The Essential Limits of Accelerators
What accelerators are good for, and what generators can do better
TL;DR: An analysis of the shortcomings of our current innovation ecosystem in accelerating essential ventures for the public purpose, and a proposal for a new structured innovation support model to generate solutions to challenges facing society. This essay breaks down the assumptions underlying startup accelerators, and provides a high-level step-by-step approach to a program model called the generator to focus our communities’ creativity on developing and rapidly validating new approaches to the challenges we face.
I love accelerators. In 2005, I founded an organization called PresenTense with the mission to exponentially multiply social innovation in my community. The natural outgrowth of that effort was the launch of an accelerator program in Jerusalem, Israel, in 2007. I worked for Aharon Horwitz’s startup at the time, and we thought to ourselves: wouldn’t it be great to bring together social entrepreneurs and techies to build technologies with social goals at heart?
Back then there were two accelerator programs we knew of — YCombinator, and Techstars. They ran essentially the same model: they invested ramen noodle money in aspiring entrepreneurial teams (between $8,000-$18,000) over 3 months, provided a place to meet and work, and in return received an 8% share of each company. The curriculum was minimal; connections were king. The idea was that the young, hungry founders would incorporate a company, build a proof-of-concept, and use the network the accelerator provided to raise their first investment. The reputation boost alone accelerated their growth, and the food and space didn’t hurt.
After reaching out to the teams running those two accelerator programs, we were told they did not believe in the idea of entrepreneurship for social (non-financial) aims. So, we read everything we could about how they worked, and built our own accelerator network that, by 2012, seeded and managed 11 franchise accelerator programs running concurrently around the world. By this writing in 2022, over 1500 social enterprises — nearly half for-profit, half not — have been launched by entrepreneurs who were PresenTense fellows.
The ventures launched by PresenTense fellows have accomplished wonderful things; each sparked a warm, healing light within the community and helped countless individuals. Nevertheless, I have come to conclude that we were less successful at affecting long-term change for the broader trends we sought to address. I believe systemic drivers have further deteriorated our present, setting us up for an even bleaker future. It became clear that while accelerator programs may support social innovators to address issues they have a burning motivation to solve, these programs are not well suited to affect the systemic changes we need if we are to avoid the worst of the polycrisis: climate change, deepening economic disparity, and a degrading democratic culture.
Accelerators are means to support entrepreneurs to achieve their aspirations, whereas a Generator is a structured means to direct innovation towards missions that address publicly defined challenges
This essay will focus on a new model I’ve been developing to drive systems change by generating the conditions for essential entrepreneurship. It is important for me to note that the generator program model I am putting forth here is not a replacement for an Accelerator, but a different model for different aims. Accelerators are a means to support entrepreneurs to achieve their aspirations, whereas a generator is a structured means to direct innovation towards missions that address publicly defined challenges, in which we recently seem to be drowning.
Accelerating companies, generating essential solutions
Before determining the best possible program to support innovation, we should first make sure we’re focused on the sort of innovation we’re seeking to support. Accelerators are optimized for instrumental innovation, whereas generators are optimized for essential innovation. The difference between instrumental innovation and essential innovation lies in their definitions of success. Instrumental innovations work to further a goal that is separate and distinct from the outputs of the company: the success of Google is not measured based on the number of search results it returns, just like the success of Uber isn’t measured based on the number of rides it carries. Both are determined to be successes or failures based on the amount of money they return to investors.
Startups are similar. Under our current innovation ecosystem, a company offering a wonderful innovation in cancer detection will go bankrupt if it does not promise to provide above-market returns to investors. That is why the press and founding teams celebrate how much money is raised in a round, why ventures are defined by their fundraising success (‘Seed Stage,’ ‘Series A Startup,’ etc) as opposed to their actual technological and essential accomplishments. It is also why our innovation ecosystem cares so much about the ‘exit’ (when an instrumental venture achieves its purpose) as opposed to the integration of the product or service offering into everyday life (which makes an essential difference).
Yet, we all know our lives depend on essential services: food to eat, a roof over our heads, someone to heal us when we’re sick. Essential services require innovation too, and the success of essential innovations should be measured by their promise being met: lives saved, children educated, water transported, calories farmed, food delivered. Whether or not these innovations generate 10 or 100 times a return on financial investment is orthogonal to their essential impact: either the essential innovation delivers the public value it intends to, or it is a failure.
We all instinctively understand the difference between instrumental and essential pursuits, and physically recoil when an essential industry pursues solely instrumental aims: no amount of financial return makes the mass marketing of opioids justified, no matter how much money it returns to shareholders. Nothing justifies adding plastics to milk if it costs children’s lives.
Under our current innovation ecosystem paradigm, the accelerator program model turns even essential innovations instrumental. This is because accelerators require a founder or group of founders to market their idea first to sources of capital that are aligned with their aims: either return on investment for for-profits, or charitable aims (and reputational capital) for non-profits. Before accepting founders into its program, an accelerator’s leadership needs to agree that there will be a clear source of financing to subsidize that venture’s losses until it can achieve a position of market dominance or sustainability. The very nature of an accelerator, therefore, distinguishes between the end goal of the venture itself and the instrumental value of the venture to sources of capital who seek to achieve their needs.
From an accelerator’s perspective, so long as the venture achieves its instrumental aim (to raise the capital they need to continue a stage or two) the venture is a success. This is no less true in social ventures, where accelerators declare success when philanthropists or impact investors give the founders the financial support they need to move their idea forward. This is why all accelerators, social or purely instrumental, focus on the tools the founders will need to sell their vision to investors or donors.
Accelerators, therefore, defocus founders from the essential impact they first envisioned. The assumption here is that their most practical goal is to raise money, and changing their idea is justified so long as it convinces a source of capital to invest. Paradoxically, we “speed up” ventures we think will better raise capital, rather than generating ventures that would better solve the problem. This leads us to support ventures that play well within our current system, rather than efforts that shift the system to correct for its perverse outcomes.
What paradoxically happens with accelerators is that we speed up ventures we think will do better to raise capital, as opposed to generating ventures that would do better to solve the problem.
At PresenTense we sought to correct for this by building a community around the entrepreneurs in partnership with local institutions that would, we hoped, be natural venture clients for the initiatives launched. In some cases, we were successful. But as a rule, all social venture accelerators seem to end up driving small, localized pockets of change that exist in contrast to the prevailing political economy, not as catalysts to change it.
If our goal is to lead to better outcomes for all — a re-industrialization that will cease the warming of our planet, a reduction of economic disparities, a stabilizing social force that will enable democracy to thrive — we need to establish a new approach to social innovation. This approach should be rooted in the public sphere where our societies express our common values. It should be aimed at collectively defining challenges. It should include cross-sectoral partnerships as its economic foundation, and it should open opportunities for many different individual actors to participate in the innovation so that entrepreneurship is not limited to the few who can afford to take risks.
The model I am proposing I am calling a generator. Unlike an accelerator, or an incubator, generators are built to regularly take up challenges and seek to solve them through experiments by transdisciplinary teams supported by, and guided by, their local institutions. Whereas an accelerator helps a founder build the right narrative and pitch to raise capital, a generator focuses on helping community members come together to learn about the challenges their community faces and to do the hard work of exploring different ways of addressing those challenges. Whereas accelerators or incubators seek to increase the value of a company for a single stack of shareholders, a generator shares in both the labor and the ownership with all of its members so that many different experiments can be run without legal fictions restricting achievement of the best possible outcome moving forward.
Comparing how accelerators and generators view innovation processes
Our startup ecosystem — of which accelerators are a core part — are based on the following assumptions: innovations are best advanced by founders who sprint a marathon to bring the venture into the world through a single company they found. These assumptions seem so natural, and yet each one presents challenges to the work of generating essential innovations.
How accelerators advance innovation
Accelerators focus their work on founders because accelerators depend on an individual or group coming forward with an idea that is already developed, even if partially. This creates a dependency: by depending on founders, we depend on self-selection by individuals who can afford to be founders (overwhelmingly, folks who look and sound like I do: male, of a certain complexion or family background, with certain educational accomplishments, and most importantly with financial flexibility to take a 1–2 year risk on earnings).
The fact is that most essential products and services are not currently developed or served by people who would fall into this group of ‘founders.’ It is nearly impossible for a working teacher, farmer, nurse, or sanitation worker to take time from their career to join an accelerator. In the event that they do have an idea that can be sold to investors or donors, accelerators largely cannot afford to provide these essential innovators the conditions they need to take a sufficient break from their wage-earning to work on innovating new products or services.
Even if an essential innovator does, through a mix of fortune or family or funds, have the opportunity to take the break they need to join an accelerator, the second assumption often takes them for a spin: we expect our founders to sprint a marathon despite full knowledge that doing so is nearly impossible, and only 2 of every 1000 founders make it to the finish line. The marathon is, of course, the building of a company that takes an initial idea and develops around it the organizational and operational excellence to turn it into a product or service. The speed of the sprint is determined by the interests of capital: if VC, then the 5–7 year window by which the fund needs to show returns to its investors (or by which venture capitalists need to claim enough of growth in the value of their fund to raise another fund, for their own professional survival); if by foundations, the length of the attention the venture can demand from donors who are often easily distracted.
Everyone involved in the current startup ecosystem knows that the whole ‘marathon’ approach to company building is inefficient. Of the 998 founders who don’t finish the race, there are many, many excellent ideas for products and services that could noticeably make the world a better place, but which are relegated to the dustbin of our innovation ecosystem due to poor system dynamics. Of those two who make it across the finish line, most underwent leadership transitions between the stages of institutionalization and validation and growth, but due to the ownership dynamics, the founders had an explicit interest to remain in position, lest their initial sacrifice be gutted by investors’ stacked preferences.
That same third assumption — housing an innovation in a single entity — is the reason why founders in both for-profit companies and non-profit ventures seek to remain in place: to benefit from the fruits of their labor. Our dependence on the single corporate entity to advance innovation makes the opportunity cost of leaving a venture nearly unbearable for a founder who sacrificed so much to turn their idea into a reality. In almost all cases, the longer a founder can keep hold of the reins of the single company, the more they stand to benefit. On the for-profit side, we’ve built a culture of investment that is ‘stacked’ last in, first out: the last money invested into a company has preferential return compared to the first money invested, receiving payout before the founders or any of the ‘common’ stockholders receive reward from a job well done.
Similarly, non-profit founders often toil at no or deeply discounted salaries in the organizations they found until the organization matures. In both cases, the assumption that an innovative product or service is best built through a single company drives founders to the rational position of holding on for dear life to their seats, even if they aren’t the best suited to take the venture forward to the next stage.
How generators challenge these assumptions to advance essential innovation
The generator model flips all of these assumptions on their heads. Generators create a structured means for the market to seek product fit. By enlisting public institutions to address challenges through a relay-race model, generators empower teams of relevant subject matter experts to apply their energies to innovative approaches to those missions.
Generators create a structured means for the market to seek product fit by focusing public institutions on supporting missions through a relay-race approach to innovation
While accelerators center their programs around founders, generators root their programs in public institutions fiduciarily responsible for public welfare, called ‘Hosts’. These institutions are driven by the need for essential innovation to address a challenge they are unable to solve on their own. The generator’s host identifies the challenge to be addressed and invites peer organizations as partners to help solve the challenge. Together, the host and its partners enable a mission-oriented effort to catalyze essential innovations whose success is measured by predetermined objectives and key results.
Essentially, this means a generator starts by defining the market a product would fit into, as opposed to starting with a product that is seeking market fit. Starting with the market as opposed to starting with the product is especially important when it comes to essential innovations because it flips the traditional script: unlike our current model, wherein a host would wait for founders to pitch their solutions hoping for product-market fit, a generator directs the host to provide clear signals as to what it needs, how it defines success, and starts the discussion as to what capital the host and its partners can deploy towards solving the challenges they’ve identified.
By basing the program on the institutions facing the challenge, as opposed to individuals who believe they have a solution to a problem, generators can create the space for more experimentation and less individualized risk. A generator reduces barriers to entry for individuals by creating clearly defined and time-limited ways for subject matter experts to contribute their unique insights and skills. It does this through program stages that have been designed to match technology and product readiness levels that almost all ventures need to move through for success: invention, innovation, validation, commercialization, and if relevant, integration.
By limiting and defining individual commitment, a generator amplifies the expertise of the community through its relay-race innovation process, instead of demanding a founder or single team race their way from stage to stage. In each stage, the people supporting the effort are directed to work towards clearly defined outcomes. Leadership shifts between stages per the expertise required: when defining the experimental design to test the proof-of-concept, Experts lead; when building the pre-prototype to see if it meets basic criteria, Venture Builders lead. All share in the reward, and all recognize that the best way to mitigate risk is to put the best person in charge depending on the task at hand.
So how does a generator work?
Generators focus on the earliest stages of this essential innovation development cycle, starting with improving the understanding of essential needs and concluding with the development of new processes, products, and services validated by independent institutions challenged by the essential need. On a programmatic basis, the generator method follows the following steps:
- A generator is launched once a public institution decides a threat it has identified is sufficiently challenging to require new innovative approaches supported by actors outside of its organizational boundaries.
- The institution becomes the generator’s Host and invites other organizations to share their commitment to addressing the challenge by joining them as Partners.
- Together, the Host and Partners prioritize aspects of the challenge to be addressed, and recruit Experts to map the challenge implications and determine missions for innovators.
- The Host and Partners then invite local entrepreneurs, creatives, and makers to join as Builders to work with the Experts on developing new processes, products, and services in line with the mission to solve the challenge.
- The program concludes as Builders apply these innovative approaches in experimental sandboxes made available by the Partners to assess whether their innovative approaches can contribute towards solving the challenge chosen.
Each phase of the model runs for approximately two months and concludes with an interim deliverable that furthers the community’s toolkit in addressing the challenges it faces. Practically, the stages are catalyzed by workshops built upon the pioneering work done by systems innovation facilitators such as Si and the design and rapid prototyping facilitation design by GV and embodied in the design sprint method. The generator integrates these methods into a model for innovation support, around the following phases:
- The Initiation phase: Host and Partners identify a challenge and its ramifications for the community. The outcome of the discovery phase is the nomination of a group of 5–7 Experts with lived experience in identifying the contours of the challenge.
- The Discovery phase: Experts map the challenge and identify missions that contribute towards solutions. The outcome of the exploration phase is a transdisciplinary workshop open to the public, in which an invited 15–20 Builders will participate.
- The Exploration phase: Builders develop pre-prototypes for testing in sandboxes provided by the Partners. Builders are joined by 10–15 Apprentices from local schools, universities, and local corporations to assist while learning by working alongside the experts.
Products and services that pass the proof-of-concept stage are then presented by the full community to the Host and its Partners to determine which innovations to invest in further, which to open-license, and which to integrate into the organizations or companies that can use the knowhow for the greatest good.
Conclusion: we can generate more solutions to the challenges we face
Accelerators and incubators have an important place in furthering instrumental innovation, and have led to many great ventures that have changed millions of lives. If we believe, however, that we have a limited amount of time in which to address the wicked problems of economic disparity, climate change, and political instability, we need to innovate the process of innovation.
The generator model better directs the energies of our most creative, most committed citizens towards building a fairer, healthier, and more democratic world for future generations. The generator program model complements accelerators and incubators by focusing on the outcome of essential innovation that these models struggle to produce. The theory behind the generator builds on the thinking behind mission-oriented innovation, providing a practical, replicable means for any institution, big or small, to quickly develop a mission-oriented community of essential innovators to address challenges faced.
Most importantly, the generator program model provides an answer to all of those creative, searching individuals who understand the magnitude of the polycrisis facing us, recognize they cannot risk their salary, and every day ask themselves: what can I practically do to make the world a better place?
