The Evolving Agency Model: Paul Frampton, MD at MPG, Discusses How His Agency is Using Data-Driven Strategies to Deliver for Brand Clients

Paul Frampton is Managing Director at MPG & Media Contacts. Here he discusses how his agency employs data-driven strategies to inform brand message/positioning for clients, the vagaries of oversupply of data and technology in the digital ad space, and the continued importance of the “human element” in the age of automated buying.
MPG Media Contacts seems to be transitioning into more of a brand/strategy-led agency? What is behind the concept of meaningful brands?
MPG has always been renowned for its planning heritage and Media Contacts for its data-driven performance expertise. When we brought the two together, we created a planning-obsessed agency with end to end digital capability. We have long believed that the brands that make a positive difference to peoples’ lives are those which enjoy greater business success (we’ve actually tracked this against company share price to prove it).
They have greater meaning to peoples’ lives because they offer something greater than their core purpose, service or product. We contest that the role that brands play in people’s lives must evolve, which is best illustrated through a brief history lesson: 1st generation brands were functional – focusing on convenience or price while 2nd generation brands looked to stand out from the crowd by differentiating themselves. We firmly believe that in order to succeed in a hyper-connected, hyper-transparent digital world, the 3rd generation of brands must make a genuine difference to people’s lives, i.e. demonstrate clear, unquestionable personal benefit.
Do you think there has been too much of a shift towards automation and “the science”? Was this the catalyst towards your recent slight repositioning?

It’s fair to say the balance of the conversation is uneven right now. The core tenets of meaningful brands are human understanding and connected thinking. For us at MPG MC, this is as applicable to the digital world as it is to the traditional media world. The danger with digital is that with an over-supply of data and technology platforms, we forget the human element. Digital is the richest medium that has ever existed and can fulfil any marketing objective, brand or response. At its heart, it provides the opportunity to understand real human behaviour (rather than rely on claimed behaviour from questionable, incentivised surveys) and connect up all the dots. The science of algorithms and attribution models are only of value if: 1) they are connected with all media touch points and 2) underpinned by a deep understanding of the audience. I’m a huge fan of a real-time world where 1st and 3rd party data can be combined to inform when, where and at what cost to compete. What has to come first though, is a clear strategy of who you are targeting and what their motivation to act will be.
Our fiercest belief here at MPG MC is that the rich insight we generate from data is the gold dust that informs idea generation. On top of our investment into content, mobile, web-build and conversion optimisation, we have created an Ideas function which works closely with our strategists and insight directors to curate bigger, better, cross-platform ideas. These ideas are often arrived at as the result of a digital insight that is then pushed back down through our business and brought to life through our cross-discipline experts.
There is a lot of talk in this industry about connected thinking, and it has become a buzzword that many don’t truly understand or offer –exploiting our data in this way is at the core of what real connected thinking is about.
MPG has always had strong data-driven qualities. How do you intend to fuse this with initiatives such as meaningful brands?
The most meaningful brands are those that tailor their communications and use data to achieve that. We deliver that through web analytics, user experience, conversion-optimisation, DCO, our DSP strategy and many other oft-inaccessible acronyms. But for these tactics to be effective, you first have to understand what the audience wants and where they want it.
The industry would do well to differentiate strategy from tactics, and then again from the platforms that enable these tactics. We have invested in Artemis, business intelligence tools and strategic partnerships to facilitate better data collection, organisation and analysis. These are necessary foundations to enable insight to feed back up to our planning teams, but it takes bright, client-aware people to identify the relevant nuggets.
How are agencies needing to adapt and refocus themselves in order to compete with regards to the changing data and technology landscape?

Data has always been important for a media agency; now it’s business-critical. Agencies need to be able to collect, organise and action data to better understand what works. For us, attribution is the biggest challenge marketing has always faced (Wannamaker’s quote about knowing 50% of advertising was wasted but not knowing which half was back in in the 1850’s!). We can employ econometric modelling, our DMP (Artemis) or Google Analytics to find the answer. At the end of the day, the platform is not the hero. We talk about attribution being the interrogation of which variables drive performance (media, creative, channels, sales points, products, regions, customer experience) not just about asking whether display receives more weighting than search. Data should be viewed as a strategic asset that can drive value back across the full marketing cycle; acquisition or retention.
However, it’s also critically important to understand how far your data remit should go, and I sincerely believe that some agencies are confused about this, to their detriment. We are not butting-up against Deloitte or Accenture but we should collect and fuse data sets together in interesting ways to drive improved targeting. With the growth of RTB (10% of display in the UK already and growing rapidly), the portability and agility with which this data can be taken in to an exchange environment will become the differentiator for agencies. It baffles me that agency groups are creating separate data agencies. If data is your biggest strategic asset, then it needs to be close to the planners and traders, not siloed off in a separate building. After all, I don’t know of any media agency that outsources its traditional research department. Specialist skills are imperative, but then they are in every discipline with digital. Connecting them is the artform.
How does the collaboration with Havas work? How central is Artemis (and other agency’s interpretation of this) to agencies in the future?
We have had Artemis for 15 years, so the talk of DMP’s is not at all new to us. We invested early in this space because we saw the need for a single repository for all digital data. We have recently opened Artemis up to accept offline and 3rd party data. To me, this is the critical piece in the jigsaw. Adobe and GA provide clients with solid analytics platforms, but they are unable to account for the chemistry of how traditional and digital channels work together. Agencies need to use their DMP strategy to aggregate all digital data, but importantly also to fuse on and offline data with social and customer data. Our current development is focused on integrating second-by-second spot broadcast data with search data and social data like Twitter. We are already beginning to create some early correlations between TV spend and the impact on positive sentiment and brand search. Brands can rarely do this piece alone.
You penned a deal with Audience Science for the provision of a DMP. What role does a DMP actually play across an agency? How does the agency engage with this type of technology?
It’s no longer enough to have the ability to select the inventory we want, we need technology to guide that process and ensure the most relevant inventory is selected. This is best achieved via combining 1st, 2nd & 3rd party data. When combined, it creates understanding of whether an impression is worth buying or not. We use two platforms to fuse these data sets.
The first is Audience Science, which we treat as an audience segmentation and planning tool. AS allows us to combine multiple sources to create cookie-level profiles. This anonymous information means that we can divide users into specific segments based on on-site behaviour, profile users based on shared characteristics and build look-a-like targetable audiences. The second platform is Artemis, Havas’ proprietary audience analytics tool. By analysing user campaign interaction rather than individual impression events we can understand how overall campaign exposure leads to business results.
Across other agency groups, where there is perhaps a more prominent holding group function with regards to media buying (read: trading desk), what role does the agency perform in these groups long term? Do you foresee MPG being in the same position?
At MPG MC, we have always believed that it is important to have RTB/trading capability in-house so that value can be truly realised from connecting audience-driven planning with technology. The problem with trading desks in many of the big groups is that they are very separate to the agency OpCo’s and their mandates are often not welcomed. We are anti building our own network, purely for selfish commercial reasons.
Our role is to drive improved performance for our clients through clever application of data and technology. We have recently been trialling a number of DSP platforms and are migrating towards a more trading desk type model using multiple DSP platforms. As addressable advertising extends from digital display out to TV and outdoor over the coming years, we believe that this capability needs to sit within the agency. Ultimately it is trading; it represents a more efficient way of trading that requires deep understanding of data and tech, but it’s still trading. For me, technology and platform development should reside at group-level for efficiency and focus, but trading must continue to be aligned with planning and be close to the account team.