Another variation of this "challenge" approach is what GE has just done by releasing two months worth of flight data as part of a challenge to come up with better flight management algorithms. But neither of these approaches seems focused enough to guarantee the results buyers want. A P&G wants to know that a vendor is working on its problem, not just throw it wide open and hope something innovative comes back. An alternative might be a hybrid of the hub-and-spoke model: The CIOs agree on a common set of needs, and then each goes off to their preferred vendors.

Speed: Industry standards groups are a dime a dozen. A number of them are competing right now to create standards for cloud computing, for example. But standards groups tend to emerge after a flurry of innovation and implementations have happened already and companies start to feel the pain of supporting heterogeneous environments. "Speed" isn't the rallying cry of standards bodies. Getting consensus quickly on business needs and doing agile iterations along the way will be tough with any cross-company effort.

Agreement: A related tension point is sure to be how industry-specific or company-specific software needs are. One startup at P&G's Goldmine event, analytics software vendor Verix, illustrates this tension. P&G is partnering with Verix to develop the Israeli company's software. Verix's early success is driven in part by its industry focus -- bringing domain expertise in life sciences and, through close work with P&G, consumer packaged goods.

During a panel discussion at the P&G event, I asked two private-equity investors how they assess analytics companies. Are they more attracted to industry specialists or broader providers? Both leaned toward focus for the startups they back.

So could companies in very different industries -- such as consumer goods, aerospace, investment banking, media, hospitality, logistics, transportation and energy -- agree on what they need in new analytics software? Passerini ultimately thinks businesses are more alike than different, so there's room to collaborate. "It's all business," he said.

Energy: Sustaining any cross-company, collaborative effort is hard, especially around technology development. Remember RFID in the supply chain? Wal-Mart in the mid-2000s tried to turbo-charge a huge, cross-industry effort to spur development and use of radio-frequency ID technology, in which tiny chips are placed on goods to make it easier to track them. That industry-wide momentum foundered, even with Wal-Mart's enormous supply chain influence.

I asked Passerini about that example, and he said one lesson is that companies pushing for better analytics all need to see a clear payoff; with RFID, the return wasn't clear to everyone throughout the supply chain.

Urgency: We're back to Passerini's impatience. Do other companies -- both buyers and builders of software -- share his sense of urgency that we need a new way to develop software? At P&G, Passerini has shaken up the IT strategy every few years for a decade -- outsourcing much of the daily operations to Hewlett-Packard, then creating a service approach where business units can take or leave what IT offers, then pushing to give 60,000 employees analytical tools on their desktop.

"We use change as a strategy," Passerini said. "We obsolete our current model. We think it's important to do that when we're in a position of strength." Translation: Passerini thinks today's software development business model is working, but we should change it now before there's a crisis. Often it takes a crisis to drive change.

So that leaves the door open to you to articulate what's needed. Share your comments below. Is there a need for a better way to build software? What will work and what definitely won't?

Can data analysis keep students on track and improve college retention rates? Also in the premiere all-digital Analytics' Big Test issue of InformationWeek Education: Higher education is just as prone to tech-based disruption as other industries. (Free with registration.)