When Facebook initially made its move to monetise onsite inventory with foreign datasets (foreign, meaning an advertiser’s first-party data), it divided opinion. We took a somewhat negative stance on the strategy.
With the launch of newsfeed ad formats being made available to RTB, does this require a new approach as to how the paid-for element of the Facebook platform will be leveraged, perhaps pointing to the future of RTB?
ASU Gold Rush
The initial FBX launch turned into a retargeting frenzy. Predominantly used to help drive greater cost efficiencies for businesses running existing RTB campaigns (FBX on numerous occasions was found to be unsurprisingly cheaper) and also to prop up the volume of click-based acquisitions that made the overall click-average-to-impression-conversions on RTB campaigns healthier.
Since then, Facebook has obviously seen enough data to inform its decision to open up more desirable real estate, such as the newsfeed. The fact Facebook saw higher yields from the ASUs in RTB is hardly surprising. By leveraging advertiser-side data, users are more likely to click and buy.
In turn, this drives greater bidding appetite—as well as leveraging impression bidding versus click bidding. Previously, social buying teams bid and bought on a CPC basis. Whereas now, trading desks bid and buy largely on a CPM basis.
Considering the CTRs are also probably higher (due to more relevant creatives matched to data sets), the eCPM and eCPC are likely to be much higher. Flipping it to the newsfeed was a no brainer.
Unlocking Greater Insights
An often discussed subject with regards to RTB protocol is the inability of a DSP to translate the information sent on a bid request. The current open RTB spec includes many parameters which are optional. For an exchange, or SSP, aggregating supply across 1000s of sites, obtaining and sending the same information on every bid request is somewhat of a challenge.
However, and according to Facebook’s developer reference documents, the parameters sent out on every bid request contains details which, a) can be extremely insightful from a reporting perspective, but also, b) enable more customised ways of bidding for inventory and building more advanced models on inventory.
Some examples of what FBX seemingly pass on a bid request include deep insight into page context, including whether it is canvas page, search page, event page, group page, photo page, home feed or messenger page.
Why does this matter? It means buyers will be able, in theory, to map out the associated performance metrics via the above values (via matching successful bid request IDs to the parameters sent to on that request to the corresponding success of that impression).
A DSP having FBX log access is key to this, but it means not only will advertisers understand the relative difference on reaching a user within a search page, versus group page or home feed, but that buying models and algorithms can be adjusted accordingly.
There is a preconception that because Facebook represents a single buying point that only one buyer should execute a performance campaign on it. The argument could easily be made that if you are able to ingest all the feeds, process it and model out bespoke ways of bidding for different impressions in different page environments, then you are adding value above and beyond simple retargeting strategies. Paving the way, potentially, for more than one buyer on FBX per client; the above modelling could be so unique, the prospect of bid conflict is highly unlikely.
Ushering in Upper-Funnel RTB
Back to the newsfeed. Facebook may well be on its way to redefining how RTB is leveraged. There is a legacy approach to how a campaign strategy is segmented: retargeting and prospecting.
Prospecting invariably is measured in the same way as retargeting (with perhaps a slightly higher cost-per-acquisition target), which is at odds with the principle of prospecting for new customers and users.
The newsfeed, made available to RTB, may start enabling the programmatic industry to provide value in the ‘upper funnel’ space which is really where prospecting should be fulfilling its objective.
The newsfeed is a canvas for what is essentially targetable and optimisable content marketing. The newsfeed ad units are essentially sponsored content. Whether FBX buyers are able to cookie-match third-party data sets (an Exelate or Bluekai model) or is strictly just pixels placed on advertiser’s sites, this is something ExchangeWire is unsure of.
At best, the newsfeed through RTB becomes a prospecting playground, targeting via rich third-data providers. At worst, it becomes a way of re-engaging existing customers, helping to develop a robust, ‘always on’ CRM strategy where the user is given something tangible: more relevant content where they most expect it. If the newsfeed is purely used as an acquisition tool, it will be a hugely missed opportunity.
Making Native Scale
The last thing that the newsfeed on FBX represents is a genuine example of ‘native advertising’ becoming scalable—through RTB. The only way that native will truly scale is when the execution onus is placed on the procurer. The data signals it kicks off are also more useful than clicks and dwell time.
The purpose of a content marketing strategy is likely content propagation. If this becomes a readily available reporting metric, then native advertising just got a lot more relevant to the buying community, with predictive optimisation to users more likely to share content.
So why should this not be extended to other platforms? Why would Twitter not do the same? The newsfeed ad unit is not a standard IAB unit, nor is the ASU. Will Yahoo! seize the Tumblr opportunity and make the ‘spotlight ads’ available through RTB? The RTB community have already moved to incorporate non-standard ad units into their workflow and the money has subsequently flowed.
This should be a message to all other prospective new-age social platforms. Don’t fight the money. Do what Facebook is doing and make your native scale by making it RTB’able.