Skip to main content

Screen Wars Thought Leader Interviews are also available on

SpotifyApple PodcastGoogle Podcast
Screen Wars Thought Leader Interviews

Ashwini Karandikar on Data-Driven Decision Making

By May 11, 2022November 14th, 2022No Comments

Ashwini Karandikar, EVP Media, Tech & Data at The 4A’s, joins Cross Screen Media CEO Michael Beach to share the importance of cross-channel measurement at the local level and data-driven decision-making for marketers and agencies. Watch our latest Screen Wars Thought Leader Interview here and read the full transcript below!

MB: Ashwini, thanks for joining us today.

AK: Of course, pleasure being here.

MB: Why don’t you give us some background on your career, and how you ended up where you are today?

AK: I’m an industry veteran. I’ve been around the advertising/marketing ecosystems for a while now. Way back, I was at a few startups, then I joined an agency that Dentsu (formerly Dentsu Aegis) bought. I ended up starting Amnet (Dentsu’s programmatic group) for them. I took it to over 50 countries.

I had a blast building that business and traveling the world. Just before COVID, I transitioned out, and I joined McKinsey for a while. I knew Marla Capozzi from before. When she mentioned the opportunity to lead the media technology and data practice for the 4A’s, I could not refuse. It’s been a year since I started leading this practice at the 4A’s, and I’m having a fantastic time doing it.

MB: I’m sure having the opportunity to have a real deep understanding of one agency, and then working with so many groups, had to be exciting and a good learning curve.

AK: It definitely was very intriguing. The time and space we are in right moves so fast, and so many things happen at the same time. And all of them affect people directly at the end of the day. It doesn’t matter which side of the equation you sit on. But I love being in this vantage point. Having built businesses before, I completely understand what an independent agency is going through, and after living at a holding company for so long, I get that side of the business as well. I think technology is the biggest equalizer of our time. Figuring out how different constituencies interact, react, and collaborate within the space that we live in… All I can say is that it is a great time to be here.

The nature of agencies has evolved from where it was from pure servicing, pure media buying, and pure TV buying, to anything that has to do with consumers.

MB: Would you mind giving us some background on the 4A’s? Who the customers are and kind of what the problem they solve?

AK: The 4A’s is a very well-established organization. It was set up in 1917 to help promote and advance the interests of member agencies. Currently, we serve more than 600 members, and we interact or have some interaction that accounts for over 85% of the total US advertising spend. What is not very well known is that the 4A’s also includes the 4A’s benefits division, which ensures more than 160,000 employees. We have a very robust 4A’s foundation that advocates for multicultural talent within the industry. Like everything else, the 4A’s evolved with this space as well. We are always at the forefront of defending and advocating for agencies. The nature of agencies has evolved from where it was from pure servicing, pure media buying, and pure TV buying, to anything that has to do with consumers.

We have been very diligent, and I’m very proud of the way the 4A’s health agencies deal with the whole pandemic. They were helping with legal questions, helping with do you go back to the office, do you not? What is the meaning of an office? What is the meaning of culture? What is the meaning of talent? How do you really hire? How do you expand the products and services that you offer? And like I said, at the end of the day, everything affects everyone at a very personal level. So, the 4A’s is in it for all of these different reasons and has been doing this for a very long time. I actually have a question for you, Michael.

MB: Fire away!

AK: It’s a similar question. Cross Screen Media is such a cool business. How did you arrive at where you are now? How did you think of the idea of Cross Screen Media? What was the genesis of Cross Screen Media?

MB: Great question. I’ve had three big work periods in my career. During the first, I worked in direct marketing for a long time, primarily in US politics. We really were on the cutting edge. We basically had a CRM file of every customer in the United States we wanted to target. And we were unique in that regard. We saw the intersection of digital and that kind of data being the future. And this is in the late 2000s. So I started an agency focused on that space and corporate advocacy. We kept running into challenges, like, how much do you allocate to digital vs. linear and other formats?

So we started to build some technology around answering that question for our agency. It was a really interesting challenge because you had to combine your digital data sets with set-top box data. Then, we saw that having all that capital expenditure for one agency was really tough. So we spun that technology into a separate company, and we exited the agency.

Our initial go-to-market was other small regional agencies that were smaller than we were. We were about 140 people at the time. To now offer all video together under one roof gave them a huge advantage in the market. And we’ve gone from there.

AK: It must have been such a cool journey, from the beginning, video was just what you saw on your TV screen. And now in the last 10 years, there isn’t a single device that does not have video of some sort.

MB: It’s a really interesting challenge, and it’s only growing. We focus so much on the allocation and where the dollars get placed, but the bigger question is, “What creative to make?” Something like TikTok, that’s gaining a lot of steam now, is a completely different video format than you would’ve needed before. That’s a lot more effort, either on the agency or the brand. Today is more complicated than five years ago. And five years ago it was so much more complicated than before that. I don’t see that trend stopping.

AK: It’s a bit scary.

MB: Absolutely. Good job security for all of us, though.

AK: Definitely.

Once you get to the local level, the questions and the potential problems exponentially multiply, because each market is a universe in itself, and you have to connect it all.

MB: Measurement on the video front is a huge area for our customer base. Are you seeing the same thing?

AK: Yeah. Measurement has always been a big question. Over the last 12-24 months, with everything that’s happened with Nielsen and broadly with measurement, I feel like the whole space is back in play and in flux. There’s been a lot of chatter around multiple currencies and single currencies. I think multiple currencies are here to stay.

For those of us who came into this from the digital side, multiple currencies were always a fact that you took into account. And then you figured out how to normalize it or map it against some base currency. If you’re currently buying at a broader scale, this sync measurement diversifies into what I call the vertical and the horizontal. Meaning, that a siloed base, like a TV network, picks one measurement provider, and it works for them. That becomes a vertical solution.

But then, for a market or an agency, you still need to connect the dots across multiple verticals. So it’s not just one or the other, it has to be both. And I think the newer measurement companies, Samba, Innovative, VideoAmp, et cetera, are leading the way in this “verticalization.” But at an agency or marketer level, you still need to be able to bring it all together. We are seeing this happen.

There’s also a growing need for verification and accreditation, which MRC offers. Even if different providers offer you different measurements, you still have to sort out how it matches to the broader ecosystem. There are a lot of moving parts and pieces. The 4A’s does extensive work with the ANA, with CIMM, with our agency partners, with you, Cross Screen Media, trying to figure out what’s a good framework for anyone to follow. And once you get to the local level, the questions and the potential problems exponentially multiply, because each market is a universe in itself, and you have to connect it all.

MB: Yeah. The world we live in is local. What are the specific challenges for local?

AK: Actually, I would love your insight on this. Cross Screen Media lives so much in the local space, and provides such a key offering that local markets need. I think the challenge is that geography restricts local access to data. The market they are in restricts what they see and what they’re able to measure, and how they’re able to make sense of what they’re seeing. There’s a need for better big data analysis to feed in the trends that any local market can use to give some context to what they are seeing in that given market. And then, being able to stitch it all together to make it make sense at a regional or a national level.

In fact, while I was thinking and preparing for this podcast, one of my questions for you was, what is Cross Screen’s view on multicurrency? Where do you think that’s going? And in terms of the data sets that you see within local, or just the request that you are seeing within the local market, how have they evolved over the last 10 years?

MB: I agree with you. It’s a multicurrency world. Coming from a digital background, people have different objectives. It always fascinated me that every TV I see has a different objective, and yet we still use the same primary currency. I know there are a lot of advantages to scale that television has versus other mediums, but we never looked at that like it was a dealbreaker. The customer base that selects us today is okay with multiple currencies right off the bat. They want to take more of a digital view of television. Not fully digital, but they look at both the same. And about 60% of the video spent on our platforms is still on broadcasting cable television.

But they want to figure out reaching the target, and everything else. Even 10 years ago, we built planning out first, which is different from most of the market. Everyone else has built measurement out initially. Part of that was just the challenge of bringing it all together. Especially when you were dealing with a moldy market plan.

You had two big insights: One, that media consumption against the exact same demo was dramatically different across markets. I’m in Washington, DC. If you looked at people in the market to buy a Ford F150 here, versus in Charlottesville, Virginia, the consumption habits of those audiences would be dramatically different.
The second thing was that the digital pricing was relatively similar. Maybe you’d find a 3 or 5% difference in CPMs between YouTube here versus in Charlottesville. But if you looked at linear, you could see a 100% difference. Because their pricing is still so human driven, it’s set in each individual market.

The CBS owner in Charlottesville doesn’t care about what NBC is charging in Washington, DC. They care about what NBC is charging in Charlottesville. And you get these huge spreads.
If you’re using a national model to plan local with your consumption habits and pricing, your plan is going to be wildly off. So we used those for a number of years, and we would end up thinking, “We’re going to get 800 points to the target in each of these markets.” And we would get 1300 in one and 700 in the other.

That was the genesis for building out planning first. And it serves us really well up to this point. Now the rest of the market is starting to figure it out. You’ve got audience data, you’ve got pricing history and all these things that a planning environment would have. It creates a lot of stickiness with the customer. That was our worldview behind doing planning first at the local level.

AK: I think that was smart. What did you have to do to get into planning? Usually, the path is the other way. You bring in data, you start reporting on it, you figure out some measurement framework, and then you move up the chain to say, “How did they plan this?” Or, “What were the decisions that led to the buyer that we are now measuring?” What were the different data feeds that you had to take into account?

MB: When we first built the product, it was internally facing back in 2013. We originally launched our first beta, and then it was set up. Box data was the first data set we had to bring in because our buying team used Strata to plan out the buy. But that wouldn’t work with a data set like RentTrack. So we built a more advanced optimizer, but then the next thing became, how do you deal with rates? How do you deal with avails? How do you deal with what actually happened? If your end report is coming out of Strata, and it’s in age and gender, but you put a plan together on the strategic target, you got a lot of data analysis work.

Those became the first pieces that we put together. Then we realized that a lot of those needed to be in one place. Those were all individual steps that we were doing as a service, so we launched Cross Screen Media. We took about nine months to rebuild everything and make that the core offering. Also, to help you answer the question, on the front end, what do I need to do? Before, if you knew you wanted to spend $500,000 on broadcast, we could tell you the answer to that question. But the minute it became, “I’ve got $5 million to spend on video in these 10 markets, what do I do?” That was much harder.

AK: What an incredible journey. I think it’s a much valued and much-needed connected workflow if you will. That’s very cool.

MB: We’ve tried to stay out of things that are in existing planning environments because a lot of our customers use Mediaocean or Strata. We’ve got to be really careful with our resources, so we don’t duplicate anything. A lot of what we’ve done is an add-on to what’s already in the market.

Sounds like 4A’s is working on a measurement report. Is that right?

AK: Yes. We are working on multiple measurement projects. We have a very robust committee staffed with lead experts of measurement across holding companies, with large independence. They live and breathe this whole question that we have been talking about. About bringing in multiple data sets, figuring out actually what the data is saying, and how do you apply it back to, either planning, or them figuring out how to further curate your bias.

We are about to launch a project with the ANA and CIMM about the overall purpose and usage of multicurrency measurement systems. Our aspiration is to come out with some sort of framework. Not dictate the partners or providers, but instead come out with a framework to say, “Assuming that multiple currencies are here to stay, you still need to make sense of it all at the end of the day. What are some frameworks and models that you could use?” We are about to get started with that. Obviously, we are in constant discussions with Nielsen, Comscore, and others.

For all practical purposes, Nielsen and Comscore are still the only two horizontal systems or the connectors. While Nielsen goes through its evolution and unveils Nielsen ONE, and Comscore goes through its product evolution, we are constantly in touch with them to not only understand what they are planning but how it impacts agencies. And in turn, how will it impact marketers? It’s a lot of good work. A lot of moving pieces.

I’m sure you are seeing a lot of it. For everyone on national, there are like a hundred local. It’s like an exponential one-to-many map.

MB: I think the model is 800 video advertisers at the local level for every one national video advertiser. Basically, in every media market, you’ve got about 250 brands by the predominant amount of national television. We’re estimating about 1,250 that are your primary CTV or national cable buyers. Then you’ve got almost a million video advertisers at the local level across the country. And it seemed really high, but in 2020, there were 3 million Facebook advertisers in the US, almost all those local. You can see how that would really work.

That leads to every media market having more buyers than buy at the upfront or anywhere else. So the technology stack needs to be totally different, because the average video buyer to local buys $120,000 a year, versus tens of millions of dollars at the national level. We’re trying to retrofit the same stack down, and it’s not going to work.

AK: Yeah. So how do you think multiple currencies are going to work in the local scenario? Meaning not only the digital multiple, but also the overarching multiple currency world that’s coming and is going to affect all of measurement across the board.

MB: I feel like a lot of our customer base has been dealing with that for a long time. For instance, almost every one of them is a Nielsen customer. At the local level, if they’re not a Nielsen customer, the station groups are all Nielsen customers. Many of them are now Comscore customers as well. So they’re getting a rating or an estimate back from one of those two people. And they’re just going into our environment figuring out, “I want to target this auto audience. And I got an 18 to 49-year-old adult target from the station.” The guarantee may still be based on that 18 to 49-year-old adult, but our customer is trying to figure out, “What program that has the same 18 to 49-year-old adult rating has the most auto intenders?”

I know that’s how we thought as an agency. There’s an asymmetric information opportunity here. I have the Nielsen numbers and the strategic number. The seller may only have one. Obviously, we have sellers that are our customers because they’re trying to have the same advantage. I don’t think it’s a big deal. I think it serves well to the sell-side to have multiple currencies as well. Because when they go into a buyer that may rely more on Nielsen data than Samba, or somebody else, it’s good to know that. And vice versa, when they go into a group they know is using another provider’s data. If they don’t price their inventory to that, they’re going to miss out on an opportunity because the other side’s probably looking at both data sources.

AK: You’re right. Everyone will have to become smarter about reading basic data sets and figuring out how it links together.

MB: One question on that. Is it common today that other groups are using multiple data sets, just like our customer base, or is that something unique that we’re seeing?

AK: Multiple data sets and multiple currencies are becoming the norm. The question is, what do you want to pay for? Are you going to pay for Nielsen, Samba, Innovate, and whoever else? And, how are you going to connect it back to the outcomes that the marketer or publisher is most interested in? The measurement equation falls on both sides. The buy-side is interested in measurement to figure out what impacted the media buyer on the outcome they want to achieve. While measurement is critical to figure out how to price the media, I’m sure the sell-side would like to understand how their consumers are interacting with their content. So we are seeing this across the board.

The buyer side has been more deliberate. Trying to figure out how to respond to the multiple currencies coming to market now. You hear a lot in terms of cost and validity from this side, about being able to benchmark the numbers for what they really are. And being able to connect them to a desired outcome. So you have a cool vantage point. You are seeing the data as it is coming in, and you can see the trends as they come in. Many agencies have created their internal measurement tools to do something similar. They analyze and synthesize the data that they see, and then try and connect it to the desired outcome across the board.

Many marketers and agencies don’t have the infrastructure to analyze all the data they receive.

MB: One of the things we get feedback is the business model, as well as, who’s going to pay for this? How do we pass this along? Is it the marketer that signs up for a data source, and the agency has to use it? Is the agency that signs up for it, and becomes part of working media? Or similar things. What do you see in the business buyer side?

AK: That’s a great question. I’ve been on multiple sides of this conversation in the past. Ideally, I think it’s best for the marketer to own the contractual relationship. Agencies also have their own proprietary measurement solutions or data synthesis solutions. The proposed models could be hybrid. It could be a subscription fee plus a dedicated team that helps with strategic questions, or big data questions. It could be a completely managed service, meaning, here’s all my data, please make sense of it and just give me back the output.

I’ve seen all of them work. It completely depends on the goals of the two parties in that given contract. Because many marketers and agencies don’t have the infrastructure to analyze all the data they receive. Especially now, with so much of it being generated.

As long as the infrastructure is taking into account all the privacy needs, it may be fine to work with a partner who manages all of this for you. I’ve seen multiple models. They all work or don’t work, based on what the two sides thought they were getting into. The common theme is that there’s a serious requirement for both sides to be legitimately educated on all of this. Data for the sake of data is just table stakes.

I remember being in conversations where a certain marketer had sales data from 10 years ago. That data is potentially useless if you don’t carry those products. I remember seeing that data, and I was like, “There are many products in this set that you don’t sell anymore. Why are you carrying this data around with you? What do you want to get out of it?” They didn’t know either. Someone had just told them that keeping data as long as possible is a good thing. Sometimes that’s just a cost that you are bearing without realizing what to do with it.
There is a need to have a basic understanding of data sets. You don’t need to be a programmer, but you definitely need to know what to do with the terabytes of data that you’re going to get.

It’s probably my pet peeve, but about five years ago, someone started this trend about wanting log files of all sorts. But what will you do with those log files? Because that’s simply a timestamp record of what happened, whenever it happened. And this notion that “If I get terabytes of log files, there’s some magic bullet in there.” There is not.

Out of all of this, what is becoming more and more obvious is that A) There is a real need for a real understanding of what you want to do with the data sets. B) There is a real requirement for marketers and agencies to work with each other. Because agencies, by definition, specialize in all these different facets.

Marketers have to focus on their business. Whether it’s developing, creating a product, selling the product, or developing markets. There’s a need for sophisticated partners to help bring all this information together in one place, so that the user does not have to spend hours pivoting and tabling everything, not knowing what to do with it. I think we are seeing all of this happen, and it’s a big positive for the agencies. Agency talent is trained to understand these different facets all at the same time. I think the partnerships are going to be good.

There’s a cost to storing. There's a cost to investigating. There's a cost to analyzing. There's a cost to recommending, even if a machine’s doing all of it.

MB: I love the log file example. That’s something we’ve seen rise up as well. Now some DSPs, or ad servers, give free access and others charge quite a bit because they’re having the same question. There’s definitely a cost to this.

There’s a cost to storing. There’s a cost to investigating. There’s a cost to analyzing. There’s a cost to recommending, even if a machine’s doing all of it. Someone has to train the machine to understand all that data. All of this can be a cost if you don’t know what you’re doing with that data. It could be a fabulous asset if it’s a feed into something bigger and better overall.

MB: Excellent. If we have this conversation a year from now, what’s one prediction that’s under the radar?

AK: I have many. In terms of predictions, I think one year from now we are still going to be working with Nielsen as measurement connected with better data feeds from other partners. But this notion that Nielsen is going to be gone? I’m not so sure. We are seeing gaming as the entry into better media planning at an even more macro level if that’s possible. Better media activation, to some extent, better consumer sentiment analysis.

Measurement for games will be a whole different ball of wax that we’ll have to sort out because gaming is a very active sport. In that case, how do you really make sense of the gamers’ intent other than seeing the game they are playing?

A year from now, we’ll still all be talking about what the metaverse is really about. Is it real? Is it not real? Is it one thing? Is it 10 things? Who knows? We also covered gaming extensively during our Decisions Conference. And the trend is leaning towards that. There’s more purposeful activation. The user is very clear about what they really want to do in that game, and they don’t want to be bothered with anything else. It will be interesting. What do you think?

MB: I love the gaming one. There’s a company, Anzu, that raised a decent amount of money last week for in-game ads. And it feels like there’s some scale there now. Measuring attention is going to be a big area. Obviously, it’s going to be eyes on screen, but do they see your ad? I think it will be a more complicated problem than when you’re in video.

AK: Do you have a prediction that no one is thinking about just yet?

MB: I’m really interested to see what Nielsen reports in their May issue of The Gauge. It will be the first time they’ve had it for a year straight because it’s gone up and down. That’s been interesting. Because when it comes out, the streaming people are beating their chest saying streaming is going to take over. Then the fall hits and the NFL and college football come back, and broadcast and cable creep up.

I’ll be really interested to see that first year-over-year comparison that they have. Because some of your Nielsen answers are that nothing’s going to replace anything overnight, but the blend is going to get more complicated. That’s the world we’re going to live in for the next 10 years. I think everyone’s looking for this binary outcome. Are we all going to streaming overnight, or are we all doing linear? Now it’s like, we’re going to be in the middle.

Even when we look at our customer plans, it’s rare to ever see less than 60% of the reach coming from a group that gets advertising from both digital and linear. Linear only and digital-only usually are the two smallest buckets. The biggest one is almost always in the middle. Now, it could be that 90% of your frequency came from linear, and 10% from digital. But the fact is that the consumer is getting it both places at some point. I think that’s only going to grow.
We’re probably just starting to see where that will be the biggest pie. Then maybe people will start to go streaming only. But I feel the same way about consumption.

And I think you see it down in the networks like your NBCs and your CBSs. They talk more about that now, that they have a streaming offering than they did when they were linear only. It’s “You can find our content in another place.”

I have one last question. Our audience is always looking to build their reading list. What should they be reading to understand video measurement, specifically local?

AK: There are plenty of media publications at the national level, local level, marketer level agency level. AdAge, Ad Bic, Drum, and AdExchanger, etc. Currently, and probably henceforth, there is so much good information, and good insights coming from non-traditional sources. The creator economy, what do you see within general talent pools? What do you see within the whole AI/ML discussions? In the past, they would all be very nicely bucketed.

Go for AI/ML only if you’re a coder, otherwise don’t worry about it. Technology is having such a broad impact on everything that is happening now and in the future. My recommendation would be to read the big established publications. Also, to follow any topic that may interest you in terms of media creativity, media types, the impact of tech on media production, on measurement, on the intersection of local and national, etc. I can’t keep up with it.

Many people use Twitter as a good source. That’s not a bad source, but it all depends on whom you’re selecting to follow. Cast a net as wide as you would like, because I don’t find information bucketed anymore. Are you trying to apply all the stuff that you’re reading about towards retail media, or is it for a market, or is it for a storefront, or is it for local point of sale? It all depends on the application. This is a very long answer to say: Read as much as you can. There is not a single source anymore.

MB: Excellent. I appreciate your time, and I know our audience is going to love this conversation. Thank you.

AK: Thank you for inviting me. I think it’s always fun to chat with you about what’s coming next.

Ashwini Karandikar is an internationally recognized entrepreneurial executive with proven experience in advertising, business development and operational excellence across the globe. Motivational and mentoring leader with an excellent track record of building diverse and talented teams in over 50+ markets. Keen and proven eye for identifying and converting opportunities into long term and sustainable revenue generating business outcomes for employees, company and clients alike.

I have been recognized as an agency innovator, leader in diversity and inclusion, and nominated as one of the 30 most powerful women in mobile advertising.

At present I lead the media, technology & data practice for the 4As. Prior to this I was a senior advisor for McKinsey’s high tech, media & telecom practice, and before that was with Dentsu where I started and oversaw Amnet, Dentsu’s programmatic group, across 50+ markets. I also serve as a director on MiQ’s Board of Directors. 

I am very active in performing arts, music and theater. I am founder/member in several voluntary groups focused on bringing performing arts and artists to local communities through theater/drama, instrumental and local concerts.

Cross Screen Media is a marketing analytics and software company empowering marketers to plan, activate, and measure Connected TV and audience-driven Linear TV advertising at the local level. Our closed-loop solutions help brands, agencies, and networks succeed in the Convergent TV space. For more information, visit

Michael Beach

Michael Beach is the Chief Executive Officer of Cross Screen Media, a media analytics and software company that enables marketers to plan, activate, and measure CTV and linear TV at the local level. Michael is also the founder and editor of State of the Screens, a weekly newsletter focused on video advertising that is a must-read for thought leaders in the advertising industry. He has appeared in such publications as PBS Frontline, The Wall Street Journal, The New York Times, Axios, CNBC and Bloomberg, and on NPR’s Planet Money podcast.