For marketing, the stakes are high. In media, everybody wins when the right consumer gets the right message.

Waste in advertising is historically endemic and significant. For example, half of all online display impressions are never viewable. Eliminating waste is a C-suite imperative in the new normal.

Big data is the gas tank of the new marketing machine, and analytic systems are becoming the engine, but we're missing a few parts.

Every car has a dashboard, and now so does the marketing machine.

The new standard for delivery for marketing analytics is the "software as a service" model. SAAS companies have higher valuations because the SAAS model strips humans from the equation in favor of scalable computers.

The inevitable outcome is that the new analytics do not come with service. What we get from SAAS companies are "dashboards" -- web-enabled views into the data -- and there are hundreds of them.

So we have too many dashboards, and lack the skills and processing capability to fully exploit them. The new marketing machine needs new drivers, and new controls to respond to the real-time information these systems are giving us -- a better steering wheel.

McKinsey & Co suggests that U.S. organizations are facing a shortage of 200,000 IT staffers with deep analytics skills. Even if we somehow fill this gap, how many of those will have advertising skills, too? Puffy résumés notwithstanding, this is a rare talent combination.

Compounding the problem in large companies is the proliferation of tools and platforms. Every brand is attached to its dashboards. Getting scale across an enterprise or across systems will require a type of discipline that is awkward for a brand process. Insights, especially, require merging previously disparate kinds of data -- and we are missing a plug-and-play way to do that .

So, for now, our marketing machine has a powerful engine, lots of gas, 1,000 dashboards, too few mechanics, bad drivers, and no chassis to hang all the parts on.

The chassis is the framework that makes all the parts work together. In the technology world, that 's called "architecture": principles, models and standards. In marketing there is little patience for investment in standards unless the client says to do it, and is willing to pay. Little wonder that trade pundits are predicting the rise of a role called chief analytics officer.

Marketing analytics has promise, but bumps along for lack of people to integrate the knowledge, and a common plan to drive scale. Where will service, skill and integration come from? How can we make progress? Here are a few suggestions:

We need a common way to think about measurements. The ARF's in-progress digital-marketing-measures compendium illuminates the problem: There are hundreds of measures available. Nielsen has a nice contribution with its "3Rs" approach: reach, reaction and resonance.

We need norms for how brands, agencies and analytics providers interact. Should agencies be the integrators of platforms that brands need, and so provide one dashboard to each brand client? Or should enterprise IT solve the problem and let agencies use the data? Agency proliferation, by itself, impedes the notion of integrated analytics, so the enterprise or brand might become the integrator.

We need the education system to crank out analytics-skilled people that know more about "why" and "what," not just "how."

We need the measurement community to partner the same way that digital ecosystem players now collaborate -- quicky, openly and fluidly. The measurement community has pole position in this race. They see the gestalt of technology, data, marketing and measurement, and know how to put insight into action. These values make up the best roots from which to create capabilities to drive the new machine.

The fragmented market situation begs for a power player to step in and rationalize the industry. Google Analytics has that potential, for example. In that scenario, agencies and media companies will have a third party controlling the perception of their performance -- probably not their preference.

One thing is clear. If we don't know where we are going, we are unlikely to get there. Right now, we're adding wings to a Chevy and wondering why the darn thing won't fly.