Brad Feld and David Cohen built the first global accelerator on a radical premise: the scarcest thing in early-stage startups is not money. It is judgment.
David Cohen had just sold his last company and was sitting on enough capital to do something interesting. Brad Feld, the Boulder-based venture capitalist, had been thinking about the same question: why were the best accelerator programs so geographically concentrated? Why did a founder have to move to San Francisco to get access to smart, experienced operators?
Techstars launched its first batch in Boulder in the summer of 2007. 11 companies. Ninety days. A cohort model built not around lectures but around access - specifically, access to a carefully curated network of mentors who had actually built and sold technology companies.
"Give first." It became Techstars' unofficial operating principle - the idea that mentors should give their best advice and connections without expecting anything in return, and that this generosity would compound into a network that attracted the best founders.
The first Boulder batch produced companies that raised significant follow-on capital. The model was validated quickly enough that Feld and Cohen began expanding to other cities within two years.
Techstars' central insight is that early-stage company building is a craft that cannot be learned from books or lectures. It has to be transmitted from practitioners - people who have made the specific mistakes, navigated the specific crises, and developed the specific instincts that separate companies that survive from those that don't.
Each Techstars program recruits 100 to 150 mentors per city - investors, operators, founders, executives - who agree to a serious engagement with portfolio companies. Not a 30-minute coffee. An ongoing relationship over the course of the three-month program, with the expectation that mentors will give substantive help on real problems.
The Techstars application asks how you'd use 100 mentors in 13 weeks. Companies that have thought about that question concretely tend to get more out of the program than companies treating it as a pitch process.
SendGrid. Before SendGrid, transactional email was a nightmare. Every developer building an application that needed to send email had to configure and maintain their own mail server, manage deliverability, handle bounces, navigate spam filters. Techstars backed the SendGrid founders in 2009. The company was acquired by Twilio in 2019 for approximately $3 billion.
PillPack. TJ Parker came to Techstars Boston with an idea about pharmacy - specifically, the problem that people who take multiple medications every day have organizing their doses. PillPack reinvented the pharmacy model around pre-sorted daily medication packaging and home delivery. Amazon acquired PillPack in 2018 for approximately $1 billion.
ClassPass. Payal Kadakia founded ClassPass after spending 45 minutes failing to find a ballet class. Techstars backed ClassPass in 2013. It reached unicorn status in 2020.
Techstars' expansion from one city to a global network happened faster than almost anyone expected. By 2010, programs existed in Boston and Seattle. By 2015, Techstars was running programs on six continents.
The corporate partnership model was key to this expansion. The branded accelerator format - the Barclays Accelerator powered by Techstars, the Sprint Accelerator, the Target Retail Accelerator - gave corporations a structured way to engage with early-stage innovation and gave Techstars a sustainable revenue model for running programs globally.
The corporate partnership model was controversial inside the startup community. Critics argued that corporate sponsorship distorted program incentives. What it actually did was make Techstars the first accelerator to operate at genuinely global scale.
Techstars' application process is more extensive than most accelerators'. Companies apply to specific programs - the Boulder program, the New York program, the Berlin program - and the selection process is managed by the managing director of each program, who has significant autonomy in choosing companies.
The most important part of the process is the in-person interview, which is closer to a working session than a pitch. Candidates are often asked to demonstrate something rather than simply describing it. The program has a low tolerance for founders who are good at talking about what they are building but can't show it.
The "mentor madness" period in the first few weeks is designed as a rapid matching process: portfolio companies meet large numbers of mentors in short sessions, and mentors who feel a genuine connection opt in to a deeper ongoing relationship. Deliberately chaotic. The chaos produces signal about which founder-mentor pairings have real chemistry.
Techstars has backed more than 3,500 companies across more than 50 programs worldwide. The combined value of those companies exceeds $115 billion. But the more durable legacy may be the mentor network itself - the community of several thousand operators, investors, and founders who have participated in Techstars programs and who continue to give back to the ecosystem.
The "give first" culture has produced a network that continues to generate value long after any individual program ends. Techstars alumni refer to each other. Techstars mentors invest in Techstars alumni companies. Managing directors become venture capitalists who back the next generation of Techstars founders. The compounding is deliberate design, not accident.
In a world where most accelerators are fundamentally capital-deployment vehicles, Techstars built something different: a judgment-amplification network. The capital followed the judgment. It always does.
Vaomi helps launch your own private program with the signal, structure, and network to find founders first.