Skip to content

Seed Accelerator Rankings Project

An academic ranking of U.S. seed accelerators led by Rice University professor Yael Hochberg with Susan Cohen and Daniel Fehder, backed by MIT, the University of Richmond and later the University of Georgia. It placed roughly 15 to 31 programs per year into tiers (Platinum+ down to Silver or Bronze) based on confidential portfolio outcome data and alumni surveys. Founders used it as one of the few data-driven, non-commercial comparisons of accelerator outcomes.

Reliability
Grade C · 7/12
Ranks
Accelerators, United States
Editions
7 since 2011
Method
Mixed
How items get in
Free, institutions submit data
Status
Discontinued

Editions

EditionPublishedTop of the listOn record
2018October 26, 2018AngelPad; StartX; Y Combinator (Platinum+)25
2017 Method changedJune 5, 2017AngelPad; Y Combinator (Platinum+)30 of 31
2016 Method changedMarch 11, 2016Platinum tier (8 programs)23
2015 Method changedMarch 17, 2015AngelPad20
2013 Method changedMarch 10, 2014Y Combinator15
2012August 22, 2012Y Combinator3 of 15
2011May 2, 2011TechStars Boulder15

How the ranking works

Eligible programs had to fit the researchers' definition of an accelerator (fixed-term, cohort-based program with mentorship and education ending in a demo day), be U.S.-based and have graduated at least one cohort with 10 or more startups. Programs supplied confidential data on portfolio outcomes, supplemented by Crunchbase and other public or proprietary sources where programs did not participate.

The 2017 report scored programs on valuations, qualified exits (IPO or acquisition of $5M or more), qualified fundraising ($200K or more), survival and founder satisfaction (a Net Promoter Score from alumni surveys, roughly 1,000 respondents), with extra weight on valuations, fundraising and exits. Programs were then grouped into tiers by clustering of the composite score, with no ordinal ranking inside tiers. Exact weights were not published in the press releases.

The method changed several times: the 2011 and 2012 precursor rankings (run with Tech Cocktail and Kellogg) used financing events, exits, VC reputation and program terms; the March 2014 release used ranks within Gold, Silver and Bronze bands; the March 2015 release was a plain top 20; from 2016 the project switched to tiers only.

Tiers: Items grouped into named tiers with no order inside a tier. No fee was charged. Researchers invited more than 150 U.S. programs (160+ in 2018) to share confidential portfolio data and to survey their alumni; programs that met the definition could be assessed, and for non-participants such as Y Combinator the team reconstructed data from public and proprietary sources.

Read the publisher’s methodology

How far to trust it

Superscout grades every ranking on six criteria, each scored 0 to 2. Seed Accelerator Rankings Project scores 7 of 12, grade C.

Transparency 1/2
Inputs and tiering approach were described, but weights and program scores were not published.
Independence 2/2
Academic project with no fees or sponsor stake in results.
Evidence 1/2
Mix of confidential outcome data and alumni satisfaction surveys.
Stability 0/2
Format moved from ordinal ranks to banded ranks to tiers and tier names changed between editions.
Coverage 1/2
Invited 150 to 160+ U.S. programs but published only the top 15 to 31.
Track record 2/2
Seven to eight editions from 2011 to 2018.

History

The project grew out of a 2010 accelerator study and a May 2011 Tech Cocktail ranking of the top 15 U.S. accelerators led by Aziz Gilani (DFJ Mercury) with Northwestern's Kellogg School, which Hochberg assisted; a second Tech Cocktail ranking with Hochberg and Kristen Kamath followed in August 2012. Hochberg's CV lists the Seed Accelerator Ranking Project from spring 2012 with herself as managing director. From March 2014 the rankings were released at SXSW by Hochberg (then MIT Sloan) and Susan Cohen (University of Richmond), with Daniel Fehder joining from 2015.

In 2015 Y Combinator was absent because it reclassified itself as a seed fund; from 2016 the team included it using public data. The 2016 edition dropped ordinal ranks in favor of tiers, 2017 added a Platinum+ tier and a Bronze tier, and the eighth and final edition appeared in October 2018 with the University of Georgia as a partner. No later edition has been found.

How it is received

The rankings were widely covered by TechCrunch, Forbes, Inc. and Harvard Business Review, and ranked programs such as AngelPad, ZeroTo510 and Health Wildcatters featured their tier in marketing (AngelPad's site still describes itself as the top U.S. accelerator based on these rankings). Its academic authorship and reliance on outcome data made it the most cited non-commercial accelerator ranking of the 2010s.

Limits noted in coverage include dependence on programs choosing to share confidential data, the need to reconstruct data for non-participants such as Y Combinator (TechCrunch, 2016), and the U.S.-only scope. Press copies of the 2016 tier lists differ slightly (Forbes and the University of Richmond list StartX or NMotion differently from Rice's release), which shows how hard the lists are to verify now that seedrankings.com is offline.

Strengths

  • Built on confidential portfolio outcome data plus alumni surveys, not on reputation alone.
  • Academic authors with no fee and no sponsor stake in the results.
  • Clear eligibility definition for what counts as an accelerator.
  • Tiers avoid false precision between programs with similar outcomes.

Weaknesses

  • Discontinued after 2018, so it says nothing about current program quality.
  • U.S. programs only.
  • Method changed almost every edition (ranks, then bands, then tiers), which limits year-to-year comparison.
  • Depends on programs volunteering data; non-participants were estimated from public sources.
  • Weights and full scores were never published in the press releases, and the project site is offline.

Open questions

Facts not verified yet. They stay visible so they are not mistaken for settled facts.