Prevention Is The Best Medicine For Optimizing Programmatic

Updated: Oct 16, 2018

Article contributed by Matt Prohaska, Programmatic Advertising Director, The New York Times

So I got to attend AdExchanger’s great Industry Preview conference last week (thank you again, John and team, for having it in TheTimesCenter so I only had to walk 100 feet in the blizzard). Our own Nick Bilton, NYT columnist and lead writer for our Bits blog, gave the opening preview. Part of his talk covered technological advances in healthcare, not focused on treatment but on prevention – e.g. next generation FitBits that involve digital tattoos or implants to track your own real-time condition better digitally. Makes me want to restore that domain I took out five years ago:, yet another decent idea way too soon.

Listening to Nick got me thinking about how healthcare relates to a lot of our daily lives in the programmatic direct/premium business caring for the health of campaigns, whether you’re at a publisher, agency trading desk, indy DSP, or client. Managing mutual expectations, pre- and post-campaign launch, takes much more time than it should. Part of the challenge is understanding projected win rates on bids, and therefore media spend. We have noticed the two most important variables needed to understand are audience overlap and bid ranges.

We need an audience discovery tool. Several months ago, we were proud to launch a new client on our private exchange (the first publisher-direct deal with this savvy desk happened to be with us) with what we both thought was the right price in the right inventory to get them the volume they were looking for. Two days into the campaign we noticed their pass-back levels over 90%.  We thought it was a technical issue at first until we asked the desk who mentioned that at the last minute they added a layer of client 1st-party targeting where only consumers who had been to the client’s site in the last seven days would be included in the cookie pool. Needless to say, this would have been helpful to know ahead of time to manage expectations around audience overlap. It seems like an exchange could build a simple Venn diagram showing cookie pool size and how many are found in a standard 30-day period on a publisher’s site to at least guesstimate the potential audience and suggest where bidding would need to be to drive X impressions in a month. Which brings me to the second need….

We need a price transparency tool. It would be as simple as what Overture (formerly, 1st paid search service pre-Google for you youngsters out there) built back in the day for search where you put in a keyword and it told you what CPCs everyone was bidding at to know where you needed to be to get in the 1st position, 1st page of results, 2nd page…. Not sure how many other publishers do this, but we’re having to create visibility of various clearing prices (can’t get into too much detail) to give some guidance for our sales team and buyers on where to start our floors and get volume desired. This takes too much time to manually build today and seems like something major exchanges should be able to provide.

Even though the overwhelming majority of programmatic direct buys are non-guaranteed, with buyers and sellers knowing that going in, it would still be nice to know about how much will be delivered, so buyers can balance margin and scale while publishers can forecast properly and setup campaigns better for success right from the outset.

I loved Chris Rock’s stand-up years ago (“Bigger & Blacker” special – not linking to it, as it has just a bit of salty language) talking about the motivations of the pharmaceutical industry. Seems like we could save a lot of time and money if we could invest a little more upfront with each other, while pushing our shared platforms to help us with more mutual transparency. It would be good for our country on healthcare and good for our programmatic industry.

Matt Prohaska is Programmatic Advertising Director at The New York Times. Look for his column on the fourth Thursday of every month.