Now I know some of you out there are going to give me a verbal slap down for coming up with a new acronym - but the Publisher Trading Desk (or PTD) could actually be a viable and profitable entity for the sell-side. A PTD is not all that dissimilar from the functions performed by an ATD, DEM or any other data-driven intermediary display buyer. The publisher would effectively leverage its own data to perform buys across exchange inventory.

But aren’t publishers already doing this? Granted they are performing some sort of audience extension via the likes of Admeld, Rubicon, and Improve Digital – but there is still no standalone PTD solution on the market. Why should publishers build out this type of offering? The reason is simple: they own proprietary first party data and have the supply. And they should be taking full commercial advantage of the data-driven eco-system that we are now working in. Let’s examine some of the challenges and opportunities around such a strategy.

Will We See Lots Of PTDs On Agency Plans?

This PTD strategy won’t work for everyone. Large publisher networks like A&NY Media are already doing some interesting stuff around audience extension – and they could well be the first to market with a defined PTD offering. I think some other big European publishers like the Guardian and Axel Springer could also do it. But the real opportunity lies with niche publishers with valuable audiences. Let’s take the examples of Europe’s two heavyweight financial publishers, namely the FT and Handelsblatt.de. Both are sitting on unbelievable first party data. Both have high-end financial audiences. Both attract the best brands. So, why shouldn’t they look at leveraging their own data to scale across dynamic and managed supply?

Doesn’t It Conflict With Our Existing Sales Channels?

This is an obvious obstacle to establishing the PTD. A clear demarcation of the premium and data-driven operations needs to be established. One sells high value integrated campaigns across their site; the other is a data-driven offering focused on performance. The first thing Handelsblatt, the FT and other publishers need to do when launching their PTD is make sure they are not devaluing their existing premium proposition. Most of these sites are near to selling out every month anyway - so the PTD will be focusing on retargeting audience across sites with a similar context. It's easy then to avoid sales channel conflict.

Should We Really Compete With Agencies?

I see no real conflict here. The FT and Handelsblatt are still going to work with agencies. These publishers are just augmenting their current offering. There’s little chance that either will be selling their data directly to agencies or third parties. Outside of the big five holding agencies the PTD could sit happily alongside ad nets, DEMs et al. Some legacy relationships will already exist with marketers at brands so it won’t be difficult to sell in the PTD solution directly.

What About People, Processes, And Company Structure?

If there was any hindrance to launching a viable PTD business this is it. The default sell-side approach to the advent of exchange trading has been a defensive one. While the buy-side has been busy developing sophisticated strategies to trade, publishers have been trying to figure how to work in this real-time environment. Putting inventory on exchanges and SSPs and decoupling data from media to sell it to third party vendors is interesting, but the real opportunity lies in trading. I don’t think publishers realise it yet but everyone on the buy-side is trying to lock down supply and data. Crtieo is doing deals with top tier publishers to protect its access to inventory. AppNexus tech is providing real-time access to Microsoft inventory. Group M Marketplace is talking to everyone. And other big agency groups are trying to lock-in dynamic and managed supply for their own private exchanges.

Publishers are in pretty strong position – as the industry becomes even more data focused. But even if they were in a position to build out a PTD, they simply don’t have the skills or in-house resource to make it happen. So what do they require? Firstly, they need some smart optimisation/ad ops people. Head hunt a senior ad ops specialist, preferably from an ad net background; then hire some smart analysts who’ll be able to run exchange campaigns. Publisher acquisition is going to be important so you’ll need resource for that too. And let’s not forget the sales people.

The Cost Isn’t Going Down Well With Upper Management

But this will never fly with senior management. It’s hard to justify the cost of circa 500K for an untested start up unit like this, but there is plenty margin to be made here. The demand form tier two and three agencies as well as direct marketers would ensure profitability in a short period of time. The vertical PTD could offer the kind of performance and audience targeting that other buy-side entities could only wish for – mostly because of access to the first-party data.

What About Technology? Do We Need To Build Our Own?

No. The tech already exists. You could either use a DSP to power your PTD – which is the standard buying strategy for buy-side exchange traders. Or persuade the SSPs to white label their solution, letting you buy across different inventory source. Not sure how possible that is. But if the volumes were large enough I’m sure they would roll-out something to service the publisher’s buying requirements. The SSPs could also help to segment audiences for targeting.

The PTD is already here in an understated way – but I think we are going to see publisher trading desks openly competing with buy-side entities in the exchange space over the coming months. In a market where lines between traditional vendors are quickly disappearing, there is an opportunity for publishers with proprietary data and inventory to deliver a competitive solution to the market.