Why LinkedIn Is Making Good on Programmatic’s Promise

October 4, 2016

Why LinkedIn Is Making Good on Programmatic’s Promise

Editor’s Note: This post was contributed by Lee Byrne, Head of B2B AdOps at LinkedIn.

From the beginning, the programmatic buying of advertising has always had great promise. And after many years of falling short of that promise, programmatic, I believe, is finally poised to deliver on its potential.

I’ve been involved in the advertising technology sector since about 2000, with stints at Lycos, Yahoo, and ZoomInfo. I also co-founded Bizo in 2008, where I was Vice President-Advertising Operations, and where I bought programmatic advertising at a massive scale for our customer base. In the process, I’ve had a front row seat to programmatic’s evolution, and it hasn’t always been pretty.

The promise of programmatic is powerful. With programmatic, you can use digital data to target your audience at the individual level across the Internet; you’re no longer guessing that your prospects are reading the New York Times or Advertising Age. With programmatic, you know that a member of your target audience is viewing the content on the page where your ad appears.

In addition to precise targeting, programmatic — in theory — is supremely efficient. It eliminates transacting and creating individual insertion orders with dozens or even hundreds of media outlets. With a single ad network, it’s conceivable for a marketer to efficiently and effectively reach her prospects wherever they travel on the Internet.  

Programmatic’s promise has fueled its growth. Next year, eMarketer projects that programmatic ad buying will total $27.5 billion (24% growth) and account for 72% of digital advertising in the United States. 

Frustrations With Programmatic Buying

Despite its growth, programmatic has often disappointed and frustrated ad buyers. Buying programmatic advertising has been noisy and messy and hard. Why? There are many reasons, but viewability, ad fraud, and brand safety are chief areas of concern.

At Bizo, I remember trafficking our first buy on Right Media’s Yield Manager platform. I uploaded the creative, layered on the Bizo audience data for our advertiser, set my bid at $1, and sat back and waited. Within minutes the campaign was delivering, and I was pleasantly surprised how easily it had worked. A few more minutes passed, and it occurred to me that I had no idea if it really was working. I had no idea what was really taking place — there was no way to be certain if the ad was viewable, if it was on a legitimate website, and if it was next to appropriate advertising and content. I have to admit, I got a little nervous about this whole Bizo endeavor at that point.

I encountered all of these obstacles first hand in the early days of buying display media programmatically.  Advertisers questioned us daily about where their creative was appearing. We told the advertisers that our blacklist was working wonderfully, and we certainly had their best interests in mind with every impression we bought.  We were hyper-diligent in digging into impression-level data to make sure we could back up our assurances, however we came to realize that we weren’t doing nearly enough.

The issues of fraud and brand safety were immediately apparent and worse than we imagined. As soon as domain level reporting (imagine not having that today!) was easily accessible, click fraud was easy to spot with CTRs over 3% on some domains.  Other domains, like wsj.com, started to show up that we knew were not selling media through exchanges.  The appearance of wsj.com was due to browser plugins loaded with adware, and we figured out how to block them. Managing our blackslists evolved from a monthly job to daily chore and then to a whitelist-only approach.  We didn’t make many mistakes, but we still weren’t comfortable.

Companies sprouted to solve these new advertiser problems. DoubleVerify and Integral Ad Science came out of nowhere and agencies insisted we allow them to track and block where and how advertiser’s creative were being delivered. In some cases, multiple ad verification vendors (each with its own technology) were used to track the same impressions. We ended up signing up with Pixelate, which was more focused on helping companies like Bizo understand what they were buying.

And then along came the viewability issue. It was another reason that, in buying programmatic ads at a massive scale, the Bizo team had to be hyper vigilant. We developed our own workarounds for buying programmatic advertising efficiently and effectively. We studied individual websites to make sure that ads were viewable. We whitelisted sites that fought fraud. We emphasized sites that performed well and paid a premium to do so. In short, it was a ton of hard work! And, all of this extra effort took time and money and made programmatic not as efficient as it could be.

Programmatic 2.0 Has Arrived

That was Programmatic 1.0. But now Programmatic 2.0 is here, and LinkedIn is leading the way. Earlier this summer, we launched Programmatic Buying for LinkedIn Display Ads. If you’ve experienced issues with programmatic like those I’ve mentioned here, i.e. efficiency, audience quality, viewability and brand safety - then programmatic buying on LinkedIn solves your problems.

With LinkedIn, programmatic buyers deal only with a single, premium site. Our audience is as strong as you can get, with more than 450 million professionals on the platform who can be targeted by industry, job title, seniority, business size, and more. There’s just one ad per page, it’s above-the-fold, and its viewability rate is twice the industry average. Bot traffic is minimal, because registration and log-in is necessary. And brand safety is essentially a given.

My only issue is that I just wish Programmatic Buying for LinkedIn Display Ads was around when I was buying ads for Bizo. 

Our new programmatic buying option is further reinforcement of our commitment to help you deliver the best possible results every time you engage with us. Learn more details about about our programmatic offering.

Topics