There is no such thing as free money. For education startups, the National Science Foundation has provided what may be the closest thing: upwards of $20 million every year to support their research and development efforts. There’s plenty of paperwork involved, of course. However, unlike traditional investment terms, this funding comes without exchanging equity or promising repayment.
But a recent reorganization of the NSF’s grantmaking portfolio and personnel has raised concerns about whether that support will continue—especially after a dozen edtech companies that recently applied for funding said they were all unfairly rejected.
Dubbed “America’s Seed Fund,” the NSF money is awarded through the Small Business Innovation and Research (SBIR) program, which directs federal agencies with research budgets exceeding $100 million to allocate a small percentage to support companies in the development and commercialization of new tools. The NSF is one of 12 federal departments and agencies that participate in this program.
Every year about 400 companies across a range of industries receive NSF SBIR funding, which is awarded across different stages. There’s a Phase I award that offers around $250,000 for building and testing prototypes. Should that go well, Phase I recipients can apply for a Phase II and follow-on grants totaling an additional $1.5 million in funding.
In their applications, companies select the technology area that best match their work—and education technology has been among the primary topics of interest identified by the NSF. This determines who reviews their applications and, if they get an award, who oversees and supports their projects.
Today, edtech is no longer a primary topic of interest, but a subtopic of other primary technology topics: artificial intelligence, information technologies and robotics.
The change is part of a broader reorganization of the NSF’s SBIR portfolios, according to Andrea Belz, a division director at NSF overseeing the SBIR program. To explain the changes, she says that “we saw that technologies serving the education market also exist in many other parts of our other portfolios. What we have done is unify the branches that were developing solutions for the education world.”
This means edtech is no longer its own separate portfolio with its own dedicated director. Rather, all future applicants seeking funding for edtech projects will be evaluated and compete for funding alongside other applications in AI, IT and robotics. This re-categorization is reflected in the latest SBIR solicitation issued in early December.
Belz, who joined the NSF in May 2019, did not say when the changes were first raised or considered. “Conversations go on, all the time, and we are constantly evaluating how to best serve different constituencies,” she offers. “There’s no sudden date when we started talking about it.”
Letter of Concern
The vagueness of the timing of these changes is at the core of a complaint letter sent on Jan. 31, addressed to Dawn Marie Tilbury, another director at the NSF. Signed by officials at 11 edtech companies, the note expressed “serious concerns” about how the changes to the edtech portfolio were communicated, and claimed that the lack of transparency unfairly impacted their applications for Phase II awards.
According to the letter, which EdSurge received a copy of, 12 companies that previously received Phase I funding for education technology had applied for Phase II funding in August 2019, when edtech was still a primary topic of interest and was its own NSF SBIR portfolio. The applications were then reviewed by a panel of judges assembled by the edtech portfolio director Rajesh Mehta.
The applicants did not hear back for several months. Then on Oct. 15, Mehta sent an email disclosing that he was being reassigned, and that all current edtech grantees would be moved into the Information Technology portfolio, overseen by Peter Atherton. The NSF later sent an email on Dec. 2 about the restructuring of the SBIR program.
Within a week of that notice, according to the complaint letter, “every single company that applied for Phase II funding was rejected.” Most were notified of this decision in an email from Belz, who declined to comment on the status of SBIR applications or how decisions were made.
Some of the applicants had received a “competitive” rating from the review panel, which is generally a strong indicator of a successful SBIR application. That was the case for Meredith Halpern-Ranzer, co-founder of Tinkercast and a signatory of the letter.
When she asked for an explanation, she recalls being told by Atherton that her Phase II application proposal did not meet the standard for technical risk, and that it was not “deep tech” enough. That didn’t make sense to her. “If you were accepted for Phase I, you are already supposed to have passed the litmus for technical risk,” she says.
Atherton did not respond to a request for comment.
To the aggrieved companies, it seemed that NSF officials had changed the grantmaking criteria after they had applied for a Phase II, and retroactively applied a different evaluation rubric to their applications.
“We are not at odds with the right of NSF to shift priorities, restructure, and allocate resources differently,” the letter states. “We do, however, believe that making such changes after companies have responded to a published solicitation with no communication, transparency, or opportunity to engage is neither fair nor reasonable. It is not in the spirit of NSF’s own guidelines or mission.”
To add insult to injury, none of the edtech companies currently appear in the web page for current NSF SBIR Phase I awardees—even under the Information Technology topic they were told they now belong to.
There is no guarantee that Phase I recipients will receive follow-on funding. But once a Phase II application is denied, a company cannot apply again. That has set back plans for those like Halpern-Ranzer, who was looking to expand Tinkercast’s children media services into the education market. “Without the support of this grant money, we’ll have to put these plans on hold.”
Many of the rejected applicants have asked the NSF to reconsider their application. Tilbury did not respond to a request for comment but notified the letter-writers that she is looking into their concerns.
A ‘Chilling Effect’?
NSF officials maintain that they remain committed to supporting educational startups. “We continue to encourage the submission of education-focused proposals” under AI, information technology, robotics or other topics, spokesperson Michael England said in an email to EdSurge.
Still, the portfolio reorganization, and the complaints raised in the letter, have already made some edtech entrepreneurs think twice before applying for NSF SBIR funding, according to Michelle Miller, CEO and co-founder of Games and Learning, Inc. and a current NSF Phase I grantee. (She did not sign the letter.) Completing a competitive SBIR application is no cakewalk, as the process can involve hundreds of hours of research and writing.
“Based on the information we have had, the changes have had a chilling effect on impact-driven digital learning developers,” says Miller, “especially when a funder that has shown historic interest appears to be shifting its investment focus.”
Funding from the NSF SBIR program has proven vital to jumpstarting the fortunes of many small companies. According to the agency’s estimates, 107 of the startups it funded through the SBIR program since 2014 have been acquired. They include edtech startups like literacy platform ActivelyLearn (acquired by Achieve3000) and AI-powered assessment tool Gradescope (bought by Turnitin).
The NSF is not the federal agency that awards SBIR funding for education projects, and edtech entrepreneurs may find some solace in the fact that they can also apply for this financial support elsewhere: the Institute of Education Sciences, a division of the U.S. Department of Education also offers SBIR awards. It is currently accepting applications through March 3, 2020.