Jay Kulkarni is CEO of Theorem, a media operations company.
Kulkarni discussed his company and his views on the industry with AdExchanger earlier this month.
Click below or scroll for more:
- DoubleClick Days
- Defining Programmatic Buying
- Theorem Target Market
- App Models
- Agency Model & Milestones Ahead
JK: Right, I was an early employee on the product side for publishers, advertisers and email at DoubleClick. Those five years were the “boom” times, if you will. I lived through the boom and the bust. From a pure technology perspective, in the early days it was a time of heavy innovation. We just kept on banging away at products.
To put it in perspective, those were the days of Microsoft and Netscape wars, where the Internet eyeballs were about 50 million. Now they are a multiple of that. So, even though there was a lot of innovation taking place, there was plenty of waste and inefficiency. There weren’t too many metrics to measure the efficiency of campaigns, either.
Today, most of the new innovations happening – for example, with the buying and selling of inventory, exchanges, sell‑sides and RTBs – those are a sign of maturity of an industry, where you can bid on a per‑impression basis, as opposed to buying bulk inventory on a CPM basis.
From a creative standpoint, you have the three major screens. But even within mobile, you have a lot of fragmentation. Ten years ago, it was about dial-up modems. If you had a cable modem, then you were cutting edge. Right now, I see a lot less talk about bandwidth and access to the Internet. It’s taken for granted.
For some time, we thought display was going away. The richness of creative was ebbing. But now it’s making a strong comeback. That’s how I look at the macro industry over that time period.
For me, programmatic buying, conceptually, is an economic problem and solution. If you can cherry‑pick certain impression‑level buys through real‑time bidding (RTB) – whether the platform on the demand side is a demand‑side platform, on the sell side is RTB/SSP – it doesn’t matter. If we are able to reduce the human element of picking up the phone and calling the media sales guy and, instead, let machines look at the budget, the audience you’re looking to target and, within a predefined time frame, get you all the audience you need across the sites that you’re looking for with minimal human interaction at the best price possible – for me, that’s programmatic buying.
How has it impacted Theorem? And, when did the impact start?
Believe it or not, when we used to deliver products at DoubleClick, this concept existed. This exchange, real‑time idea has been around for a while. The market, the technology and the economics of it – all of it lined up four or five years ago. Only in the last two years, you’re seeing it become real‑time. I could be off, but the volume of buys going through these automated platforms is approaching a billion‑plus in revenue.
So, the product started to be fleshed out, 5-6 years ago. Only in the last two years has it become more mainstream, where people are saying, “Well, let’s give it a shot.”
As for how it impacts Theorem, it impacts us positively. We have strong relationships with most of the companies that are growing and thriving in this space.
There are two ways our process and our business is changing. The first one is the ability for us to do higher volumes of campaigns has increased because of this programmatic buying. Similar to search you have a different mindset of performance‑based trafficking or ad optimization that we are embracing and thriving on. Also, because of the sheer volume of banners that are consumed in programmatic buying, that helps our creative business since the number of ads that are being produced by Theorem has increased. Finally, we have a strong and mature reporting and data business, and programmatic buying has had a positive impact because the sheer volume of data means the amount of reporting that needs to be done increases. We are looking at reporting two ways – one is the human element of reporting, and then there’s technological perspective. Because the number of data sources has increased, we have launched, in beta, a product called SAM – Smart Adapters for Marketers – that helps automatically download data from different digital sources.
Our clients are actually well‑rounded and holistic. We work with large networks and technology providers, publishers, and the new DSPs and SSPs that are coming. This gives a very strong view on the market. We do three things. First, creative production – and once the creative is produced, it needs to be “set up,” or trafficked, whether it’s in search, email or display. We do all of that. Once it is set up and trafficked, then we need to report.
On the technology providers, like I mentioned, we work with most of the DSPs and SSPs out there. We also work with Microsoft, Google and Yahoo!. We’re also seeing a strong upsurge on the publisher networks. Publishers and networks make sense because, as content is migrating from traditional to digital, we are seeing a strong uptake. From a creative production perspective, mobile is one of our fastest‑growing areas.
Often, Theorem is compared with a company like Operative. How do you explain the differentiation today between yourselves and others in the ad operations marketplace?
Ad operations is one element of it. As I said, we have three major footprints. The fact that we have a strong creative production and expertise is very big. That’s unique since few companies like us have invested in that area.
We work with every rich media provider and every major mobile ad network out there, so that is an instant differentiation.
Media operations and ad operations are second.
Obviously, the third one, reporting, is also key. The fact that we have those three footprints is unique. In terms of companies, we may be one of the few that is so holistic and broad. We have a global operational, sales and account management presence, which we own. We don’t sub-contract. We don’t re‑outsource. We don’t have contract people. Right now, we are touching 800 people globally. We were born and our operations are in India. We also have production in the Dominican Republic. The point is, we own our operations, which are all ISO-certified. That’s one of the reasons some of the biggest companies, when they do deals with us, are multi‑year deals, because they know that we have invested in that kind of a mature infrastructure.
Lastly, in terms of the competitive mindset, there are two spectrums of it. One is the small freelancers and consultants that we run into, which is good, because they bring something to the table, which is niche, but they lack the scale we have.
On the other end of the spectrum, we have bigger, publicly-traded companies who are seeing digital marketing grow quickly. They are starting to make forays into it. They may have the global scale and infrastructure and be publicly traded companies, but they don’t have the expertise or the familiarity with the technology that we have. We know every ad server and rich media provider out there. We know every data source, so that’s where we have an edge and we hold our own.
There seems to be a trend with the “app” model in ad tech. Do you see the app model – an interoperability model – coming to Theorem? What do you think the impact of an app infrastructure could have on digital advertising?
It’s an innovative model, but it’s not new. We are looking at it. In terms of what it could help. I believe it will help us scale as an industry. If I look at Theorem’s business, there is a lot of duplicity of effort. There’s a lot of redundancy and redoing things. If you look at the concept of an app, it means it will operate across different platforms. That’s great for the business because the more we can reduce the friction and the operational challenge/ And then we can take our talents and look at the more innovative part of it. Conceptually, it is a great idea, but it is still too early to say where it may take us.
We have agency customers and work with some agencies as our partners. What the agencies have is the strategic media buying, the branding, the media capabilities. Whereas, we are complimentary to that and heavy on the technology, innovation, automation and the production side of it. They are complimentary. We don’t do the branding piece of it, which is where the agencies excel. In fact, some creative is done for the agencies and they bring the storyboard, the concept while we excel on the production side of it.
The market is relatively new and all the business models are in flux. At some point, it may seem one is similar to the other. But the core capabilities of what Theorem has and what the agencies do are complimentary to each other and testified by our support of agency campaigns.
Are you starting to see marketers come directly to you?
Not really. One of the categories I didn’t mention when I described our client base is we don’t have too many brands directly in our model. Our scale comes from ad networks, publishers, the larger agencies, if you will, and the large tech companies. Most of our business is concentrated in these four categories.
Looking ahead, in the next 12‑18 months, what are some goals you’d like to achieve?
In February we had our 10th anniversary, so we are currently doing a deep dive into what the next opportunities present, if you will. In terms of the next 12-18 months, two or three things come to mind. Creative is one of the fastest growing parts of the business. Within that, the mobile piece is growing well. This is a specialized skill set. We plan to grow aggressively and educate our teams on it.
We do email, display, search and mobile. We have stayed away from social media but have more agencies and brands in place in those platforms to take advantage of Facebook. As Facebook publishes its growing display ad revenues, that will be a catalyst for more volume. We need to look at this opportunity closely.
Lastly, data automation. I mentioned our new product that we are launching, which is SAM. When I ask my teams who do reports, where they spend their most time, it’s actually counter‑intuitive. They said 80 percent of the time goes into getting access to data, and only 20 percent in compiling reports. That piece is ripe for automation, which we hope to address.
By John Ebbert