Joe Zawadzki is CEO of MediaMath, a media trading solutions provider.
AdExchanger.com: When “last we left” MediaMath, the New York Times had just featured you. What’s been happening over the past 9 months? Any updates you can share about traction, product line, etc.?
JZ: We’ve been fortunate to continue our two-year consistent growth trajectory, despite the broader economic challenges in the market. In fact, the greater focus on ROI accountability in today’s market has actually accelerated the adoption of our platform (a trend we didn’t call when we started the company). As a profitable firm, we’ve been able to “lean in” to this opportunity while others have pulled back, building out an already great team of smart, passionate, and nice folks (a rare combination, imho!).
In Q1 of this year we released our SaaS media trading platform that allows our agency clients to opt for either the turnkey results of our managed service offering, or a self-service and whitelabelled offering giving them the tools, technology, and training to do media trading in-house, informed by the MediaMath lessons learned from the monthly execution of millions of dollars in spend and billions of impressions bought.
It’s been really fun, although it takes until Sunday morning for the kids to warm up to dad again.
Any insights you can share in online display advertising such as overall eCPMs, CTRs, category trends, etc.?
We’re focusing on making sure publishers don’t regret it by bringing high-quality, Fortune 500 advertisers into the spot market. A healthy spot market paying fair value on an eCPM basis for good audiences on good publishers, benefits the entire ecosystem and encourages more supply and demand to enter the system.
Is Media Math a services business or a technology business?
Proudly both. This market is a ridiculously dynamic one: over a dozen exchange entities ranging in sophistication from impression-level bidding to fairly manual UI processes, all focused on growing supply; third-party data sources that are transforming media properties into information marketplaces and blurring the line between media and audience; agencies, media companies, networks, publishers, and advertisers all trying to figure out their place in this new ecosystem.
To innovate in this market requires having both the right tools and the expert practitioners that know how to design and use them. To us that meant building a platform that could integrate with all exchanges, provide turnkey campaign setup for our clients, deliver best-in-class optimization, and provide campaign reporting and insights. But when we started two years ago, there were no practitioners around to use and refine that platform. So we did it ourselves. It’s hard to imagine building a media trading platform without actually getting our hands dirty using all the existing exchanges, ad serving technologies, and analytics platforms on behalf of real life client goals and constraints.
For the past two years, our managed service team has been responsible for delivering results, over hundreds of campaigns and tens of millions in media spend. That makes our solution “real world” and cuts through the vapour out there. In parallel, our SaaS development team distills this learning into the shared technology that enables both our internal traders and our agency clients to succeed.
It turns out the dual service-and-technology model fits well with the two types of agency needs we see. Some are looking for us to solve this problem, now, i.e., generate immediate client returns, provide insights and visibility, and work within my existing business model. Others are asking to “provide me a platform” that will allow me to build out an internal media trading discipline they are looking for a technology solution. In many cases, we see agencies starting out as the former and migrating to the latter.
You call yourselves ‘digital trading experts’. Can you give us an example of what a typical MediaMath engagement might be like?
Well, the definition of typical is changing. Twelve months ago, we were working “bottom up” focused on proving the performance across brand and direct response with hard and soft objectives, insights, and power of MediaMath’s technology atop biddable display through a high profile advertiser or two within our agency’s client roster, and then scaling horizontally within the agency.
Of late, we’ve become more “top down” — prospective agencies know we can produce results and the discussion is more about “how do I structure this within the agency or holding company to roll-out aggressively across my enterprise?”
Frankly, we’ve learned a lot from our agency partners as well! We’ve seen a tremendous amount of smart thinking and entrepreneurialism from our clients, and the product set has grown to meet their evolving needs.
Can the agency model survive given the rapid transformation toward digital media trading? What are key steps in your mind that agencies must make to be successful in this new, trading environment?
Absolutely and without question. Rumors of the agency demise have been greatly exaggerated: the agency model is on the brink of resurgence.
Marketing is becoming more complicated, not less. Digital is becoming a bigger percentage of the total, not smaller. Channels are integrating, not growing more siloed. The consumer is getting smarter, not dumber. Metrics are getting deeper, not shallower.
Marketers can’t quarterback a coherent program spanning all of this complexity on their own. An agency – the transformed, modern agency, well-versed in technology and analytics – is needed to orchestrate this dynamic landscape, pulling together the best tools and the best practices to achieve success for their clients, under incentive models making them partners in that success.
As to how to succeed? Find a trusted partner to provide the technical and quantitative DNA, who understands and believes in the agency model and can “get the flywheel spinning”. It gets easier with some success. Perhaps obviously, we think agencies are better served by working with a best-in-class and extensible solution instead of trying to assemble a hodgepodge of tools and providers.
Will Media Math ever trade its own book?
I’m not totally sure what this means. We happily bear performance and payment risk for our partners, because we’re confident in what we do, as well as supporting those who prefer to handle both themselves. Our singular focus is delivering superior results for our agency clients, whether through our SaaS solution or our managed service.
Does Media Math help facilitate the selling of display advertising media back to the exchange? Or is it just facilitating media buying at this point?
We are buyers only. We think selling media creates a conflict of interest that would put us at odds with our clients. The question of: should this impression get sold to our client at $1.00 CPM when it clearly performs for them, or should we sell it to the exchange at $1.20 CPM so we can profit, creates divided loyalties. We don’t want to deal with that, and thus, we only buy. This said, more and higher quality demand in an auction model helps media suppliers, so we view the relationship as “symbiotic”. Analytics creates value the game is not zero sum.
Of the ad networks and exchanges out there today, who’s getting it right?
That’s a tricky question. Everyone seems to have standardized on the right vision at this point: real-time impression-level bidding, prices driven by fair market supply & demand, easy access to third-party data sources, open access to APIs and support for “bring-your-own-algorithm,” preference to overlap versus underlap when resolving sales conflict, integration of premium and remnant, guaranteed and spot. Where people are on that roadmap, of course, varies. Some have chosen to emphasize certain parts of that vision over others, but most agree on the horizon and acknowledge, and the need to get there sooner rather than later.
It’s our proprietary pixel-serving technology that allows us to normalize information across supply sources, append third-party data, manage attribution and metrics, and build bid models based on our clients end-advertiser goals. Agencies tend to have standardized on DART or Atlas, and we focus on streamlining their workflow, not disrupting it.
Is there room for brand advertising with performance display advertising, or is PDA strictly about direct response?
Definitely both. Exchanges are about providing sources of biddable media where prices are driven by supply and demand, and where data, media, and audiences can be bought and sold programatically and algorithmically. Whether the goal is brand impact, engagement, direct response, etc., doesn’t matter — all types of goals can be, and are, supported. Considering exchanges “remnant” or “direct response” is hugely limiting.
Put another way, all marketing should be “performance based” just with a broad definition of the goals that include brand (that traditionally are viewed as unmeasurable) and direct response metrics. Yes, even brand marketers have goals, whether they be cost-per-audience member reached, brand awareness, net promoter score, or what have you and the tools exist to measure these. Simply replace “CPA” with these goals, and the entire power of this medium cost-efficiency, targeting & optimization, transparency, scalability, etc. are all brought to bear on brand marketing. This really needs a whitepaper or a book to answer well. Short answer: yes.
There is a lot of talk about better analytics and attribution capabilities for online advertising. Are you satisfied with ROI metrics that you’re tracking? What needs to be changed?
MediaMath’s systems support a whole range of upper funnel brand metrics and post-transaction “downstream” metrics. Our technology lets us develop real-time bid models against billions of impressions per day against all of these metrics today, while continuing to aggressively support more traditional online metrics like CPM-against-target, CPC, CPA (thank you page).
As the company’s name suggests, we’re metrics pioneers but we look to avoid the proverbial arrows by tempering zeal with an acknowledgement of market readiness for metrics broader than reach and deeper than CPA. Our goal is to make available the tools for agencies to help their clients clearly define, report on, and action against more sophisticated measures of return, with systems built to handle them, without requiring the adoption of same.
How does the online ad exchange model shift to a premium AND remnant inventory model?
Power buyers – like a MediaMath – move upmarket to create a “premium spot” market with deep, quality supply relationships, pricing above remnant, explicit avoidance of channel conflict, and quality advertisers.
The beauty of the free marketplace is that the cream rises to the top. We firmly believe that top publishers who have distinctive content, valuable audiences, and aggressively innovate can garner higher prices for their inventory in a premium spot market than the blended economics of direct sales.
How will Media Math evolve in the next 12-18 months?
Take advantage of our two-year lead in this market. Go from being a successful first-mover in media trading with a stellar client roster to the industry standard solution for top-tier agencies. Despite the allure of profitability and a track record of successful organic growth, we’ve elected to bring in a financial backer to provide the substantial capital base and expertise to help us extend our lead rather than just maintain it.
We hope the adexchanger audience helps us succeed! We strongly believe that
a healthy ecosystem helps our partners on both the buy side and sell side win. We’re very excited about being participants in the new world that the adexchanger is chronicling and increasingly helping to define.
Follow AdExchanger.com (@adexchanger) on Twitter.