It was 2002. The dot-com bubble had burst and you only mentioned the internet in hushed undertones now. Industry executives had consigned IT to the back office once again. Sure, companies needed to buy PCs and physical servers but that was just a small cost-of-doing business. Typically, the cost was accounted for under depreciation. And since you weren’t being accounted for within EBITDA, you were not even a part of most board room discussions.
The insurance industry was little different. The call center revolution had taken place in the 1990s but few believed in the power of the internet for selling a product as complex as insurance. It was marketing 101, really. Insurance was a push product. You could only sell it in person or over the phone with a sales rep.
Against this backdrop, a few mavericks in South Wales were convinced that the internet could help disrupt insurance. These individuals were managers at Admiral Group, an up and coming car insurer in Wales. Together, they imagined a world in which a single website would act as a knowledge broker for insurance quotes. With this goal in mind, they launched Confused.com in 2002. The name of the website was an ode to the relatively taxing task of buying car insurance. The website design was inauspicious and simple. Here is how it looked in its first couple of years:
I did a Google ‘news’ search for the 2002-2004 period for Confused and it seems there was no mention of the launch or anything in any major newspaper. Nothing. Nada. Zilch. It is as if Facebook has launched and it doesn’t even make it to Page 22 of the tabloids. And yet, in the coming years, this one company would completely transform a £10B+ industry. Here is how the share of car insurance sold through these ‘aggregators’ would track over the next 16 years:
Confused’s success also spawned a series of successful copycats - notable among them GoCompare and Comparethemarket - companies who would go on to beat Confused at their own game. While Confused would soon lose its market leading position, the shift towards aggregators would still be hugely beneficial for Admiral who would become the largest player on this new channel. As a result, Admiral would very soon go on to become one of the UK’s largest and most profitable motor insurers.
Buoyed by the success of the aggregator model in the UK, Admiral was convinced it could replicate its success in the United States. In 2013, it launched a similar price comparison service in the US. Initially called CompareNow, it subsequently took on the simpler Compare.com name and domain. The rationale for the launch was quite clear. The US was a market 10x the size of the United Kingdom. The scale benefits of aggregating marketing spend at this scale were much bigger. The smartphone revolution had taken place by now. As a result, everyone owned a hand computer. And pretty much everyone was on the internet. All the time. Moreover, the Admiral team had also accumulated a decade of experience of running and managing a large price comparison website.
Only a foolish man would have challenged the launch idea and team. And yet, over the subsequent six years, Compare.com struggled to gain traction. There are many reasons as to why comparison shopping hasn’t taken off West of the Atlantic. However, this essay isn’t really about any of that. Rather, this post is about an idea that I call science envy. It is about how our innate desire for order causes us to see patterns where none exist. It is about why you should never read a business history book and become convinced that there is a ‘right way’ to promote innovation. It is about why timing matters more than teams and ideas. And ultimately it is about reconciling the differing reception that Confused and Compare got when they launched.
First, let’s start with a primer on what science envy really is about.
The phrase actually borrows from a related term, ‘physics envy’. Physics envy is typically thrown around by physicists at economists. And boy, do economists hate it. Here is the official Wikipedia definition for the concept:
“The term physics envy is a phrase used to criticize modern writing and research of academics working in areas such as "softer sciences", liberal arts, business studies and humanities. The term argues that writing and working practices in these disciplines have overused confusing jargon and complicated mathematics, in order to seem more 'rigorous' and like mathematics-based subjects like physics.”
I believe that modern humans suffer from a more general disease. I term it ‘science envy’. Physics envy then really is just a sub-set of this original disease.
Here is how it works. Our education and experiences train us for a world where we are taught to see patterns and identify cause and action. You study, you get good grades. You press the accelerator, the car moves forward. You kick the ball, it flies in the air. And so on. And such knowledge works tremendously well for us in most contexts.
However, these experiences subconsciously condition us to expect that the business world works in a similar manner. I.e. if you do these ten things, you may also have a successful business. If you do A, B and C, you are guaranteed to become a successful entrepreneur. There is a multi-billion dollar self-help and coaching industry that relies on this insight.
The truth is that there is no magic bullet for business successes. Sure, you can find some interesting patterns among successes. But that is all ex post facto. The dirty little secret is that nobody still knows what works. Often there are so many forces that combine to shape reality that having a checklist is just not an appropriate way to evaluate ideas.. Sure, Zuckerberg’s hacker mentality might have helped Facebook a lot in its early days. However, that is not the reason Facebook is such a breakaway success. If anything, there exist probably a billion parallel universes in which Facebook gets eaten by the MySpaces and Orkuts of this parallel world. And in those alternate realities, historians would likely mention how Facebook’s hacky culture meant that it lead to platform which was often breaking and skirting privacy regulations. Popular media representations like ‘The Social Network’ gloss over such truths to create a compelling narrative. Ultimately, people end up glamorizing the wrong qualities and traits. Quite often, they confuse signal with noise when in reality what you have is merely a hodge podge of factors. What works in one context quite often fails in many other contexts. As such, the only thing you are able to extrapolate is a few general set of rules but these are so high-level that they hardly constitute a scientific or quantitative approach.
Successful VCs understand this dirty little secret. That is why even for top-decile funds most investments still go to zero. A few return 1-3x after 7-10 years. And the select few that really make it return all the fund. A few successful operators have even admitted that their most successful investments were ones in which they did not do any due diligence or modeling and just acted on their gut. The cynic among you might say that the gut of these VCs was really trained by decades of experience picking successful teams and ideas. However, the annals of history are filled with seemingly great ideas run by good teams that didn’t go anywhere. Let’s look at a few.
Think Grocery delivery over the internet is a great idea? I mean you can’t really argue with that. Instacart, founded in 2012, is worth $8B+. Traditional players like Amazon and Walmart have been pouring billions into the segment. However, unbeknownst to most, Webvan tried to achieve exactly the same in 1999 and failed spectacularly. This despite the fact that the company raised $800M+ and had a stellar executive team.
Convinced that Uber is the best idea and can’t imagine your life without it? However, it wasn’t the first ride-hailing company of the smartphone era. A company called Taxi Magic had been on the horizon and was offering the same service a year before Uber’s arrival. (except the skirting regulations bit) The company, eventually renamed itself to Pivot but remains a small niche player with less than 1/100 the value of Uber.
Love your pooch but haven’t got time to take it for walk? There is a $650M company called Wag built on solely that idea. Admittedly, that is a SoftBank valuation and could be taken with a bucket load of salt. But decades before Wag made dog-walking cool, Pets.com raised over $100M in 1998 on the more reasonable promise of selling dog food online. However, the company was forced to close its doors just two years later. It was 2000 and market wasn’t ready for home delivery for such a product just yet.
Addicted to Amazon’s 2-hour delivery guarantee? Well a startup named Kozmo promised to do just that in 1998. And it was no slouch in the fundraising department either as it raised over $250M. Notable among its investors - Amazon.com - yes, Amazon invested $60M in this company.
Use Wikipedia daily to learn about random things? Well, long before Wikipedia made Encyclopedia irrelevant, a company called Nupedia had worked on the same idea of an open-source and collaborative knowledge-base? It’s fault? Well other than one of timing, the fact that it had a much more rigorous peer review process.
Anyways, I won’t belabor the point further. I do hope that I have convinced you that success in the business world is a bit like playing blackjack. Sure, knowing probability, being able to do mental math and experience all increase ones odds of winning. But ultimately, it is very much a game of chance. As such, very good players often end their nights in the red. But the very best ones - they know not mix skill with luck - even as they stack up losses in successive nights.
Consider this, Sequoia - universally considered to be one of the best venture firms in the valley - was actually a lead investor in Webvan (mentioned above). It must have been a humbling experience. And yet, Sequoia still led the $8.5M Series A for Instacart in 2013. For literally the same idea. And that stake has likely increased more than 100x in value by now.
Looking back on our earlier examples of Confused and Compare, most VCs would likely have backed Compare and passed on Confused. Everything about Compare.com at its inception was better - the market was bigger, the team was more experienced, the timing (in terms of # of internet users) was better, the UI/UX was sleeker and even the domain name was way better. And yet, the fortunes of the two companies have played out in reverse.
Ultimately, it is worth always remembering the role that luck plays in the business world. As such, the very best ideas with the very best teams will sometimes struggle. Some good ideas with bad teams might make it if the timing is right. And there might even be a few bad ideas with bad teams that will survive for years and remain hidden under the sizable income statements of large public companies (aka the ‘rounding errors’). All startups and most VCs know this point around luck and chance. Startups and entrepreneurs live by this lesson everyday. VCs see it with the carry in their portfolios.
Publicly traded companies often act as if they have forgotten the point around luck. Many internal innovation programs fail because they ignore this reality. Armed with large balance sheets, such companies expect to use capital as a proxy to increase innovation throughput. The sad truth is that capital is no longer a constraint in the Western world. That is why even today you have over $3 trillion in assets that sits idle and earns 0%. And there exist 30-year old bonds that promise 0%.
Many large successful companies today are natural or unnatural monopolies which rely on a data or network effects moat to generate a lot of free cash flow. And yet, they actually produce very few new ideas. That is something a few folks like Peter Thiel have been lamenting for over a decade now. Here is a very nice short clip of Thiel confronting Google’s chairman on the same issue:
The only way to get better at innovation is to actually do a lot more of things. And even then you will likely have many more failures than you will have successes. And your failures will have lost you tens to hundreds of millions of dollars before the breakout success(es) will return the first penny back. Now, if you are a CFO, this might sound like corporate lunancy. If you are a C-suite executive, this is hardly actionable advice. But in the absence of any action, your company is no different than the 30-year old zero coupon bond. Your company’s value is completely priced in. Sure, you might throw an earnings surprise here and there. But for all effective purposes, you will only be able to take share from an existing pie, not grow the pie.
The only actionable advice here is to set up your company, its functions and operations in such a way as to simply increase the frequency of experiments. Many failures will likely ensue but success will ultimately follow. The most innovative public company of the past decade, Amazon, has harnessed in on this insight. Bezos’ decision around decentralisation (outlined in a 2002 company-wide memo) was a key enabler here. But equally important is Bezos’ recognition that most such experiments will fail. He often brings this up at conferences and mentions this in his shareholder letters. That is a way to say to himself, his investors his and employees: Failure is okay, risk-seeking is encouraged here. And this mindset has helped Amazon become the behemoth that it is today.
Every time you are tempted to give in to science envy, think of the actual reality of our physical universe. Quantum physics has codified uncertainty at the very heart of our current understanding of nature. The very equation that social scientists and spreadsheet gurus have come to imitate is basically a sham. It is no different from a probability distribution. And if the physical world still can’t create an equation for a rule-following universe, you likely shouldn’t create one to evaluate business ideas.
To receive more posts like this, sign up for our newsletter!