Among all the changes in the news business, one trend is clear: foundations are playing a bigger role in journalism than ever before. They’re funding startups; they’re starting their own startups; even established news organizations are talking about wanting foundation money.

We journalists hear a lot about this relationship from our side — that is, the side that wants the money. But it’s instructive to look at it from the other side: What are foundations looking for when they fund media projects? How do they judge a success? What grants do they come to regret?

We get that sort of perspective in this article (PDF), written by Susan King, who among other things leads the journalism efforts of the Carnegie Corporation. It’s from the current issue of Carnegie Results.

King writes about the first grant she ever approved — $354,000 to the American Communications Foundation, a nonprofit that made radio and TV stories for commercial outlets. The money was intended to fund stories about issues of interest to Carnegie, but King, for a variety of reasons, wasn’t happy with the results.

You’ll want to read the whole thing to see Carnegie’s perspective, but here are a few takeaways I drew from the piece:

— The demands of free-market journalism don’t always align with the interests of foundations. Carnegie wanted stories on topics like public education and campaign-finance reform. And it got them — but King criticized them as “thin, often cute and most of the time not squarely focused” on Carnegie’s needs. ACF responded by saying, in essence, you can’t put white papers on commercial radio — that a certain amount of sweetening was required to reach a broad audience. There may be truth to that, but Carnegie wanted stories with more heft.

— The measurements journalists use to show success may not be the ones foundations use. King notes that ACF did put its stories before a large audience — one measured in the millions. But she felt that, despite their reach, the stories hasn’t made much of a difference. “[ACF] might have been able to tell how many stories were produced, how many millions of people listened to or saw each story, but foundations now defined that information as ‘output, not impact.'”

For online journalism startups, the sheer measurability of web sites can be a comfort. One can find solace in the knowledge that you have X unique visitors a month, Y subscribers to your mailing list, and Z followers on Twitter. But King is saying that may not be enough — evidence of impact and engagement, elusive as it may be, is more powerful than a count of eyeballs.

— That shift of focus could be a good thing for investigative journalism. After all, tough watchdog stories may not reach a huge audience, but they can have direct, tangible impacts: changed laws, fired officials, shifts in policy. Those are things both journalists and foundations looking for impact can point to. If foundations end up funding more hard-hitting journalism and less filler, there’ll still be a lot of unemployed journalists — but the hit on democracy will be lessened.

— Funding for middlemen is going to be more of a struggle than funding directly to news organizations. ACF was in the business of creating stories that it would then present to radio and TV stations — which would in turn present them to their audiences. The wave of foundation support we’ve seen to sites like MinnPost and Voice of San Diego cuts out the middle player in that transaction — the foundations can directly fund the work they want to see, without worrying about navigating the needs of one more layer.

The logical extension of this trend is something like Kaiser Health News, in which a foundation cuts out all the layers and produces the journalism itself, directly hiring reporters and editors. (Although, to be fair, Kaiser also works with established news organizations.) As the Internet destroys the power of established channels and opens the audience to all comers, the need to work through existing players goes down.

Here’s King’s conclusion:

Collaborations with journalism organizations will never be easy when news leaders fundamentally believe in independence and foundations want results. I think both sides need to be much more specific and clear when creating partnerships, so there is no misunderstanding of what each side’s goals are and how much collaboration is possible in a relationship with an editorially independent free press. I do believe that, along with the serious players in the news industry, foundations share a desire to serve and to inform the public. But having clarity about what “public service” means, what good journalism is and can accomplish, demands conversation, great debate and frank understanding of each side’s different perspectives.