The ARPA-E agency is known for its high-risk energy research projects. But this year’s annual conference had elements of a business bootcamp, offering would-be energy entrepreneurs tips on how to raise money and build a commercially viable product.

That focus on the practical is deliberate. The agency’s mission of pursuing potential breakthroughs hasn’t changed. But four years after launching, it’s increased the emphasis on commercialization, says deputy director Cheryl Martin. “We believe big ideas are important but only if they make a difference,” Martin said during the opening keynote of the conference earlier this week.

Energy secretary Steven Chu hired Martin one year and a half ago to lead ARPA-E’s Technology-to-Market program, which has resulted in some changes in how research projects are structured. The goal of an ARPA-E grant, which are usually about three years, is to produce a first commercial prototype. Researchers aim for specific cost targets and now need to contact potential business partners to help in the commercialization, Martin says. (What ARPA-E Does Well: Making Connections.) “I brought a significant focus on the market and how we should engage with the market to understand what those needs are. That wasn’t as present in as many places,” she says.

The research programs themselves have matured, too, she says. Rather than focus on an individual component, say a better battery electrode, the ARPA-E program managers try to take a systems view of how to design a battery for an electric car with a 500-mile range. “You can rethink the whole idea of form factors and architectures if you move to system-level thinking,” Martin says.

Bringing some business savvy to energy researchers can certainly help produce the technology home runs ARPA-E was created to produce. But a few attendees this year reminded me that the DOE still hasn’t adequately addressed the process of commercializing research, which has loomed over it for the past few years. (See, What ARPA-E Can’t Do.)

ARPA-E was modeled on the applied research method at DARPA where the military serves as a target customer. But energy lacks that “ecosystem” for advancing promising research into the marketplace. In some cases, grants have funded a group of researchers, but once the grant expired, many couldn’t continue work in that area of expertise, one attendee me.

A few years ago, many clean-tech venture capitalists trolled through the ARPA-E looking for promising university research and to connect with entrepreneurs. This year, it seemed as if there were fewer venture capitalists and startup companies. One VC joked he and a few colleagues where the “last ones standing.” This year’s conference had panels on how to work with large energy companies. These “strategic investors,” which are looking for access to technology, can play a significant role in advancing ARPA-E funded work but building strong, stand-alone startup companies isn’t really their mission.

The good news for ARPA-E is that it has widespread support, both from politicians of both parties and from large companies which could help bring some technologies to market. And ARPA-E can legitimately make the case that its relatively small investments—typically less than $10 million—are, on the whole, well spent. This week the agency announced that 17 projects, started with $70 million, have attracted $450 million in follow-on funding. Twelve were launched as new companies and over ten partnered with other government agencies for additional investment.

In the months ahead, ARPA-E will need to be reauthorized and a budget for next year appropriated during a period of across-the-board budget cuts in research. (See, R&D Faces its Own Fiscal Cliff.) Examples of ARPA-E research turning into successful commercial products could help make the case for keeping or increasing its budget. “We’re serious about this part of the mission,” Martin says. “It’s the way we will be judged.”