By: Eric Picard and Brad Beal, Pandora
The IAB Tech Lab has done something historic, important, and worth discussing. It is likely that if you’re reading this post, you’re somewhat familiar with viewability, one of the most important issues facing the digital advertising space today. Below we’re going to cover an exciting step with a change in the way that viewability is measured and vendors interact with mobile applications. This issue is so important because – despite the complexity of supporting numerous viewability vendors on the web for any publisher, advertiser, agency, or vendor in the space – the challenge of supporting multiple vendors in mobile in-app environments has been nearly impossible to solve so far.
The proliferation of vendors with their own software solutions that need to be supported in-app was causing complexity. Publishers were finding it difficult to maintain their apps and stay up to date with Software Development Kits (SDKs), and, there were discrepancies between vendors that were hard to understand and debug. Buyers were finding that inventory supply was becoming fragmented since not all publishers supported all vendors, which either restricted their options in media planning or forced them to work with different measurement partners across different publishers. And, as an industry, we frankly were not collaborating as much as we should have been to end the proliferation of fraud.
Today, the IAB Tech Lab Open Measurement Working Group is announcing the beginning of a new era for the industry, and for the organization itself. As opposed to “simply” issuing guidelines or standards (hard enough to do), we are entering an age where complex shared software frameworks will be centrally managed, with collaboration across multiple companies. Nothing like this has been done in the advertising industry before, and we’re proud and excited to announce the IAB Tech Lab’s new Open Measurement SDK (OM SDK).
You likely understand that buyers of media require viewability vendors to verify that impressions purchased in digital media were viewable to a human. These vendors also support numerous other types of analysis, such as how the human paid attention to the inventory, the quality of the inventory and protections against fraud, etc. And it makes sense that, in order to get access to the data needed to verify this information within a mobile app, some special software from a trusted third party is needed to be integrated into the mobile app.
What is unlikely to be obvious to most people in the industry is that integrating software code from a third party into the code of the mobile app itself is fraught with all sorts of complexity and risk. Mobile app developers must incorporate a Software Development Kit (SDK) into the code of their app to support the vendor’s tracking and analytics. From a resourcing and time perspective, integrating with just one SDK can be costly for an organization. The process requires not only coding, but also significant testing on an ongoing basis to ensure the stability of the app. Couple this with regular SDK updates, essentially repeating the cycle, and the costs become prohibitive. And even then, SDKs can and do cause instability.
It also may not be obvious that massive scale mobile publishers like Pandora face challenges that might not be faced by smaller publishers. Any SDK can cause a variety of issues, ranging from memory leaks, heavy battery consumption, apps hanging, or even crashes. Now the vendors in our space are really great, and the number and frequency of problems an app developer might experience impact a small percentage of users. For smaller developers, a tiny fraction of users having a bad experience may be acceptable in order to support the viewability vendors required by media buyers to unlock spend.
As a company with nearly 80 million active users, most of whom use Pandora’s mobile app, the instability caused by integrating just one SDK into our app can cause millions of our listeners a month to experience crashes or odd behavior in their app experience. Since the code is owned and operated by a third party, we have no way to debug or easily fix issues when they occur. Incorporating SDKs from multiple viewability vendors simply is untenable for a company operating at the scale we operate at.
That said, media buyers demanded that we support every vendor – not just one. Advertisers understandably want the ability to use the measurement vendor of their choice, regardless of the limitations or problems this may cause for a publisher. And, they want to do so, without having to wait for a publisher to integrate a vendor’s SDK and test it to ensure stability. This created tension in the ecosystem that was building last winter, and reached a boiling point in the spring.
So, when Alanna Gombert, the former GM of IAB Tech Lab, reached out to us for support with a new approach, we were extremely excited to dive in. The solution will rid us of the SDK bloat issue that causes instability, maintenance issues, and negative user experiences across all publishers in the space. And, it will also solve for the frustrations of advertisers around lack of options and slow time to market. Finally, it will enable a common industry standard for the collection and reporting of viewability data, benefiting both publishers and advertisers alike by reducing inconsistencies in reporting.
So, what is that solution? How did we cut this seemingly impossibly complicated Gordian knot of technical, business, and financial issues?
The IAB Tech Lab, working across the entire complex ecosystem of vendors, publishers, media agencies, technology companies, advertisers, and trade organizations, has driven the standardization of viewability and verification measurement signals via a common framework – the Open Measurement SDK. The OM SDK will provide participating measurement vendors access to the data needed to support their verification and analytics needs, and they, in turn, will focus on supporting the OM SDK in place of any proprietary SDKs, and alternative API-based viewability measurement. In short, we’re creating a single SDK implementation that works for all vendors. The amount of cat herding that needed to be done to get this massive hairball of a problem to resolve is frankly impressive. And the speed at which the IAB Tech Lab and all the various constituents needed to operate at was unprecedented.
We’ve been part of many IAB taskforces over the years, and this was unlike anything we’ve seen as an industry. The level of technical capabilities that the Tech Lab has developed, the willingness to take on governance of a complex open source project, and the willingness of the vendors to operate with good will and in the best interest of the whole industry is really remarkable. And the speed of getting this project to market has been really heartwarming – because as a large publisher, trying to balance the requests of our media buying customers with the user experience needs of our listeners has been a real challenge.
This project has been actively developed for over a year and is now available in a Limited Release. Viewability is only a start that will make way for additional measurement capabilities within the OM SDK. But we, at Pandora, feel strongly that this is the way forward for the whole industry, and we’ve devoted significant product and engineering resources, including dedicated engineers, to work on this project as a driver in demanding and integrating this single OM SDK. Other companies in the Open Measurement Working Group–including DoubleVerify, Google, Integral Ad Science, Moat, and Nielsen—are also contributing code, resources, and time to this project and to the authoring of the code and we welcome our newest member comScore and look forward to their contributions. Now that the first version of the SDK is available for Limited Release and review, we’re working to integrate it into the Pandora Mobile App.
The collaboration and good will we’ve developed as a working group is unprecedented and makes us proud to be part of this effort. Every once in a while, the fact that our industry is made up of people who care about doing the right thing, and making this work becomes very obvious.
ABOUT THE AUTHORS
Eric Picard, VP of Ad Product Management, Pandora
Eric Picard is an advertising technology expert with more than 20 years of experience in the ad technology industry. Eric is currently Vice President of Advertising Product Management at Pandora Media. Previously Eric led product strategy for the DSP business at MediaMath, which acquired his startup Rare Crowds in 2014. Rare Crowds built micro-segmentation technology for finding valuable highly targeted audiences, at scale, and was a spin-out from TRAFFIQ, where Eric was the Chief Product Officer.
Previous to this, Eric was at Microsoft for six years, where he had a variety of roles in technology strategy. He helped set the company’s long-term vision for advertising technology, emerging media product planning, planning for incubation & research teams, and designed next generation advertising products. Eric managed advertising technology acquisition planning from 2005 through 2007 – including reviewing hundreds of deals, and participating on deal teams for numerous pivotal acquisitions, such as Massive, Screen Tonic, adECN, aQuantive and others.
Eric started working in ad technology as founder and director of product management at Bluestreak from 1997 to 2004, where he oversaw advertising products, such as third-party ad serving, ad analytics, rich media products and led development of many company technologies. A frequent speaker across the industry, Eric has written about advertising technology since 1999.
Brad Beal, Sr. Product Manager, Listener Ad Experience, Pandora
Brad Beal is a Sr. Product Manager on Pandora’s Listener Ad Experience team and responsible for the company’s Display, Rich Media, and Video products across all platforms. In this role, Brad also focuses on measurement and ad quality. Brad started his career in advertising at Horizon Media and has also held a variety of Product roles at AT&T.