In 2014, there’s been an awful lot of hype around programmatic advertising, which, at its most basic, involves automating the process of buying advertising inventory.

It looks set to continue into 2015.

Programmatic has reshaped the mobile advertising landscape, with many mobile ad networks reinventing themselves as demand side platforms (DSPs), focusing on their advertiser clients, and looking to the exchanges for their inventory.

After all, so one of the arguments goes, in the world of programmatic buying, it doesn’t really matter where your advertisement appears as long as it appears in front of the right person. It’s the algorithm’s job to find them.

So if your 50-year old Daily Telegraph reader also happens to be an avid Candy Crush player on the commute home, it’s just as easy and probably less costly to reach him via an ad between levels of Candy Crush than it is to reach him with a banner on the Telegraph’s mobile site.

Some of the stats and forecasts around programmatic are impressive.

At a recent event hosted by supply side platform (SSP) PubMatic, which is primarily focused on representing the interests of the publishers in part by making it easier for media buyers to spend money with them, IDC analyst Karsten Weide noted that global programmatic spend, including all mobile and online display and video advertising accounted for US$4.8 billion in 2013.

This is almost doubling to US$8.5 billion this year, US$14.7 billion next, and rising to US$38.1 billion by 2016 and US$55.3 billion by 2017.

At that point, if that figure is achieved, it will represent 59% of total global display ad spend that is bought programmatically. “Anything that can be automated will be automated,” the analyst noted.

And programmatic is evolving. Programmatic used to equate to real-time bidding (RTB), where you were effectively buying inventory in a blind auction from the exchanges, not knowing where the ad would appear, but having confidence that the right people, or perhaps to be more precise, person, would see it.

Now programmatic also embraces private marketplaces (aka programmatic direct). In these situations, the advertiser knows where the ad will appear and can pay a pre-agreed price for it (usually higher than in an RTB auction) in a forward market as opposed to a real-time auction. However, the inventory is still bought in a programmatic, automated manner.

For media buyers and their advertiser clients, it gives them planning certainty. They can put a certain type of inventory in place, with impression volume guaranteed, at the right time to support a launch. Because the workflow is integrated and automated, it’s faster, less prone to human error, and so easier and more cost effective for both the buyer and the publisher.

All of which has led some to speculate that programmatic is about to take over the advertising world (including TV, radio, print, etc.) and that media agencies are about to become disintermediated as big brands set up their own trading desks to buy their own media programmatically.

This would remove the need for all those expensive, clever young things in Soho doing all that research and planning stuff. Already, Procter & Gamble, Kimberly-Clark, Kellogg, and others have made moves in this direction.

Smarter and more knowledgeable people than me have argued against both points, sometimes driven partly by vested interests (that’s the media agencies). To cut to the chase, I think they’re right.

Firstly, as Steve Wing, once of the Guardian and now doing his stuff at CBS Interactive, argued convincingly at the PubMatic event, media buying has always been about data. So the idea that data is a dark art to the people who plan and buy media is flawed anyway.

For sure, programmatic may be cutting down on the paper/e-mail chain and the admin involved in buying ad space, online, on mobile, or anywhere else. However, it doesn’t negate the need for some intelligence and insight as to where the best place is for the ad to run.

In its first incarnation, when programmatic equalled RTB, that statement may perhaps have been harder to defend. But as programmatic evolves beyond RTB into private marketplaces, I believe it makes much more sense.

Because once you start re-attaching some value to where the ad appears and what you are prepared to pay for it to appear there, as opposed to merely who sees it, then the job of planning and buying – note the word “planning” in there – reassumes the importance that the media agencies would argue it has never lost.

As Paul Silva from trading desk MEDIAiq put it at the PubMatic event: “If P&G thinks it is as easy as licensing a DSP and hiring a couple of graduates and … that will lead to business growth, it is incredibly misguided and misinformed. …The programmatic execution piece has become very commoditised. If you look at where the most value is being created, you have to be good at ingesting massive amounts of data and determining value from it.”

Perhaps the last word – for now – on programmatic should be left to Sir Martin Sorrell. Speaking about the pace of change in advertising at an IAB event in London in October, he recently admitted: “The tech side is frightening. I’m not an engineer … so we are fumbling around in the dark. … It pays to be paranoid. You have to be paranoid in this area because you just don’t know what’s going to happen.”

Then a couple of weeks later, speaking at ad:tech London, Sorrell told delegates that advertising now was not just about “the mad men and women, but also, the maths men and women. … Brand communications have to embrace the left brain and the right brain,” he said. “The maths men and the mad men, the technical and the marketing.”

He may, by his own admission, be getting on a bit, but he still talks a lot of sense.

As ever, the immediate future of programmatic versus “traditional” advertising is likely to be less a case of either/or, and more a case of fusing the best bits of the two, combining tech with human insight to deliver more than either could achieve on their own.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .