09/30/2013 07:15 pm ET Updated Nov 30, 2013

My Obamacare Anxiety: America's Largest Digital Start-Up Probably Lacks the Right Digital Talent

"If a team can't be fed with two pizzas, it's too big." -- Jeff Bezos

The Bezos two-pizza rule, and its insistence that only small, intense, swiftly-moving teams can be effective, has become a start-up item of faith. I hope its religion has shaped the process by which the health care exchanges -- which reveal themselves tomorrow -- have been built.

But I doubt it. I'm deeply skeptical because governments have a bias to bigness, consensus and rudderless leadership. You can't launch a start-up with sprawling teams. In fact, you probably can't run anything that way, as the teams-don't-work work of Harvard Professor J. Richard Hackman has found.

So as we await the grand unveiling, we should all be concerned that the fate of the largest transformation of our health care system is largely in the operational hands of people who aren't part of the digital world, who've never run a consumer start-up before -- let alone a consumer ecommerce business, which is essentially what Obamacare is. Eighteen percent of our GDP is no time for on-the-job training.

And by the way, the problem is multiplied by the fact that because states have the right to run their own exchanges, we have a federal program in the 36 states that opted in, and state-run exchanges -- entirely separate brands -- in the states that built their own.

Even worse, many of the actual companies who are building the exchanges -- as the Daily Beast reported -- are traditional government IT contractors. Check out their websites and you'll see the recipe for disaster. Companies like hCentive and CGI have never built anything that touches the consumer directly. They've never created a brand, never dealt with the intuitive presentation of visual information, have no experience with conversion funnels and the engineering of choice.

Then there's the pesky detail that behavioral psychologists call cognitive bias. It explains why human beings consistently act against their best interests, why we are "Predictably Irrational" as my friend Dan Ariely described our maddening species in his best-selling book. Selling health insurance hits every neuron in our primitive brains, and it's why we need digital marketers who understand the amygdala and how to tame it.

I know the kind of talent required for this, and it doesn't want to work for the government, or for government contractors. I've spent more than 10 years advising CEOs, entrepreneurs, and dozens of online brands including -- from which the federal and state governments can earn a lot. I've helped grow and build exchanges and platforms in personal finance, travel, art and antiques and partnered with AT&T wireless to sell millions of cell phones online when no one thought it could be done. In fact, the complexity of the "choose a phone/choose a plan" sequencing is not dissimilar from the challenges the health care exchanges face.

The exchanges need the absolute best talent out there, not the talent that happens to be around, or the talent that has the patience and previous experience to submit to the exhaustion of a government RFP. (I've seen those documents and they are soul-crushing. With so much exciting work for so few talented people in the digital world, why would anyone let themselves be pulverized into contractual dust?)

But as complicated as they sound, on a structural level the health care exchanges are no different than any ecommerce website. You need to build a marketing front end that explains the product and lures consumers into the funnel, and a transactional back-end that, in this case, gets tied into the back-end of the heath insurance companies.

Creating a seamless, branded experience out of this involves a tight collaboration with marketing and product teams. Even in organizations that do this every day, there's inevitably friction and stress. I shudder to think of what it's like inside a government-run structure that's never done this before.

That's probably the reason that expectations have been set so low. We've already been told that Day One of the exchange is going to be messy, with epic glitches. That's to be expected, although I hope that they've done a ton of usability testing during these last few months. They also should have created a friends-and-family alpha and beta test -- as it's called in the digital world -- where a couple of hundred people actually run through the process of buying insurance, and report on their experience so the site can iterate and reiterate.

With that context, here's my advice for the exchanges going forward

  • Innovative ways to display visual information is core.

Ordinary eyes glaze over when faced with a spreadsheet. The good news is that there is brilliant work being done in presenting complex data with breathtaking simplicity. (Here's a terrific TED talk on making information beautiful.) That's what it's going to take to win young people over, especially millennial men who are the toughest customers for the exchanges, believing as they do in their invincibility.

Now, take a look at this page I found on the HealthSource RhodeIsland. Do you know any young guy who wouldn't pay the penalty of $95 a year to avoid this pain?

  • Remember the Paradox of Choice

It's a seminal book by Barry Schwartz, and it should be required reading for anyone involved in creating the exchanges. His thesis, that more is less, is particularly relevant in a situation when consumers are being drowned in a sea of moreness. Even in markets where there are only two plans, there is an army of contingent variables to deal with -- including level of plan, size of deductible (see below) and which drugs are covered.

What's important is to break the complex into digestible nuggets that consumers can absorb and accept in sequence. When they accept one nugget, they're more likely to accept the entire nugget cluster. Robert Cialdini in his book Influence, describes this as the power of "commitment and consistency.

  • Don't create cognitive dissonance in your packages

I worry that this has happened already, based on the Bronze, Silver, Gold, Platinum shtick. The problem is that the convention of that laddering -- borrowing from the credit card business that invented it -- is that value linearly increases with cost. But it's asymmetrical here: The lower cost Bronze has the highest out-of-pocket cost; different drugs are covered on different plans; and different doctor networks are included on different plans. So the "metals" model sends the wrong message. A middle-income household with specific prescription drug requirements might be better off with the Platinum Plan; but it's branded the wrong way, as luxury item, and will scare customers away who need it.

This goes beyond the one-dimensional financial calculators that are going to be part of every site. It's going to take some exceedingly insightful info-structuring to assure that people get what they need, and need what they get.

  • Obsess over abandonment

Because the number of consumers who buy the first time is going to be low, perhaps shockingly low, it is imperative that those operating the sites study the places where people jump ship, understand the pain points, and address them promptly.

Those who operate ecommerce websites know that they are a continual work in progress; Amazon and others are ceaselessly testing and optimizing, often in real time. (I've heard that Amazon can be testing more than 100 cells at once.) And there are innovative companies who are pioneering new ways to reduce abandonment; for example, CommerceSciences, presents prompts to users based on the precepts of behavioral psychology, recognizing that some people are motivated by price, others by social proof.

Government-run sites are static -- whether they are built internally or by contractors, they are launched and left. So it will require a dramatic culture shift to move into the dynamic world of nothing-is-ever-finished. It also requires leadership that is plugged into the leading edge of Big Data, algorithmic science, loyalty marketing and rewards.

  • Build effective follow-up and re-targeting mechanisms

If the federal government and the state-run exchanges haven't planned extensive, big-data informed retargeting campaigns, then we're wasting taxpayer money by attracting consumers to the sites without follow-up up strategies.

Retargeting is an essential part of running an ecommerce operation today; essentially it means that you don't stop marketing after your customer clicks off your site. You never surrender. You continue to market your product or service to them wherever they are on the web, often with a promotion. (But you don't know how much to offer until you have an ROI model built.... see below.)

Many, if not most consumer acquisition campaigns would be ROI-negative if not for retargeting, which now has a complex ecosystem of companies who offer these services; MyBuys is just one example. You can, by the way, retarget on Facebook through the Facebook Exchange. Here's an example of how that works.

The exchanges also need to use sophisticated CRM strategies to follow up with prospects who've provided email addresses but haven't bought yet.

  • Build an ROI model.

Every self-respecting ecommerce site understands customer lifetime value, and how much they're willing to pay to acquire a customer based on that. They know cost-per-lead, cost-per-conversion, and all the attendant metrics and sub-metrics. That's complicated in this case, because there are so many unknowns; I'm also not quite sure of what if any percentage of the insurance premium goes to the government in a rev-share agreement.

But as a taxpayer, I'd be mightily chagrined if the insurance companies kept the entire premium, while the government invested its millions in generating traffic and leads without any immediate economic benefit. That would be a ginormous taxpayer investment in the insurance industry with sketchy return. What they call corporate welfare.

Yes, government has an obligation to keep people healthy, and yes, reducing the cost of health care via the exchanges is a positive vector, but you shouldn't be spending money on fancy TV advertising without a rigorous ROI model. That fancy advertising includes, as Kaiser Health News reports:

"A folk singer playing guitar in front of a mountain stream. A Disney-like animated video about how 'a new day is coming.' An announcer talking about 'change is here.' A woman jumping up and down in celebration in a baseball team locker room."

{Bulletin: The Arkansas Legislative Council today rejected a $4.5MM expenditure on TV advertising for their exchange. Note that with a population of $3MM people, that translates to a national advertising budget of over $450 million dollars, which is about Budweiser spends. Sounds like a prudent veto to me.}

There are some things government does an okay job at. Building roads, rescuing people from floods, threatening Syria. For the sake of our health care system, let's hope that building friendly, functioning, convincing and intuitive ecommerce sites is one of them.
We shall see.