Why Better Place Failed:A Lesson for Partner Networks
Why Better Place failed despite $850 million: expansion replaced evidence. The same sequence error can sink your partner network. Here is what to do instead.
Most startup obituaries blame a shortage of something — demand, product, capital. Understanding why Better Place failed is harder than usual, because none of the usual suspects apply. The technology did what it promised: pull your electric car into a station, a robot exchanges the depleted battery for a charged one in roughly ninety seconds, and you drive off. The funding was almost without precedent: $850 million, which at the time ranked as the fifth-largest venture round ever raised. And yet by 2013, the company was bankrupt.
The vision had launched six years earlier, when founder Shai Agassi took the stage at the World Economic Forum in Davos in 2007 and described a global network of battery-swapping stations that would end oil dependency — no charging waits, no range anxiety, no fossil fuels.
What destroyed the company was neither the market nor the machine. It was the order of operations. Agassi scaled a model he had never finished proving, and no amount of capital can compensate for a broken sequence.
Ron Adner, the ecosystem strategist who examined why Better Place failed in depth, treats it as one of the most instructive failures in modern business. If you run a consultancy, agency, coaching practice, or training company and you are building a partner network, this story is about you — just with fewer zeros.
Why Better Place Failed: Expansion Is Not Evidence
An Impressive Map Concealing an Unproven Model
While the first deployments in Israel and Denmark were still finding their feet, Better Place was already opening offices in Australia, Japan, Hawaii, the San Francisco Bay Area, the Netherlands, and China. Agreements were signed with governments and automakers across several continents. Each announcement produced fresh headlines and fresh investor excitement.
From the outside, this looked like momentum. Markets entered. Partnerships inked. New pins on the world map every quarter. The story kept getting bigger.
Look beneath the map, though, and nothing had actually been demonstrated. The swap stations in Israel were costly to construct and slower to roll out than planned. The Renault partnership had yielded exactly one compatible model — the Fluence Z.E. — and demand for it was weak. Even the basic offer was muddled. Was the customer buying a car, an energy contract, or a subscription? You got a different answer depending on the country and the presentation.
"Too much of the time was lost to the distraction of global expansion."
— Ron Adner, on Better Place
Adner's diagnosis cuts to the bone. Every dollar that funded an Australian office was a dollar withheld from making the Israeli pilot actually function. Every executive hour spent courting a Chinese partnership was an hour not spent untangling the customer experience in the single market where the system was live.
Better Place was not starved of opportunity. It was flooded with it — and the flood is what drowned the company.
Why Better Place Failed: The Beachheads Were Already Chosen
Israel and Denmark Could Have Proven Everything
The bitter irony is that Better Place never had a market-selection problem. It held two of the most favorable pilot markets a startup could hope for, and it needed only one of them.
Israel measured just 470 kilometers north to south, with dense population centers, punishingly high fuel prices, a government actively backing alternative energy, and Renault committed to producing compatible vehicles. A few hundred stations would have blanketed the entire country with swap coverage. National infrastructure, at startup scale.
Denmark offered a similarly compact geography, deep environmental consciousness, and tax incentives that made electric vehicles dramatically cheaper than their combustion counterparts — a population already culturally primed for the product.
The sequence writes itself in hindsight. Saturate Israel. Show that battery swapping holds up under real daily use. Validate the station economics. Capture the case studies and codify the playbook. Then carry the proven playbook to Denmark, and from there outward, one evidenced market at a time.
Adner is explicit on this point: in his analysis, "The markets in Israel and Denmark would have allowed Better Place to reach sustainable scale." The atomic network — the smallest self-sustaining version of the ecosystem — was sitting right there, waiting to be completed.
Agassi chose everywhere instead of somewhere. That single choice set the clock on the bankruptcy.
Why Better Place Failed: The Sequence Error in Service Firms
Eighty Certified Partners and Nothing Proven
You do not need nine figures of venture capital to re-run the Better Place script. A partner program, a methodology that demos well, and the very human habit of mistaking expansion for validation will do the job at any budget.
Watch how it plays out in the service world. A methodology founder certifies a first cohort of twenty partners. The training lands well — energized participants, strong evaluations, a framework that feels robust in the room. Before that first cohort has delivered even five real client engagements, Cohort 2 opens in a new market. Cohort 3 follows in another geography. Twelve months later there are eighty certified partners spread across four countries, and the conference slide shows a map with pins in twelve cities.
The LinkedIn feed says success. The dashboard says something else entirely:
- 35% partner activity. Two out of three certified partners have never delivered an engagement.
- Three case studies total — every one of them from the original cohort, none from the new markets.
- A 4% referral rate. Nearly all new business still flows from the founder's own marketing.
- Sliding partner satisfaction. The people who were enthusiastic at certification are now frustrated that the promised client leads never arrived.
- $22,000 of revenue per partner — nowhere near enough to sustain a practice.
Read those numbers carefully, because they are not describing a scaling problem. They describe a proof problem. The atomic network was never completed in market one, so the expansion simply photocopied an unvalidated model into three more geographies.
This founder did not torch $850 million. They torched something with no replacement market: the patience and goodwill of the partners who signed up first.
Run the Sequence in the Right Order
Four Moves Better Place Skipped — and You Shouldn't
Adner's remedy is not complicated, and it transfers directly from national charging networks to consulting partner programs:
1. Commit to a single proving ground. Israel had every structural advantage — compact geography, state support, a committed automaker, cultural readiness. It deserved to be the only market for the first two years. For a service ecosystem, the equivalent is one segment, one geography, one cohort.
2. Close the full loop. Enough stations to cover the country. The Renault Fluence selling at a steady clip. Drivers genuinely treating swap stations as their primary way to "refuel." Unit economics demonstrated, with each station covering its operating costs inside a defined window. Until the loop closes end to end, nothing is proven.
3. Convert the loop into evidence. Swaps per day. Consumer satisfaction scores. Cost per swap against the cost of gasoline. Station payback period. Captured and documented, this becomes the most persuasive expansion asset that exists — not projections on a slide, but a live system running in a real market.
4. Export the playbook, not the ambition. Carry the Israeli evidence — operational, financial, and consumer data — into Denmark. Adapt it to local conditions. Prove it a second time. Enter market three holding two proof points instead of zero.
Notice what changes at every step: the expansion decision rests on evidence rather than enthusiasm. The conversation with investors, governments, and partners stops being "we believe this will work" and becomes "it already works in two markets — here are the numbers."
For your ecosystem, the translation is one-to-one. One segment. A complete delivery cycle. Documented proof. Expansion driven by what the evidence licenses, not by what the vision craves.
The Asset That Doesn't Refinance
Capital Was Lost. Trust Was Destroyed.
The headline failure was financial — Better Place simply ran out of money. The structural failure ran deeper, because what the collapse consumed could not be raised in another round.
Early employees gave years of their careers to a mission that evaporated. Israeli drivers who had bought the Fluence Z.E. were left holding cars whose entire refueling infrastructure ceased to exist. Government partners who had spent regulatory capital and public money were embarrassed. Renault absorbed a write-down and walked away permanently soured on battery swapping as a concept.
Scale that down to a service ecosystem and the same cascade appears. Partners who paid for certification feel cheated when the lead flow they were promised never materializes. Early clients who received a half-refined methodology become its loudest critics. The internal champions who sold the program to their organizations watch their credibility drain away when delivery falls short of the pitch.
Money is a renewable resource. Trust is not. A partner who exits after a bad experience does more than leave — they become the cautionary story told to every prospective partner who asks around. An executive who sat through a mediocre engagement does more than churn — they actively steer their peers away from your methodology.
Trust compounds in whichever direction you feed it. Strong early experiences compound into referrals and advocacy; weak ones compound into warnings and avoidance. Premature expansion loads the system with negative experiences before you have engineered the positive ones — and the compounding does the rest.
That is why Better Place failed: Agassi commanded $850 million and two near-perfect launch markets, and the consequences of scaling before proving caught him anyway.
You are working with less capital and fewer advantages than he had. Which makes the discipline non-negotiable: close the loop, capture the proof, and only then scale the thing you have proven.