I've recently come to terms with the fact that the agency business is almost 100% commoditized as a result of its own ineffectiveness and unwillingness to adapt. The agency of the future will be 90% automated, with only minimal human curation of the data and the systems that are put in place, and a modicum of strategy applied as well. The human component will exist for top-level strategic thinking, and to develop research and POVs on data opportunities and evaluate new platforms -- all of which serve to further automate the media planning, buying and execution process. Oh, and let's not forget humans will be there to provide IT assistance for when the systems inevitably have hiccups.
The digital side of the business is headed to this realm of automation and there's not too much anyone can do about it. The simple fact is, there's too much confusion, too much clutter, too few trained professionals, and too many fly-by-night solutions that profess to be expert but actually serve to bring down the average performance of the category.
Agency/client relationships last an average of four years in the digital world, and slightly longer in the offline world. The debate between content buying and audience buying in online won't even be relevant in five years; the "content" buying component of digital media is being automated, for better or worse, and the "audience" buying model is proving to be effective. The results are clear. Using data to drive targeting works -- and this can all be mechanized.
The offline world won't be far behind, once RFID tags or some other measurement solution are placed in print ads (assuming they don't all shift to iPad-esque versions) and TV shifts to a 100% digital, accountable model (which will definitely happen in the next 10 years). Even outdoor ads are going to be tagged with GPS transponders that will measure the volume and average length of exposure by monitoring coordinates for cell phones in a strict vicinity of line of sight to the ads.
The agency of the future will have a creative team that develops a concept and a list of assets, and then loads those assets into a dynamic ad delivery system that tailors the ads to whatever location they're delivered, whether it be TV, online or other. Those ads will report back to a centralized dashboard of activity that also measures sales in real time. Optimization will happen in real time with those same dynamic ad systems coordinating assets against proven targets. It may sound like a pipe dream, but even video can be executed in this way. It may not be as creative as some of the executions you see now, but the almighty dollar rules the game, even in the world of mad men.
The agency of the future will have a media analysis team responsible for inputting a media brief into a DSP-type system that then spits out recommendations, across multiple media formats, and plugs into those ad development and delivery systems I just mentioned. The real digital dashboard will be available for them to review, as well as for their clients to review (which means account execs are out of a job, too).
I know it sounds dismal, but think positively for a moment. What this model will do is reduce the headcount in the agency structure and focus the open headcount against specialized, data-centric professionals who can't be commoditized and who can't be outsourced. These specialists will provide true value, and these people will have to be trained to use the systems and understand the process by which that new agency paradigm will operate. This model makes the agency valuable again, outside of just "shaping" emerging media for brands.
You can argue until you're blue in the face, but if you know how the major holding companies are run, with their profit margin goals and growth projections on an annualized basis, then you understand that a more specialized, lower-headcount, lower-cost solution means a higher margin and more reinvestment opportunities. Technology is the way of the future in most categories -- and I now firmly believe that it will be the way of the future in this one.
Tell me I'm wrong, please. I would love to see it another way!