Paul Frampton is Managing Director at Havas Media

We are all well aware that the retargeting model is broken. Still, three years into an RTB world, advertisers are frequency spamming the same users over and over again.

Whilst there are some positive innovations in the prospecting space, the reality of first-party data integration is still nascent and will never scale to the same level as the US.

My view is that the next evolution required is a new model for prospecting consumers further up the funnel, namely whilst in the awareness and interest stage. This requires a different approach; an engaging and more contextually-relevant deployment than is present in today's remnant exchanges.

Exchanges and DSPS's have efficiently opened up access to both the big boys, but also the long tail, of inventory eyeballs. For me, that's where the next play will be, but it will be a content play and not a performance one.

To bring new customers into the funnel, content creation is critical but the problem is that content creation is difficult to do at scale today. The best content comes from independent publishers, specialist content companies, plus a combination of journalists and bloggers - imagine a world where you could curate and target content at scale through a platform like a DSP. That content could be created from scratch, or quickly tweaking existing content to make it specific for an audience need.

The good news is that there is already innovation in this category. One of the most exciting companies in this space is Roll Up Media who have a vision very similar to the above. They are building a platform that promises to bring vertical publishers, writers and brands together.

If you were in the US, you would likely categorise this as native advertising. Combining native advertising with a technology platform where created or tailored content is housed, and targeting can be as precise as within a DSP, and you have an operationally efficient way to "do" content.

Think of it as a marketplace for brands that want content. A brand (the buyer) puts out an RFP for relevant content to reach a variety of different audiences and a variety of sellers (independent publishers, writers, perhaps even publishing houses) then bid for the job.

If the same platform could then automatically optimise for SEO, and reporting analytics could surface which content brings which audience into the funnel and how long they play with it for, suddenly advertisers have a chance to achieve their ambitious content marketing objectives.

The upside for brands is clear. For publishers who only sell 60% of their inventory, it offers a fresh opportunity to monetise their audience. At the moment, that 40% will be sold off in bulk or transferred into an exchange for a low yield, but if that same "space" can be adapted to deliver this model, there's obvious value for the publisher.

Applying the best of the performance world (technology, efficiency, analytics, targeting) to the brand and content space will only encourage more brand spend which is good for the whole market.

The challenge is then to use smarter attribution to understand how a content strategy like this "greases the pole" for the performance channels that harvest that interest at the bottom of the funnel.

MediaMath, Invite Media & many others have built scalable models that aggregate mid- and long-tail inventory and make it easy to buy, analyse and optimise. The question for me is who will be the first MediaMath of content marketing?