Why Battery Startups Fail: Developing Technology vs. Shipping Products

Authors: Stephen Fratamico, Bud Collins

Over the past 25 years, battery technology has improved dramatically, yet battery startups have struggled. At Catalus Capital, we have unique and extensive insight into the industry, resulting in an informed opinion on what young battery companies must do to succeed.

Over decades, energy density has increased, costs have fallen, and power capabilities have climbed. For illustration, pack prices reached a record low of $108 per kilowatt-hour in 2025, 93% below where they stood in 2010 in real terms. [1] Yet, unlike many other high-tech industries, most of these advances have come from large, established organizations, not startups. In an unusually persistent degree of concentration, about 80% of battery patent families filed between 2000 and 2018 came from a limited group of very large companies, even as battery patenting grew 14% annually (four times the 3.5% average across technologies). [2] Meanwhile, battery startups have consistently failed to turn working technology into working businesses.

When these startups fail, technical reasons are usually given: the chemistry did not perform as expected, cycle life was too short, or costs were too high. But after examining hundreds of battery startups, we identified a different pattern.

In many battery startups, business development focuses on distant future markets while R&D is fixated on achieving step-change improvements for a next-generation product. The result is a familiar pattern: companies pursue large customers that are realistically four years away from actually buying—while devoting a large amount of R&D resources to pursuing breakthroughs—instead of delivering real products to smaller customers, gathering real-world data, and learning quickly.

As part of our work at Catalus Capital, a battery technology-focused investment firm, we have seen and done diligence on hundreds of battery startups over the past seven years. From this work, a pattern appears consistently: the dominant reason promising battery startups fail is the separation of technology development from commercial execution.

Historical Performance

To explore this phenomenon further, we assembled a cohort of the 14 most prominent innovative battery companies from the first cleantech wave (roughly 2005 to 2015) and analyzed their outcomes.

As of today, twelve of the fourteen have not reached commercialization. [3] Seven entered bankruptcy, shut down, or underwent an asset-level wind-down; three were acquired before reaching profitability; and two remain active more than a decade later with no publicly verified sustained production. Only two achieved meaningful product adoption, yet neither has publicly demonstrated both durable commercialization and profitability.

A second, partially overlapping cohort of the eight innovative battery technology companies that completed SPAC mergers between January 2020 and December 2022 produced a similar result. Three have yet to meaningfully commercialize their original battery thesis or have abandoned it entirely. The remaining five have generated product revenue, but none were profitable as of June 2026. Out of this cohort, in FY2025, Enovix, Amprius, and SES AI achieved positive gross margins of approximately 19%, 11%, and 54%, respectively, but remain loss-making, while Eos Energy and ESS Tech ran highly negative gross margins. Notably, large amounts of revenue at companies in this cohort do not indicate full commercial adoption: Solid Power’s 2025 10-K says revenue to date ($77.7M since inception and $1.1M in 2025) has primarily come from R&D licensing agreements, a line-installation agreement with SK ON, and government contracts. Solid Power’s 10-K goes on to describe the company as “a research and development company with no commercial operations.” [5] Across the cohort, none has yet converted its battery technology thesis into both durable commercialization and profitability.

While these cohorts are small cross-sections, they have raised over $15.8B without demonstrating profitability, and they underscore the same patterns we have seen at Catalus after seven years of doing due diligence on innovative battery startups.

Technical Success Is Not the Bar

The lack of success we are illustrating may contradict the press releases and media statements made by some battery companies, so it is useful to distinguish among the levels of success for a battery startup. For example, several companies in the above cohorts produced functioning batteries, ran pilot plants, supplied product to customers, and deployed product into commercial systems.

For a battery technology company, success can be grouped into four categories:

• Technical success: the chemistry or architecture works under laboratory or pilot conditions.
• Initial commercialization: the company ships samples, small batches, or early commercial units.
• Durable commercialization: the product achieves recurring customer adoption and continuing production.
• Economically sustainable operations: the company can make and sell the product profitably.

Most of the companies analyzed above have reached the first stage, several reached the second, very few reached the third, but none have publicly demonstrated the fourth.

The Illusion of Progress: Attention from Marquee Companies Is Not Traction

To understand why battery startups fail, it is critical to understand their rocky relationship with prestigious commercial partners. Battery startups almost universally gravitate toward attempting to sell product to marquee companies—electric vehicle manufacturers (aka automotive OEMs) being the best example of this. On the surface, this makes sense. These companies represent the largest market, the highest volumes, and the most obvious path to a massive exit, a primary objective for many venture investors.

However, engagement with a large company is often misleading. For example, automotive OEMs are willing to evaluate nearly anything a startup can provide: anodes, cathodes, separators, electrolytes, or cells. However, the gap between getting an OEM to accept samples and actually becoming a real customer of the startup is vast. Hundreds of companies are competing for the attention of these firms, and evaluation programs are relatively inexpensive for them.

The asymmetry is structural. According to the U.S. Department of Energy, a new battery or battery component supplier “must generally build a full-scale facility to undergo qualification, incurring significant upfront capital expense (on the order of hundreds of millions or billions of dollars), but may not see revenue for as much as 12 to 18 months while the qualification process is underway.” [7] The quantum of capital required to sell to large companies is immense, so it is not logical for a startup to view it as their beachhead market.

Startups often do not understand this asymmetry, and they frequently mistake program activity for progress. Meetings, technical exchanges, and pilot projects often look like traction. Some large companies further propagate this misunderstanding by providing startups with aspirational sales and volume targets without appropriate caveats. Investors often reinforce this illusion as well by seeking investment candidates with engagements with blue-chip firms. In reality, these relationships rarely lead to meaningful commercial outcomes.

That illusion is further perpetuated by the conflation of technology readiness level (TRL) across applications. TRL describes how mature a new technology is. Startups and investors often look at TRL levels to assess the investability of new technologies. More often than not, TRL is treated as a single value for a technology; however, it is more useful to use multiple TRL values for a technology across different applications. The same battery that counts as an advanced prototype at TRL 7 in an aerospace program sits at TRL 4 (laboratory scale) when it is measured against an automotive one. [8]

It is critical that a startup understands the barriers and considerations when attempting to sell to a large commercial partner. Aside from occasional payments or non-recurring engineering revenue, years can pass before a startup realizes it has made little progress toward being included in a real product.

Why Laboratory Testing Is a Poor Substitute for the Real World

Battery startups must be data-driven because of the scientific nature of their products. However, many companies rely almost entirely on laboratory testing data, which can be misleading.

Early-stage startups often use very small sample sizes, sometimes as few as five to ten cells per batch. Cell testing equipment is expensive, and long-duration tests take months. At the same time, early manufacturing conditions frequently introduce significant noise into the data.

Even with the best-designed testing protocols, laboratory environments cannot replicate the real world. Actual battery deployments experience complex combinations of vibration, misuse, temperature gradients, partial charging, long dwell times, and other environmental stresses. These conditions are difficult or impossible to fully reproduce in a laboratory.

As a result, unknown failure modes inevitably slip through testing programs. In one striking example, we encountered a startup that assiduously followed testing conditions specified by a commercial partner. When the startup learned how the batteries would actually be used in the product and modified their testing protocol accordingly, every single battery failed catastrophically.

A 2025 study quantified this gap in automotive applications. Researchers cycled 92 commercial cells over 24 months against 47 discharge profiles spanning constant-current baselines, periodic duty cycles, synthetic profiles built from field data, and real electric-vehicle field data. They found that industry-standard constant-current cycling underestimated cell lifetime by as much as 38% compared to realistic dynamic cycling protocols. [9] In 2018, when Idaho National Laboratory cycled two Nissan LEAF packs alongside individual cells from the same vehicle on an equivalent protocol, the packs lost capacity roughly twice as fast as the cells, and a fast-charging penalty of under 1% at the cell level grew to 5 percentage points at the pack level. The authors concluded that pack-level aging cannot be directly extrapolated from cell evaluation. [10]

A battery startup must understand that laboratory testing is an inadequate replacement for field data.

The Importance of Yield

The step that separates a working cell from a working company is manufacturing. Taking a cell that works in the lab and figuring out how to reliably manufacture it at scale is a herculean task with unforgiving economics.

During early ramp-up, scrap rates can reach 70 to 80%. [12] Rates of 15 to 30% are common in the first years of battery production, and often hover around 10% after five years. [11] The International Energy Agency states that competing in today’s battery market “requires reaching average production yields exceeding 90%,” and that for most facilities “it can take more than 5 years from the start of operations to reach levels close to nominal output.” [13] Figure 1 plots these ranges against both thresholds.

Tying into the previous point, attempting to sell into a large program like automotive further compounds the problem. In automotive, the governing metric is defective parts per million, and pack reliability multiplies across every cell in the pack, of which there may be hundreds or thousands. As three battery-industry veterans put it, “a defect rate of even 0.1% is much too high.” [14] Attempting to scale a new technology to automotive gigafactory scale before improving yield on smaller manufacturing lines inevitably leads to higher scrap rates. A startup would be best served to forgo large-volume manufacturing until they are able to improve yield on a smaller scale.

The Forgotten Middle: Product 1 Pays for Product 2

Another one of the most common strategic mistakes battery startups make is delaying commercialization while waiting for a next-generation product.

Many startups already possess a Product 1 that could deliver strong value in niche markets today. Instead of deploying it, they focus on developing Product 2, a hypothetical advance promising a 50% improvement over the current state of the art. This is almost always a mistake. In practice, Product 2 cannot reach the market before Product 1. Meanwhile, while focusing on Product 2, the startup gains no field data, no operational experience, and no revenue.

If EV manufacturers are unlikely to adopt a company’s Product 1, and if real-world applications will inevitably reveal hidden issues, what should startups do?

Startups should deploy early products in smaller, less glamorous markets where their technology provides meaningful value, where there is higher tolerance for early failures, and where there are much higher margins. These markets can provide lower-cost externalized testing, thousands of real-world operational datasets that can validate products for larger-volume customers, and modest revenue with positive contribution margins through higher prices that can absorb lower yield rates.

Most importantly, these deployments compress the learning loop. By shipping Product 1, a company learns how the product actually behaves in the field, and those lessons help improve it further to inform the design of Product 2. Skipping Product 1 is not a shortcut. It is often a fatal mistake.

China’s Advantage Is Not Better Batteries

The dominance of Chinese battery manufacturers is sometimes attributed to superior technology. In reality, a large portion of their advantage lies in development philosophy and manufacturing execution. According to the IEA, “over a decade of mass production experience enables China to build and efficiently operate highly automated factories: manufacturing efficiency alone accounts for over 40% of the battery production cost gap between China and Europe.” [13]

Companies across China, Japan, and Korea frequently adopt an incremental approach to innovation. Instead of waiting for large technological leaps, they release products earlier and improve them continuously. This approach (best known in the West through the Japanese concept of kaizen) prioritizes learning through deployment. Products enter the market sooner. Mistakes are discovered earlier. Improvements happen faster.

The result is visible in the market. Seven Chinese manufacturers accounted for 72.6% of global EV battery usage in the first five months of 2026, and seven of the ten largest producers worldwide were Chinese. [15] When attempting to navigate the battery industry as a startup, reorienting development philosophy to align with Eastern incumbents will likely pay dividends.

Conclusion: Batteries Don’t Fail, Strategies Fail

Battery startups continue to repeat the same strategic mistakes as previous generations. The solution is not better chemistry or better engineering, nor is it the perfect technology developed in isolation with commitments from prestigious commercial partners. It is better integration of technology development and commercial strategy so startups can ship products early, learn quickly, and improve continuously. Startups that effectively implement this strategy will tightly link R&D and market deployment to enter markets earlier, learn from real-world product failures, and iterate rapidly.

This article is published for informational purposes only. It does not constitute investment advice, a recommendation, or an offer or solicitation to buy or sell any security. Companies are named solely to illustrate patterns in battery technology commercialization; naming a company is not a recommendation with respect to its securities. Views are those of the author as of the date of publication and are subject to change. Information has been obtained from public sources believed reliable but has not been independently verified in all cases. The author, his firm, and their affiliates, funds, and personnel may hold positions in, and may buy or sell, securities of companies discussed, and may take positions inconsistent with the views expressed.

References

1. BloombergNEF, 2025 Lithium-Ion Battery Price Survey, 9 December 2025.
2. European Patent Office and International Energy Agency, “Innovation in batteries and electricity storage: a global analysis based on patent data,” September 2020, sections 4.2 and 5.1–5.2. The 80% applicant-concentration figure and Figure 5.4 appear at pp. 55–56; the 14% and 3.5% growth rates appear in the Executive Summary p. 5 and section 4.2.
3. Constructed cohort of 14 prominent battery technology companies that received material institutional financing between approximately 2005 and 2015, assessed against public bankruptcy filings, acquisition announcements, SEC filings, and company disclosures, with status as of August 2026.
4. Cohort of 14 prominent battery technology companies that received material institutional financing between approximately 2005 and 2015. Capital raised compiled from SEC filings (Forms D, S-1, 10-K, 10-Q, and 8-K), U.S. Bankruptcy Court dockets (A123 Systems, D. Del. Case No. 12-12859; Xtreme Power, W.D. Tex. Case No. 14-10096; Aquion Energy, D. Del. Case No. 17-10500; Ambri, D. Del. Case No. 24-10952), USAspending.gov federal award records, and company announcements. Government funding is counted only where received, not where awarded or conditionally committed.
5. Constructed from U.S. Securities and Exchange Commission filings for the eight innovative battery technology companies completing SPAC mergers between January 2020 and December 2022. The screen covers companies commercializing a novel cell chemistry or architecture and excludes conventional cell manufacturers, pack integrators, recyclers, and storage software and integration companies. Revenue and gross profit from each registrant’s own FY2025 annual reports; company status as of August 2026.
6. Cohort of the eight battery technology companies completing SPAC mergers between January 2020 and December 2022, compiled from each registrant’s SEC filings (Forms 8-K, S-4, 10-K, and 10-Q), comprising pre-merger private capital, SPAC trust proceeds net of redemptions, PIPE proceeds, post-listing equity and debt drawn, and government funding received. Status and figures as of August 2026.
7. U.S. Department of Energy, 2021–2024 Four-Year Review of Supply Chains for the Advanced Batteries Sector, December 2024, p. 102.
8. J. T. Frith, M. J. Lacey, and U. Ulissi, “A non-academic perspective on the future of lithium-based batteries,” Nature Communications 14: 420, 2023.
9. A. Geslin, L. Xu, D. Ganapathi, K. Moy, W. C. Chueh, and S. Onori, “Dynamic cycling enhances battery lifetime,” Nature Energy 10: 172–180, 2025.
10. T. R. Tanim, M. G. Shirk, R. L. Bewley, E. J. Dufek, and B. Y. Liaw, “Fast charge implications: Pack and cell analysis and comparison,” Journal of Power Sources 381: 56–65, 2018. Idaho National Laboratory, INL/JOU-17-43278.
11. C. Dahmen, F. Degen, M. C. Eckstein, K. B. Pouls, and H. Walter, Fraunhofer Research Institution for Battery Cell Production FFB, with R. Ludwigs and A. Scheibe, PEM RWTH Aachen, “Mastering Ramp-up of Battery Production,” 17 October 2024, pp. 4–6.
12. McKinsey Center for Future Mobility, “Battery 2035: Building new advantages,” 1 January 2026.
13. International Energy Agency, Global EV Outlook 2026, pp. 138, 144, 150.
14. P. M. Attia, E. Moch, and P. K. Herring, “Challenges and opportunities for high-quality battery production at scale,” Nature Communications 16: 611, 2025.
15. SNE Research, global EV battery usage January–May 2026, released 3 July 2026.