As a longtime software developer and product manager, I value innovation. Unfortunately, innovation and easy, seamless operation aren’t related. Or if they are, it’s often an inverse relationship. After all, what distinguishes an early adopter is a willingness to put up with an incomplete, unevenly polished product simply because it’s new. Ad tech provides unsettlingly consistent examples. The whole industry didn’t exist two decades ago, but in that short time it has evolved and innovated constantly – criticism of ad technology never faults it for its slow pace. So in effect, whether we like it or not, in this industry we are all early adopters at all times; by the time one piece of tech stabilizes and is adopted widely, it’s no longer sufficient and the next new thing comes along. But these are integrated business applications we’re talking about, not the latest consumer gadget. Adopting new solutions is not so simple when you need everything to be connected and interoperable, when you’re trying to run a business at scale, for example. So for all its innovation and rapid change, ad tech quality and effectiveness often seems to be going backwards, at least from the perspective of actual users trying to run actual digital media businesses. That’s why so many people hate ad tech.
To say this poses a challenge for ad tech product developers is an understatement, which is why many in the ecosystem simply choose to ignore it completely. But this is exactly the problem we decided to solve at PubMatic. How can we deliver ad tech solutions to publishers that are at once innovative, high quality, flexible and adaptable to inevitable change? How can we provide a consistent, reliable, easy-to-operate platform that is also radically innovative and flexible?
Two years ago we set about answering these questions. In order to get a better handle on the problem and understand it clearly from users’ point of view, we talked to hundreds of people in the industry—day-to-day users and business stakeholders of all types. We investigated across all regions globally, both on the publisher and the demand side. While we saw a lot of variation in the feedback we received, there was a consistent pattern of frustration with ad technology that is either considered legacy and generic, or niche and difficult to integrate.
Frequently, “legacy” and “generic” were defined in terms of the traditional ad servers that were designed primarily for desktop display inventory, making them poorly suited to today’s multi-channel, multi-format, multi-device businesses. In the category of newer, niche solutions, they cited those that focus only on mobile or native, which may do one specific thing well but are difficult to integrate into a scalable enterprise solution and workflow.
We also took a hard look at our own technology and reassessed every piece of it, both individually and in the wider context of a full, integrated solution, and asked ourselves a few important questions. How completely does our tech address the full spectrum of modern demand channels, screens, and formats? How easy is it to use all together, to operate at scale and to integrate with other solutions? How adaptable is the foundation, and is it flexible enough to deal with whatever tomorrow will bring?
While we were proud of the products we’d built, we quickly realized that in order to solve the bigger problem we needed to rethink how our products work together and how other technologies can plug in. We needed to think of everything we built within the context of a complete, interoperable full-stack solution—one that was at the same time entirely modular, extensible and customizable. Pieces that weren’t compatible with this objective needed to be redesigned and rewritten, while others could be adapted and better integrated.
The main challenge was that everything needed to fit together seamlessly and yet also be decoupled and pluggable. Internally, we used the metaphor of a Lego kit. You can use all the pieces that come in the box and have a complete end-to-end publisher ad tech stack, including direct ad serving, indirect monetization, cross-channel optimization, mobile, native, video, and comprehensive advanced analytics for all of it… or you can rearrange and swap out pieces from other kits to develop endlessly varied, customized and differentiated solutions to meet your needs and easily integrate new innovation.
Developing a seamless full-stack solution that’s also completely modular is, of course, easier said than done, and it did not happen overnight. And I’m happy to say that with the recent announcement of our new unified, modular platform, PubMatic SEVEN, we released exactly that. It was worth the time it took because the investments we made in providing not just point solutions or “pipes” that work, but in refining the workflow and interoperability, the integration points and APIs, made a real difference in the experience publishers have running their businesses. And because the platform is entirely modular and extensible, we have a great foundation to innovate further and to make it easy for our publishers to integrate the innovations of others in the ecosystem.
The exciting part is that although the recent release of SEVEN is the culmination of a couple of years of work, this is really just the beginning. We see the pace of innovation and product development only accelerating in 2017. And because we have a completely modular, well-integrated full stack platform on which to build, we’re confident that future innovation will support easy operation and overall quality, rather than get in its way.