Ad tech is a funny business. Experts and onlookers alike have been hammering the concept of ad networks for the better part of three years, yet when Facebook seem poised to launch one, it has many salivating over the prospect. Just last week, we saw GigaOm fuel the speculation that the great FB ad network is imminent, but what can we expect? What will it look like? Why will it be valuable? More importantly, why will publishers want to participate?

Exporting Native?

Much discussion has centred around what would the FB ad network’s ads look like? There has been huge speculation about whether or not there will be a 'native' solution. Many believe the rise of consumer web monetisation has been predicated by native solutions, and many are investing in its future. But how can Facebook export a native solution? Would sponsored stories sitting within an MPU be a strange user experience? Would it outperform standard ad units? Perhaps as Facebook becomes more and more synonymous with the user’s general internet experience, then these ‘native’ solutions would not look out of place?

Ad Net or FBX: How Facebook Extends

An important consideration of how this new product plays out will be whether it will be a managed service/typical ad net solution or whether Facebook simply aggregates publishers onto their platform and passes the inventory into RTB channels, overlaid with their data.

Initially, it will likely be the former; and it is not unreasonable to think that Facebook Ads Manager could encompass basic campaign management tools in the same way the Adwords UI did, and does. Allowing buyers to have some degree of control over the way ads are managed and optimised. Would they create an internal auction model? Sure, why not? They’ll need to offer all they can in terms of yield to make publishers want to jump in.

For large scale advertisers/agencies, they will inevitably want more control over the execution beyond some simple interface, they’ll want complete integration into whatever stack they’re assembling. Its not the first time an exchange has aggregated supply beyond its own O&O sources.

To make FBX a viable solution for all parties, it needs to leverage FB’s vast first-party data pile. As soon as Facebook start pumping out bid requests out by the boatload, with rich user data appended to said requests, this then will open up a huge privacy and political problem.

Who will FB trust with any cookie sync? It’s unlikely they’d ever devalue or commodotise the only asset that differentiates its model from any other behavioural based ad network -- namely their audience data. If FB starts passing bid requests appended with males, aged 21, living in London and a huge fan of Ford (based on some strange obsession with liking every Ford Facebook brand page), what happens when a major trading desks fulfills the impression, serves its ad, with its own pixel server, dropping a cookie on said user? It's possible that the trading desk could continue to retarget that user without needing to pay a penny to Facebook (of course until they need to refresh the pool). This could create some awkward conversations with senior executives.

Facebook have the ultimate multi-screen publisher trading desk to start scaling brand spend

If Business Insider rumours are true (and that's not always the case given their recent track record), then FBX is expected to grab 50% of all RTB spend. Facebook clearly know what they’re doing. For some bizarre reason people think this is a great scenario,  including publications that depend on performance budgets to run their "expert" editorial content.

Facebook has an enviable user base. It has nearly 1 billion email addresses on record. By working with a solution such as Adtruth, it could, safely and securely, map individuals across every screen. Now imagine the story FB execs could use in their next meet with General Motors Ford.

Not only can Ford supplement their inevitably large investment into content on Facebook, by paying through the nose to actually get that content in front of eyeballs that resembles something beyond 10% reach, but they can also engage precise audiences across multiple screens, throughout the day. A storytelling nirvana, if you will.

Technology exists for Facebook to identify individual devices (which have previously been mapped to specific user email addresses and the open graph data associated to that user), and buy in real-time inventory across any device. This could even be sequential and/or consequential.

Now, what else does Facebook have up its sleeve? It potentially has the answer to programmatic premium for brands. Inventory acquired in real time. Targeting precise audiences with very rich user data. It also has the measurement solution. Whether it is Nielsen’s OCR product or Datalogic’s solution to track offline sales, Facebook has the ability to target, optimise and measure on metrics and measures that is relevant to the very brand advertisers that they have been courting ever since they had to get serious about making money beyond DR focused ASUs.

Why this probably won’t happen

One word: publishers. FBX has been an incredible success. Why wouldn’t it be? Retargeting campaigns now have an even more bountiful supply of inventory, invariably cheaper than the market average. It is too early to say whether new budget is coming to the market, or if it's a case of budgets being shifted around. It is fair to say that publishers have been the biggest losers with the rise of FBX.

If FBX is more cost effective on DR metrics, why wouldn’t an advertiser invest in this channel versus what they may believe are over-priced, premium private marketplaces? The question a publisher needs to ask itself is whether they want to continue to support a business sucking budget out of the media plan? A business whose ad network solution is built on data, which many of these publishers are likely contributing towards.

This could well be a false economy for publishers. Why would it deliver anything above and beyond the options currently available to publishers in terms of monetisation? Time and again, it has been proven that third-party DOES NOT scale well for DR campaigns. Granted Facebook’s data is richer, has more depth to it, but how would a Facebook ad net solution compete with existing options? This will not be an AdSense killer.

Like it or not, context still rules the roost. Facebook won't be killing Google’s business anytime soon. So, if yields are unlikely to be any higher than average, would a publisher rather support a more open ecosystem, or support a business that is slowly killing theirs? Time will tell. The only ace up Facebook’s sleeve is potentially the offline measurement piece, but even then, holes will doubtless be found by the more discerning data-focused clients.

Becoming the go-to agency alternative?

To make the FB ad net really scale and provide the win shareholders are really hoping for (the real motivation here), Facebook needs to become a viable buying partner for major brands. Facebook would need to become an agency alternative. If the brand’s audience lives on Facebook, and if Facebook has the means to target this audience, community etc., across device, across the whole web, doesn't this become a credible alternative to any buying service of an agency?

It likely won’t happen. Google has arguably felt the effect of agency distrust in certain markets, particularly in Germany. Would Facebook make the same mistake by alienating agencies? It depends on how important owning more of the media plan becomes crucial to share price performance.

Of course all of this is pure speculation. The Facebook ad network might well not have any significant impact at all. One last thought: Facebook owns social data, it knows the interests and motivations of the entire consumer web. It does not necessarily know about intent -- the real time intentions of users. Why not do a deal with Amazon, bringing interest and intent, in real time, across every screen, to every major marketer? This might appease the analysts on Wall Street who seem to be in a perpetual state of despair at the Facebook share price.