. I quickly realised that growing the business was going to require constant work and constant expense.

For all my ambition, the awkward truth is that Flathunt never really worked as a stand-alone business.

After about three months I’d burned through our savings and needed to start paying my share of the bills. Flathunt was making enough to cover costs, but not enough to pay me a salary, so I started doing some part-time contracting work.

Around that time I had lunch with Sam and Phil McCaw, who along with his colleagues had just invested in Trade Me. They had put in $100,000 and in return got 50% of the company! They needed a software developer. I agreed to work three days a week to start with. There was no shortage of things to do:

Trade Me homepage, as it was in 2000 when I started working on it
Trade Me homepage, as it was in 2000 when I started working on it

Soon after that they offered to acquire Flathunt provided I agreed to join the team as a full-time employee. I don’t think they had much interest in Flathunt. And “full-time” was a bit of an understatement.

Within the first year we completely redesigned the site, introduced success fees, and usernames (rather than just showing everybody’s email address in listings) and built photo uploading, and autobidding.

It’s hard to imagine today that Trade Me could have worked at all without those features, but somehow it did.

There were also a number of experiments that were not a great use of our time in hindsight - such as a WAP version of the site for early smartphones that was likely only ever used by the team at Telecom who paid us to build it, and an SMS bidding system that was many years ahead of its time.

Flathunt meanwhile continued to operate in the background. Years later it would be folded into Trade Me Property.

I’d stumbled into what would eventually be a very lucrative gold mine. In seven years the $10,000 I’d invested in getting Flathunt started would return $30 million.

More importantly to me at the time, I’d found something exciting to work on that had momentum, and a complementary team of people to work with that I could contribute my skills to. They were the good old days. I was all in.

Mojito island

A flashy launch or a lucrative exit are news. But all the interesting steps come in between. Let’s fast forward to March 2006 when the small group of Trade Me shareholders gathered at the wood-panelled Buddle Findlay offices in the Wellington CBD on a bright Sunday morning to sign a pile of paperwork and finalise the sale of the business.

David Kirk, Fairfax CEO and former World Cup-winning All Blacks captain, was there on behalf of the buyer, the Australian public company Fairfax Media. The deal had taken several intense weeks to finalise, and everybody was anxious to close it out.

My main memory of that morning was the complications of arranging child care. Our 18 month old had been up all night vomiting. We had all been sworn to secrecy, so we were in the awkward position of having to call a baby sitter and ask them to come despite that, without being able to adequately explain why. “We’ll tell you tomorrow”, we said, “You’ll understand then!”

After the formalities were completed, we walked home through crowds of people doing their weekend shopping. I had a strange feeling, knowing what was likely to be the headline news story the next day. I had no clue how that would land with the general public, most of whom were Trade Me members by then, but more importantly with my teammates and family members. Was this an inflection point? And if so, up or down?

The next morning I went to the Intercontinental Hotel where Sam and David held a short press conference to announce the sale. I stood at the back with Mike O’Donnell (known as MOD), who was Head of Operations at Trade Me. I recall the reaction from many of the business journalists present when the price was announced. Bernard Hickey, who was working for Fairfax at the time, asked us to confirm that the decimal point was in the right place! He wasn’t the only one who thought that Sam had somehow tricked them into adding an extra zero to the price tag:

Tom Scott in The Dominion - March 8, 2006
Tom Scott in The Dominion - March 8, 2006

$750 million was (and is!) a staggering amount. But the real news that day was just how profitable Trade Me was. The price paid was a multiple of the earnings and reflected the high-growth trajectory the business was on. This was the first time that information was in the public domain.

To attract the interest of an investor or acquirer for any venture the two basic motivating emotions to tap into are fear and greed. We need to have new (and often more importantly) growing revenue, or we need to be threatening existing revenue elsewhere.

Trade Me had both.

We had decimated the revenue media companies previously earned from classifieds. We joked that we were just working our way backwards through the sections of the newspaper - we’d done Motors and Property and were poised to launch Jobs then Travel at the time of the sale. And we had great margins. A large portion of every dollar we earned was profit. In the year before the sale we made $26 million. In 2007 we nearly doubled that to $45.5 million and then $70.1 million the following year.2

In those days Trade Me was a fun and relaxed place to work. We didn’t wear expensive suits or take ourselves too seriously. The office was a bit chaotic and had graffiti on the walls. But we were 100% committed to the job. There were many people who confused our lack of formality for a lack of ambition. But it was a mistake to underestimate us.

After the press conference David came back to the office to meet the rest of the team. He understood this better than most people, but we still suggested he remove his suit jacket and tie, so he didn’t seem completely out of place. Compared to the thousands of employees in the other parts of his corporate empire I imagine it must have seemed like a small and very young team. Within a few years the revenue that Trade Me contributed to the Fairfax group would become a significant part of their business.

Did we make the right decision to sell? Was $750 million enough? It created so many opportunities for myself and others, I never stopped to think about that at the time. I had the same feeling when Vend and Timely were acquired in quick succession in 2021.

In all three cases, half the uninformed reckons were “You sold too soon!” and the other half were “How is it possibly worth that much?”

Kiwi business people are sometimes criticised for a lack of ambition — selling out once they have enough to buy a bach, a boat and a BMW. That’s nonsense. Of course we should all aim higher than that, but I don’t think this means aspiring to own a private jet or super yacht instead.

If we measure ourselves based on how many expensive toys we can buy, we’re never going to be satisfied. There is always another level above – a more expensive car, a bigger house, a longer race.3 As soon as we achieve something, we quickly normalise it and start to take it for granted. There is nothing so amazing that we can’t get used to it. Then we start to compare ourselves to people who are slightly ahead of us. Our appetite is always slightly bigger than our stomach.

To make this worse, as I’ve experienced multiple times, when we are part of something that is successful the very first thing everybody asks is, “What are you going to do next?” That weighs you down!

Consider this formula: 4

Happiness = Wanting What You Have / Having What You Want

How much time should we spend on each part of that equation? Anybody who understands how fractions work realise that the important thing to focus on is increasing the numerator (the wanting) not the denominator (the having).

Until we experience it, it’s difficult to describe how much better it is to work on something we’re invested in. It’s not 2x or 3x better, it’s more like 100x or 1,000x.

After the excitement of a startup that has momentum, and the rewards that come from that, sitting on the sidelines can feel like a step backwards. Even the biggest boat we can buy won’t compensate.5

We have to understand why we’re doing it all in the first place. Financial rewards are a way to keep score, but we need to be careful we don’t get too hung up on them, because even if we do everything right, it still may not work out the way we hope.

The only good reason to sell is when we’re more excited about what we’re going to do instead.

I didn’t realise it at the time, but I had a lot of unfinished business.

  1. Mike Carden, Twitter, May 2010. ↩︎

  2. Fairfax Media Reports Net Profit Of $386 Million, Scoop, 21st August 2008. ↩︎

  3. For example, the Ultra Ironman↩︎

  4. Measuring what makes life worthwhile, by Chip Conley, TED Talk, 2010. ↩︎

  5. Mojito island is a mirage, by David Heinemeier Hansson, 37 Signals Blog, 21st September 2009. ↩︎

Related Essays