The paywall is not a panacea for the current economic woes being experienced by traditional publishers. In fact outside specialist and business critical content providers, like the WSJ and the FT, you are going to struggle to get anyone to part with their hard earned money for commoditised news. Some premium publishers are now betting their future on the subscription model - and in the process giving up on data and ad inventory to build out sustainable business models. There are a few European publishers railing against this prevailing trend. DMGT is a notable example. Its ad network layer, A&NY Media, currently offers advertisers a network of premium content sites with a user reach of over twenty-six million. In an in-depth interview with ExchangeWire, Nicholas Timms, Head of Network Products at A&NY Media, explains how a large publisher CAN build out a successful data-driven display business and why access to proprietary first party data and content is critical to achieving this.

Can you give an overview of the A&NY Media ad network offering – and how it relates to Associated's group’s ad business?

NT: The group has a huge ‘quality’ network (reaching 30+ million people), most notably made up of premium newspaper sites such as Mail Online, Evening Standard, Metro and Northcliffe Media. A&NY uses this to garner a comprehensive knowledge of the performance of advertising products, unsurpassed by other networks and unmatched by all publishers.

Rather than cannibalising existing press advertising models, A&NY Media has applied a new group model allowing it to compete against portals and networks rather than just other publishers. This model has also helped spread digital across the group, with A&NY Media developing new network products that existing sales teams can easily sell.

A&NY Media takes charge of all advertising models online rather than adopting the historic publisher 'wait and see' approach. Traditionally many online partners claim to be doing something for you, when they are actually doing something to you. And now we have developed Marketmaker: complete and thorough real-time audience insight and refinement; where the campaign finds its own success.

How important is having access to proprietary first party data and premium content?

NT: It is vitally important – for efficiency and insights. Our audience product operates across a huge breadth of editorially rich and vertically well defined content. It is unrivalled when as a publisher we have direct dialogue and first party relationships with our readers – ie, understanding a consumer’s net worth by the value of the properties they are trying to find, by the type of jobs they are looking for, by the type of articles they are reading, and to the searches they have made, or simply the type of editorial they are reading.

We have over 1,000 audiences that can be defined in this way – with upmost attention to the number of times a user must perform an event, and how recently that event occurred, before they qualify for the segmented audience.

The difference is also evidenced when we operate ‘lookalike’ modelling. We tag up a part of our advertisers website, ie a conversion page, and then index consumers who have reached this page against our 1,000 pre-built audiences to find audiences with the highest affinity before a campaign goes live, yielding in greater performance for advertisers.

Because of the transparent nature of how our audiences are built out of proprietary behaviours, predictive modelling acts as a ‘real-time panel study’, showing the interests and intents their consumers have exhibited across any of our group’s sites.

Our approach in modelling our reader’s behaviours has ultimately allowed us to significantly increase yields for our group publishers, whilst simultaneously increasing ROI for our advertisers.

Is this a massive point of difference for A&NY Media in terms of what they can offer advertisers and agencies?

NT: Three years ago we cut off third party ad networks buying our tier two inventory. We bought transparency and incremental reach to display schedules to compete against blind ad networks for DR business. Our development over the last two years has been in the application of technology to take greater value from first party access to data and content, increased expertise within our analyst team, and concentrating on offering granular insights on live campaigns.

Today, we have joined up different licensed techs to create a real-time audience optimisation platform to drive display efficiency dynamically. Analysts who manage our network campaigns can now make more data-driven decisions rather than assumptive targeting – resulting in greater results for advertisers.

Analysts are also tasked with simultaneously providing audience segmented conversion data that is unrivalled in the market, and leads to actionable insights in brand planning.

Let’s talk about the resource and technology A&NY Media has invested in. Can you give some more insight on the “intelligent” supply you have access to – and the granular targeting you provide across your network?

NT: Every ad network claims to have a NASA grade black box that powers dynamic optimisation of campaigns to the most valuable impressions. However, the black box is only as efficient as the data you feed to the algorithm before a decision is made to accept or reject impressions.

We licensed Right Media two years ago to take the role of our performance ad server, and the quantity and quality of data we feed the prediction algorithm consists of:

1. Environment data – Site level, channel/sub-channel level, contextual relevance of content

2. Audience data – Demographic, interest and intent audiences.

In this way, impressions are fed through to our private Right Media seat from the Mail Online (and other group publishers) pre-defined on a site level, channel level, contextual relevance and rich audience segmentation – the outcome of which is that Right Media’s prediction algorithm will now dynamically optimise between all these different variables.

For example, on an audience level, the algorithm will on a given campaign look at the conversion rates across 1,000 audiences and for example, rejecting an impression delivered to a female audience, or a sports enthusiast, but simultaneously prioritise inventory delivered to a high net worth individual or a business traveller.

This is what I call “intelligent supply” – every single impression is valued by it’s ability to provide a given advertiser with maximum ROI. The valuation process goes through on-the-fly learnings on both the environments and audiences that have converted most efficiently.

We have seen the greatest gains in a campaign looking back over conversion data by audience, rather than environment to identify the behavioural intents to focus delivery and drive ROI.

Has the data-driven approach been restricted to DR campaigns? Are there any case studies you can provide about how real-time optimisation delivered for brand advertisers?

NT: Cost-per-engagement partnerships have generated a lot of media exposure for innovation on brand pricing and content creation. Aside from the creation of content for clients such as P&G and Diageo, we were tasked with driving users to the hub to engage, a potentially risky strategy since our payment was being made on an engagement/action basis. Therefore the network’s DR products and expertise were throttled to the max to cost-effectively drive the most relevant users through to engage with the content hubs.

Our unique ability to deliver on user engagement in these brand hubs is powered by real-time audience optimisation, where similarly traffic drivers to these hubs are prioritised dynamically to audiences that have the lowest number of impressions required to drive an interaction – without requiring assumptive targeting before the campaign kicks off to manually target audiences that ‘feel right’. Insights from these campaigns are supercharged with understanding the conversion rates and individual engagement times across 1,000 pre-built audiences that interacted with our client’s brands.

Why haven’t other publishers, with the similar reach to A&N Media, built out their own ad network offering?

NT: There are some other publisher groups that I can think of who have a similar scale in first-party access to audiences. However, it is a difficult strategy to suddenly cut off third party demand for tier two inventory, not to mention the collaboration of different publishers within an organisation to create a group resource that if not managed carefully may conflict with primary sales channels.

Andy Mitchell, our group commercial director has driven forward our internal network strategy, and the products that have come out of this route are being utilised through the whole group – not just within A&NY Media digital sales, but also within Mail press display and classified, Jobsite recruitment sales, Metro display, London classified, and Evening standard display.

How is A&N Media using RTB and access to exchange inventory to deliver performance for its advertisers? Is it all about re-targeting existing users and augmenting reach?

NT: Firstly, I applaud supply side platforms for playing a pivotal role in the evolution of the market in offering publishers a non-threatening solution in bringing quality inventory into the exchange space to RTB buyers.

Our move into this space will be data driven – I don’t mean just buying inventory that matches an audience data set, but instead, buying inventory that campaign data has told us converts supremely well.

There is an incredible sea of aggregated inventory on display exchanges, and we have licensed Invite Media to access it. However the decision in scaling a campaign to exchange inventory will be done on an insight from running the campaign in our private ad exchange of group inventory. As previously mentioned, on a given campaign we can see which of the 1,000 pre-built audiences are driving efficiency, and therefore it’s a low risk decision to scale these channels on a campaign by campaign basis.

This differentiates our buying decisions on exchange space as we are not using a predictive ‘lookalike’ targeting methodology to choose which audiences to go and find on exchange space – we have already found audiences that drive efficiency on a campaign by campaign basis, so finding more of that audience presents a similarly low level of risk as advertiser re-marketing.

Data clearly underpins everything A&NY Media does - particularly, network optimisation and exchange-buying. Can you give some idea on the resource, technology and data infrastructure A&NY Media had to have in place before it delivered for advertising clients?

NT: We have licensed Audience Science for four+ years, however in the last two we stepped up considerably ensuring every publisher in the group (even those that are not part of A&NY Media as an inventory partner – eg Euromoney and DMG Events) had our behaviour capture tags implemented. We increased our breadth in our audience warehouse by working with 10+ publishers external to the group to give us intelligence on verticals that we previously lacked in – this brought us to understanding 30 million+ UK users in the granular ways Audience Science captures behaviours.

However, I do not believe it was until in the last six months when we made big back end developments to allow our various licensed techs to speak to each other in different ways that gave birth to predictive targeting and real-time audience optimisation – and these two models are driving efficiency and insights for our display clients.

Are analysts the key to the success of A&NY Media? Are we seeing a shift in skillsets in display buying?

NT: We have certainly not disregarded the importance in back end expertise at A&NY Media – network analysts, brand ad ops, and intern positions represent a third of our digital team (with an average age of 22!). However, the key to success is down to all parts collaborating together – there is a huge culture in the sharing of expertise and knowledge throughout our floor – between press and digital – and success is an outcome of the sum of all parts.

The shift in skills will be towards an aptitude for problem solving – choosing which technology to apply, which ad exchange to connect to, which audience to target...the market is evolving very quickly, and publishers, ad networks, agencies, and clients are all tasked with figuring out how to apply technology and expertise to make display work harder under more competitive conditions.

Is the “MarketMaker” the next evolution of the A&NY Media offering? Can you provide an overview of it – and outline what value will it have for advertisers?

NT: Market maker is where A&NY media’s offering has evolved to – market maker is already in operation, that comprises of the products and methodology that is operating our advertiser’s budgets. Our next evolution is to make all insights and knowledge on live campaigns readily accessible by our clients – which will be the market maker platform.

We have seen from our clients an increasing demand for intellectual property around campaign management. We are building a web based dashboard that will pull in live campaign data and present to clients comprehensive insights – from an overview of how analysts are re-locating budgets between different optimised channels, down to the granular insights in the audiences and environments that are driving scalable ROI.

The purpose of this dashboard is in line with the ethos at A&NY Media of complete transparency and gives clients full visibility, with maximum accessibility.

The platform will be ready by end of calendar Q1 this year.

AND’s owned-and-operated sites clearly have a lot of global traffic. What approach is A&NY Media taking in terms of how it trades and monetises its non-UK inventory?

NT: The Mail Online is at the front of our international stage, reaching 55 million uniques globally (20 of which are based in the US), and we have now opened direct sales teams in the US, Australia and Ireland.

For non-UK tier two inventory we are in the process of setting up links from our Right Media seat to agencies with RMX seats, where we offer all the products and granularity we use in the UK – separating our inventory out by first party definitions of behavioural audiences, contextual relevance of articles, and content channels. And uniquely, we are able to give agencies real-time audience insights on their campaigns that are targeted ROS to our inventory, without the need for them to target that individual audience in a “target and test” methodology.

With demand growing from DSPs and trading desks, are we likely to see A&NY Media launching a private exchange to make inventory available through RTB?

NT: We have in essence created a private exchange; however the invite is limited to agencies for our UK inventory.

I appreciate Right Media is not yet RTB-enabled, but this is not currently a limiting factor and offers price reduction and flexibility for agency trading desks that connect to us to choose the inventory they require, at the market price.

Our flexibility to opening our supply to agency trading desks remains a focus for A&NY Media – and we want to develop more models of commoditising inventory into our private exchange so that the value of an impression is wrapped with transparent layers of audience segmentation before it reaches the buyer to influence greater demand for more valuable audiences.