Over the past year, MediaMath has just about doubled its staff to 306, as the company works to broaden its identity from being a demand-side platform to being viewed as an end-to-end enterprise solution. The company also created a new role—the “chief commercial officer”—in which it hired former McKinsey & Co. executive Michael Lamb. As Lamb noted in an interview, 80% of the New York company's revenues come from software sales, with a share of media spending making up the rest.
Before his decade-long stint at McKinsey, Lamb was co-founder of Poindexter Systems (now known as [X+1]), where he worked with MediaMath CEO Joe Zawadzki. Asked about the new post, Zawadzki said, "Ad tech, historically, focused on feeds-and-speeds, almost reveling in acronyms and its abstraction from how that technology was to be used and its impact on outcomes. We thought that was crazy. We wanted to create a discipline that bridged our clients' strategic needs with our platform's capabilities, and helped translate the potential into actual results. Thus, the role of chief commercial officer."
Lamb spoke with AdExchanger about the company's evolution and his role in it.
AdExchanger: What is a chief commercial officer and why does MediaMath need one?
Michael Lamb: The role of the chief commercial officer and the non-traditional title entails oversight of corporate strategy and development. It refers to commercialization and operational effectiveness.
It's also about the medium-term direction of the company. Specifically, I'm the bridge between setting the mission and vision from Joe Zawadzki on to the execution on the sales side. I'll be helping set the 12-18 month corporate strategy, which will look at potential M&As, products and revenue management. This role and this discipline is something every large software company has. As we go from focusing on campaign-oriented ad tech sales to an enterprise/software model, it's a new arena for MediaMath commercially.
What is MediaMath’s status in its move from a traditional ad tech business to a software provider? Will it charge based on a complete SaaS model – or will it be a hybrid?
I would describe it as "SaaS-like." It has all the characteristics of a SaaS model, with recurring revenue from software licensing and margins. It's still mostly a percent of media model, however, though 80% of MediaMath's revenue today comes from software sales with master service agreements. The rest is of the revenue comes from campaigns.
The change that's going reflects how clients who had typically just used MediaMath's platform to manage their programmatic performance in a narrow way are now looking more broadly at managing all their display.
A couple of years ago, we would have been offering clients our particular skills at optimizing their campaigns for them. Increasingly, it's a bit more self-service and we're working more on helping clients integrate their various sources of data together or helping build their larger programmatic practice, such as what their spending should be and what sort of goals make the most sense over the long term. Secondly, we're doing more advising on direct buys, which used to be outside MediaMath's platform. We're starting to see more programmatic direct activity on our platform, which allows the clients to get all the reporting in one place.
With that evolution in mind, should we still think of MediaMath as a DSP or has it become something else?
We don't mind that term. We always saw ourselves as a one-to-one marketing tool for advertisers and agencies, offering data storage, insight, media connectivity and execution. We're trying to define the category of one-to-one marketing on a multichannel basis.
In terms of focusing strictly on the "demand-side" and how DSPs are evolving, we've always worked with publishers and buyers directly. If the old world had companies like ours confined to exchange inventory, maybe you weren't connected to publishers. Or, if you worked with agencies, you never spoke to the brands. In truth, I don't think it was ever the case. But if it was, it's not true now because of the emergence of programmatic direct. Our inventory is sourced directly from publishers and we can, in keeping with the demand-side orientation we've always had, complete those direct deals. In a larger sense, as we more fully embrace the enterprise model, the focus is on automating the entire buying media [process], including planning to execution.
Does that mean workflow becomes a more formal part of MediaMath?
We do deal discovery, where we have arrangements with publishers and then help buyers find the right inventory offered that way. That's a form of workflow. But we have to finish that thought and bring the inventory securing process under the MediaMath umbrella. But we haven't concentrated on taking the wrinkles out of the insertion order process, which we feel is withering away anyway. We attach workflow from the digital native direction, as opposed to a Mediaocean, which is still largely looking at the traditional form of media buying workflow.
There's a second version of workflow we're working on, which is about building systems that lets clients manage all the complexity of their media without the need of a PhD. That's about making the interfaces and dashboard's a little more intuitive. In that sense, we're in the workflow business and that's a problem we're trying to solve.
A number of companies like PubMatic and AppNexus have built out separate consultative services, which stand apart from the technology and software offerings. Is that something MediaMath is doing or considering given your background at McKinsey & Co?
That kind of consulting and sales is a natural, internal part of the enterprise model, so yes, we do provide strategy in addition to technology. But it's also not that new. In a sense, MediaMath has always offered guidance of how best to use its tools. Whether we do some separate, branded group is another thing.
I think these kinds of services will likely remain as part of the larger "cost of sales" offering. And it all begins with answering a client's basic question: "How do I translate the old marketing strategies I used to do to more quantifiable, goal-based, programmatic means?" This is just a part of the natural evolution that's taking place in the category we all work in.
About The Author: MediaMath
More posts by MediaMath