The biggest challenge is going to be selling the idea that 4-year degrees are worth the ~$100,000 and debt burden they can cause. You cannot really sell the unsellable, so the industry will fail to meet this challenge, just as the newspaper industry failed to meet the challenge of selling the idea that a thick sheaf of dead trees at your doorstep for a buck a day or more is still worth it.

The marketing challenge is in a way irrelevant. The disruption challenge to the fundamental business model in education is far more important. Once a new business model emerges that employs the new technologies in smarter ways to deliver better education at lower cost, THEN you can talk about marketing challenges.

My bet, if I had any player to place it on, would be on completely reinvented continuing education. Many trends point this way (this is US specific, education tends to be very highly country-dependent as a market, but as the leading higher-ed supplier in the world, where the US is today, the rest of the world will be tomorrow).

1. Many students already take 6-8 years to finish a 4-year degree, juggling a part-time job, carrying as much debt as they can manage, and so forth
2. The population of non-traditional students, who are older is rising far faster than the "typical" 18-22 full-time model
3. More cost-effective for profit universities are taking away much of the practical/vocational training market, which in information work, has a very blurred boundary with "real" eduction (i.e. the mechanical engineer/auto mechanic distinction is far stronger than the CS grad/IT tech or business major/Excel wrangler distinctions). This is because modern education is very friendly to free "higher" self-learning around vocational learning. I know many autodidacts from vocational training backgrounds who know vastly more than 4-year types.
4. Marketable job skills become obsolete in less than 4 years. It is starting to seem silly to students that their "foundational" coursework as freshmen becomes irrelevant by the time they graduate.
5. The abstract/liberal education end of the spectrum is increasingly well-served by free content ranging from TED talks to more TV-academic material (Stephen Levitt started this trend with Freakonomics, and thanks to his one book, many liberal arts majors now understand stats far better than they learned it in their dull, uninspired required course which turned them into uncomprehending SPSS monkeys).
6. The entitlement burden due to tenure is increasing, as students pay the costs of faculty who are either deadwood, or so absorbed in research that teaching takes a hit.

About the ONLY thing traditional 4-year degrees have going for them is the opportunity to build a strong peer group/cohort of friends who will last you the rest of your life, get you jobs/opportunities etc. But the price tag just for that is increasingly seeming ridiculous, and there are other ways to find cheap or free peer groups with whom to come of age (example: just move to the Bay Area startup community, or land in Washington and learn politics/governance hands on through free internships/volunteering etc).

I am not denying the value of higher ed (I've spent nearly 12 years earning 3 degrees and a postdoc badge, mostly at the expense of taxpayers in India and the US, and spent 8 semesters teaching at various levels, including designing an original course). I am questioning whether it is worth it for those who have to pay near-sticker price for the base 4-year offering.

Used to be that you could reliably claim that every additional year of higher-ed would lead to a 6% increase in annual income after, so the calculation was a no-brainer.

Not so anymore. The standard ROI equation still works (though the return rate is going down) for some fields, but in most it is falling apart. And in plenty of fields (marketing, software), if you know what you are doing, you can absolutely beat that return by gaining 4 years of properly chosen work experiences instead.

So... what new business model might save the industry?

The answer is on-demand education. And I don't mean MIT's Open Courseware or Khan Academy. I mean on-demand education in the fully interactive sense where you can interact with teachers and peers.

The way to offer this is to let students put in 3-4 semesters upfront to get the basics down and start their peer networks, and then earn credits for the rest of their lives in an "unbundled" way wherever they wander, through some sort of credit-exchange scheme, where they'll get their degree from their home university, which will be responsible for QA, but otherwise be free to earn their credits anywhere, at whatever cost, through a mix of work-learning and full-classroom experiences. There can even be schemes to pay up front for the remaining 4-5 semesters worth of classes at discounted rates, like a pre-paid cellphone. Or pay-as-you-go.

The University's job will be to maintain its brand, control the quality in its exchange offerings (priced by the individual course, rather than a uniform rate per credit hour), and negotiate access to work-learning experiences through far stronger work-study/co-op type programs that make it far easier to get on the job learning experiences compared to getting a real job.

Employers will need to adapt to work with this transient, continuously-learning labor force.

Universities will need to shrink/eliminate tenure based faculty models for a far more fluid and open model, involving a lot more temporary instructors, visiting instructors, faculty exchange programs and the like. This is already effectively happening, due to the exploding adjunct population. It used to be seen as a supply/demand problem (too many PhDs chasing too few faculty spots), but now it is part of the solution: this churning instructor soup will allow universities to offer courses in far more agile ways, taking advantage of ephemeral, but not vacuous, trends, as well as taking advantage of people who like to teach, but not all the time or all their lives.

This applies both to fundamental courses that have been around (and will continue to be around) for 100s of years, and things like iPhone game design, that might appear and disappear before most curriculum committees even notice.

Many people interested in education have in fact lost interest in tenured single-university models entirely, and PREFER contracted short-term teaching gigs. I count myself in this camp now, even though when I finished my postdoc, I was briefly disappointed at not finding a tenure-track position. Now, if I go back to any teaching, I'd rather do a semester here and there at random universities as a roving adjunct/guest than as a full-timer.

Think of this as basically unbundling, opening up of, and partial amateurization of the education industry in an open format (i.e. not iTunes, but MP3s with no DRM). There will also be increasing prosumerization. You might be teaching one course while taking another, before you've completed your degree. In these turbulent times, the graybeards have much to learn from their young paduans.

In fact, the 4-year cumulative marker (which I think is still important) will be more an indicator of a certain amount of robust life experience with a good balance of doing vs. learning, than a life stage.

John Hagel noted in "Unbundling the Corporation" a few years back that every industry will start to break down into 3 kinds of businesses: infrastructure, customer relationships and technology/product. Universities will basically shift into the infrastructure end of the business. The "product" part will move over to increasingly to the free-agent adjunct market, and the "customer relationship" part will increasingly move over to learning communities all over the place, which will operate sort of like the craft guilds of medieval times.