Anatomy of a start-up’s failure

In what has to be one of the most costly bankruptcies in startup history, in May 2013 the Israeli electric car charge company known appealingly as Better Place tossed in the towel, having burned through more than $800 million of investor equity on virtually no revenue.

It’s easy to blame externalities such as lack of investor confidence, slow development of the market, or the time and cost involved in purchasing land and obtaining permits for battery swap stations, but the real causes of the failure were internal.  I worked as the chief technology officer of the Australian branch for two years until February 2012, so I saw much of the collapse unfold.  This account is a story about the parent company, not the wholly owned subsidiaries in Australia or Denmark.  The story is important not just as a historical reference but as a what-to-avoid lesson for future entrepreneurs and investors.

The charismatic founder, Shai Agassi, had an enviable track record in software development, having co-founded and sold TopTier Software to SAP in 2001 for $400 million.  He then worked as a senior executive at SAP till he left in 2007 to found Better Place, a company with the laudable mission to rid the world of its dependence on oil by creating a network of battery-swap stations to provide instant range extension to a new fleet of electric cars.

Mr Agassi knew, correctly, that in order for electric cars to become mainstream there needed to be a solution to “range anxiety”.  The 100 mile range of modern lithium-ion powered electric cars was enough for most people most of the time but restrictive enough that it would be occasionally inadequate, leaving drivers with no option but to limp home and plug in the car to charge overnight.

‘Carry a spare battery in the trunk’ is the instant alternative that most people think of, but electric car batteries weigh up to a thousand pounds or more, so that is not an option.

The Better Place solution was for you to drive your car into a battery swap station about the size of an automated car wash, inside which the driveway would open and a fully robotic system would remove your depleted battery from below and replace it with a charged battery in a couple of minutes.  Since Better Place owned the battery, you would not care whether the replacement battery you received was older or newer than the battery your drove in with, as long as it was fully charged and serviceable.

Of course, most of the time you would not deplete your battery during the course of a day so instead of diverting to a battery swap station you would plug in at home.  Not to an ordinary outlet, but to a special outlet owned, installed and managed by Better Place.

Great model, so what went wrong?  Many things.

Very few people can change a market overnight.

To start, management.  The members of Mr Agassi’s top management layer of vice-presidents were chosen for their loyalty rather than their domain expertise.  For example, Mr Agassi appointed his sister as the head of marketing, his younger brother who had no experience in the construction industry as the global head of battery swap station deployment, a software colleague in charge of the automobile alliances, another software colleague in charge of managing the electric utilities alliances and a medical doctor colleague in charge of all hardware and software development.  The truly sad thing is that while in the early days there were domain experts keen to join the senior ranks of Better Place, they were kept out of the inner circle.

Next was over-estimation of the company’s ability to create disruptive market change.  Very few people can change a market overnight.  Steve Jobs did it several times, but besides the brilliance of his vision and his talent for implementation he had the advantage that his products were individually much cheaper than a car, which meant that the consumers and the manufacturing supply chain could respond quickly.

In its initial planning, Better Place assumed that it would have three million battery swappable electric car customers on the road within five years of its founding, and it built its product development plans around this extraordinarily optimistic assumption.  To meet this requirement, Better Place overreached and built for scale in a single step.

With new technologies, building for scale in a single step is simply not possible.  One has to go through a series of proof of concept, pilot and early releases in order to take into account market feedback to enable the progressive improvement of the product.  To make matters worse, the early designs of the electric chargers, the in-vehicle navigation and energy management systems, and the battery swap stations, were put into mass production under contract through offshore suppliers before the designs were field tested.  Without field testing, these products were inevitably overdesigned, overly complex and unreliable, but even as the problems unfolded there was no easy way to fix them because volume manufacturing had already been commissioned.

In his anxiety to reach scale in a short period, Mr Agassi instructed the development team to use billing systems from Amdocs and enterprise management systems from SAP that were fundamentally unsuitable for a startup company.  The companies, Amdocs and SAP, and their products are superb at what they do, but their only attraction for a startup electric car charge company like Better Place was that their CEOs were acquaintances of Mr Agassi’s and that their involvement would look good to investors.  To make Amdocs and SAP work in ways they were never intended to operate, the development team hired consultants from IBM.  Highly capable, but too expensive for a startup.

To operate its customer relationship management and electricity supply software systems, Better Place built its own server farm at a data center in Spain, at a time when most other companies were shifting to cloud-based solutions.

To manage its project development schedule, the development team was instructed to use sophisticated project management tools, however the projects were too exploratory and variable for the tools to be of any assistance.  To the contrary, they were a burden and a distraction.

Instead of testing prototypes in the market to learn what customers needed, the company employed teams of process engineers whose job was to conjecture what the market would require and design to the nth degree the details of the unproven products.  This kind of approach might be required for a new aeroplane but it is expensive, slow and totally inappropriate for the design of a first-of-a-kind consumer product.

Drivers simply want to charge their cars whenever and wherever they happen to be.

The common problem was that instead of challenging his development team to create the best solutions for each task, Mr Agassi told them how to go about their business.

A simple belief that there was only one solution to range anxiety was another problem.  Despite Mr Agassi’s early success with Renault no other major car manufacturer bought the vision.  To the contrary, when the car company executives were told by Mr Agassi that they didn’t understand the future despite their incredible track records of success and deep knowledge of the markets, their response to Better Place in many cases was hostile.

The Better Place arrogance manifested itself other ways.  There was a belief that customers would want to be members of the Better Place charge network and that they would value a special corporate relationship.   The reality, of course, is that drivers simply want to charge their cars whenever and wherever they happen to be, and they have no loyalty nor desire to invest time into their relationship with the electricity or gasoline supply company.

Better Place also assumed that customers would accept being restricted to a single supplier of charging electricity at home rather than change provider plans whenever they liked.

Finally, there was the assumption that all customers would purchase on the basis of lowest cost of ownership rather than functionality or other intangibles such as the reputation of the vehicles, the fit-out options and the styling.  The reality is that customers saw beyond cost.  The battery exchangeable vehicle that Renault built was brought to market too quickly, with significant limitations such as a tiny trunk.

Then there was simple ignorance on many fronts.  First and foremost was the lack of appreciation that car companies would not give up control of the battery design and configuration.  In an electric car, the battery is the most expensive single assembly and it is a critical part of the drive train.  It determines the range and acceleration of the vehicle, and its location in the vehicle has a huge bearing on handling.  Further, its construction and attachment to the vehicle determines the vehicle’s safety under crash conditions.  All of these features are part of the reputation of the vehicle that distinguishes it in the market place.  This meant that if twenty manufacturers had indeed agreed to participate in the battery swap model they would have insisted on twenty different battery designs.  This alone would have been a showstopper, especially in the early days because Better Place’s battery swap stations were limited to a single model of car and battery.

Banks and other lenders are wary about lending money to startups and, if they do, they do so at much higher interest rates.

Another simplistic assumption was that between battery swaps car drivers would plug into charging outlets whenever they were parked, be it at home, at work or in the shopping center car park.  But that is not normal human behavior.  Most of us postpone filling up with gasoline as long as possible.  It would not be natural behavior to pull out the cord and plug in several times per day.  And if you did plug in, would you accept anything less than a full rate of charge?  Better Place’s model assumed that you would accept reduced or even zero rates of charge so that Better Place could preferentially supply electricity to drivers more needy than you.

Then there was the lack of appreciation of the financial astuteness of paying customers.  Potential customers already knew that electricity was much cheaper per mile than gasoline.  However, the ultimate cost of electricity to the customer via Better Place was higher because of the financial impact of Better Place owning the batteries.  The reality is that the cost of capital to a startup is much higher than the cost of financing for a vehicle owner.  Banks love car owners, because most people would rather cut back on going out to dinner with friends than have their car repossessed.  In contrast, banks and other lenders are wary about lending money to startups and if they do, they do so at much higher interest rates.  So inevitably, Better Place’s ownership of the batteries placed a huge premium on the monthly subscription price.  The electricity itself, per mile driven, was most definitely cheaper than gasoline, but by financially passing through Better Place’s hands rather than being supplied directly from the electricity company, the cost to the customer was higher.  Instead of tackling these problems by revising the model, the company employed numerous business analysts to develop scenarios under which the pain might somehow be tolerable.

Because the financial model suffered from high built-in costs, Better Place tried to add value by re-purposing batteries that were no longer fit for vehicle use to be used as storage batteries for renewable energy supplies such as solar and wind.  Trouble is, the economics of using batteries to smooth intermittent electricity supplies are poor, even if the batteries are purchased at a low, second-hand price.

Then there was over attention to brand rather than product.  It started with impressing investors and potential corporate partners by building a network control center that would have looked striking in a Hollywood studio reconstruction of a national nuclear electricity control room.  At the other end of the scale, the branding agenda manifested itself with over attention to colors, fonts and language rather than attention to the functionality of the products that the customers would interact with.

Talking about brand, Better Place missed out on the opportunity to promote any carbon dioxide emissions credentials.  Electric cars are indeed excellent in this regard, but the electricity overheads running the battery swap stations and the losses introduced by charging cars at less than their maximum charge rate threatened to eat up the carbon dioxide emissions benefits.

Finally, there is the model itself.  When I joined the company I was enamored by the battery-swap model as I had seen it portrayed by the evangelist himself.  The trouble is, it is dangerous to assume that there is only one worthwhile solution to a technology problem.  Human ingenuity is so powerful a force that alternative solutions will always emerge.

At least three alternative solutions to the battery swap model have come to market since Better Place was founded, none of which require the construction of a massively expensive network of battery swap stations.

The first alternative is the Tesla Motors approach, in which Tesla simply installs higher-capacity batteries.  The Tesla Model S has a 250 mile range, which is so large that it would be rare indeed for a Tesla driver to need to go further in a single day.   Since it is easy to charge overnight, at home, the range anxiety problem is effectively solved for Tesla drivers.  Yes, the car is expensive but sales are nevertheless booming because it is a luxury car, quite beautiful and with stunning performance.

The second alternative is the BMW approach pioneered in the i3 electric car, in which a range extender unit can be installed as an option.  The range extender is a gasoline powered generator that can continuously top up the battery while the car is driving.  It only cuts in on those days where the car is driven more than usual.  Because it is never called upon to drive the wheels, it operates at constant speed and load, thus the design is easily optimized for efficiency, low cost and small size.

The third alternative is to install quick-charge stations throughout the road network.  These are publicly available charge units that have sufficient power to recharge an electric car in an hour or less.  Not something you would want to do on a regular basis but as an alternative to running out it is pretty good.  And compared to battery swap stations quick charging is a cheap network to deploy.  The Nissan LEAF electric car that relies on overnight charging for ordinary driving and quick charging when necessary is already selling extremely well in the US, Japan and Europe.

It would be hard not to blame the investors and the board.

Battery swap survives as a fourth model.  It was given a boost (about a month after the Better Place bankruptcy filing) by the announcement by Tesla Motors that they released a battery-swap version of the Model S.  Compared to the Better Place battery swap network the proposed Tesla battery swap station roll out has the advantage that the Tesla Model S vehicles can drive a long way between swaps, thereby minimizing the number of expensive swap stations that must be deployed along the highways to make interstate driving practical.  However, interest by their clients has proven to be minimal and the program might not continue.

Many of the decisions that Better Place made, such as developing a solution that would manage the timing and level of charging of every car plugged into the Better Place network, were based on avoiding a doomsday scenario that was the concern not only of Better Place but also of governments and electricity distributors at the time.  The scenario is that on a hot day, massive numbers of workers would come home at 6 pm, plug in their electric cars, go inside, turn on their air conditioning units and televisions, and collectively cause a blackout.  The reality is that this will not happen, firstly because people don’t arrive home all at the same time, and secondly, as a result of the ever improving range of electric cars, drivers won’t plug in every time they arrive home.

Whose fault is it that so much money was spent so fast for no return?  It would be hard not to blame the investors and the Board.

To convert his early vision to reality Mr Agassi raised very large sums of money on sky-high valuations from well-established venture capitalists, investment banks and corporations.  They were so enthralled with his evangelical message that they based their due diligence on Mr Agassi’s optimistic assumptions rather than the more conservative assumptions from the mainstream car industry.  That’s not a smart way to make an investment.

The lofty valuations were themselves a problem for the company, a millstone around management’s neck that prevented the company from retreating for a while to absorb the implications of external developments and possibly resurface with a more modest business plan.

After a few years of operation, the management problems became obvious to many of the senior staff, and to executives of the subsidiary companies, but the Board didn’t ask questions despite the fact that they were all appointed to the Board to protect the interests of the investors.  In addition, the problems with the model itself were increasingly obvious, but the Board did nothing.  For more than a year towards the end the company did not have a full-time CFO.  Whatever the reason for this, it should have set the alarm bells ringing.  I can only explain the Board’s acceptance of the flawed management and model by them being mesmerized by the charismatic Shai Agassi.  I have no doubt that Mr Agassi himself believed the model and was confident in his own management abilities, but Boards exist to question, to guide and to provide the wisdom that might be missing in an enthusiastic and overly confident management team.

Great leaders have vision and can convey it.  As CEO and founder of Better Place Mr Agassi had the vision and the ability to inspire.  But great leaders also have operational talent and willingness to modify their vision to meet the emerging market realities and competition.  This was sorely lacking.

Please login to favourite this article.