Coming Full Circle in Research by Making Big Data Smaller. An Interview with Maru/Matchbox’s Bruce Friend

“Technology-enabled research is a trend in the marketplace,” noted Bruce Friend, President of Global Media and Entertainment, Maru/Matchbox. 

Harnessing the power of technology in the pursuit of research insights is something at which Friend excels thanks to his previous work at OTX and Vision Critical. Now, much of his effort focuses on “building capabilities that are platform-based [unlike] a traditional research company, which still tends to be very focused on one-off, ad hoc studies, employing different survey-based methodologies.” 

Friend helps clients create single source panels that leverage a company’s own consumer assets, essentially connecting behavioral data from their DMP and other data assets with survey-based demographic and attitudinal data to track over time. “Furthermore, we build custom panels where our clients own the panel members and all of the data collected on those members. So, we are helping them create their own insights platform and program that has tangible asset value.  This value then continues to grow, as the panel size increases and we collect more data on the members,” he explained.

Charlene Weisler:  So how has the custom research business changed?

Bruce Friend: We are seeing that companies want to work with research partners in an ongoing relationship - always on, always delivering in terms of data and in terms of thinking. That is where we are heading. We are doing so not only by continuing to partner with Vision Critical, which spun off the research consulting part of thir business to become Maru/Matchbox 2 ½ years ago but by also acquiring new companies that complement the part of the business that we are concentrating on as well as enable us to do things on a standalone basis, as well as on private panels and communities. 

Charlene Weisler: And the media landscape overall is changing.

Bruce Friend: It’s been an interesting time in this “pending merger world” that the media industry is in now. We are seeing (content) companies building out full capabilities to support the process from script-to-screen. And beyond script-to-screen, really – being able to control the entire ecosystem from the standpoint of developing content, marketing it, distributing it, continuing to build franchises and monetize the businesses going forward. It is also about the platform – not just the content anymore. Its about how we get the content to the consumers in different ways. We see that even from companies such as Amazon (Prime Video), Twitter (TV, Video), Facebook (Watch) and others, who are adapting to video being a new and  increasingly dominant content form. It always seems that it all comes back to video and we are certainly seeing that more and more in the online space.

Charlene Weisler: Yes video is important. But more people are talking also about voice.

Bruce Friend: Yes, obviously voice activation is going to be the norm. In the not too distant future, I see companies conducting surveys through Alexa and Google Home. There are certainly some privacy issues around that sort of thing and I am sure there will be ways of working through that. Probably by creating panels of homes where people will opt in. In addition to voice, audio is also making a resurgence. We currently work with four or five companies that are very audio-focused.  Just like with video, audio is finding many new areas where it can exist and thrive. The emergence of podcasting is only going to continue. It is an indication of where our business is heading, where people want to listen to what they want when they want to listen to it – just like video.

Charlene Weisler: In creating panels from a clients’ own dataset, it sounds like you are able to fully leverage first-party data. Is this a trend? And what if a company doesn’t have a lot of first party data?

 

Bruce Friend: We are leveraging both first and third-party data. Certainly there are many resources for third-party data. Obviously the first-party data is better because it is essentially a 100% match rate as we recruit customers directly from the client’s database to become panel members. But in some cases some people don’t have first-party data. In those cases, we look for third-party data matching opportunities. We also run our own panel here in the U.S. and in Canada – both have around 250,000 members – so we can leverage them, as well as look to match more data sources into them. Our panels can also be used to look at communities outside of the company’s own panel. If you want to look at competitive viewers and competitive distribution services for example, you can then leverage our panel in addition to your own. Clients don’t always want to talk to just their own customers. 

Charlene Weisler: Do you see any evolution in how online communities are being built and used?

Bruce Friend: When communities started, they were about better, faster and cheaper. Communities were the start of agile research. The client could control and use the platform as a DYI tool. Most of these panels were 5,000 to 15,000 members. The trend is now not to have these smaller siloed communities within one company and across different brands, but to build a mega-community or an enterprise-wide community. We can now put all of these communities together with an organization with 50,000 to 150,000 members across the organization. Going bigger is better and when you then connect your DMP or other specific first-party data. You then have a very powerful asset with enough scale to do some very interesting things with the data and with surveys on top of it to give you more strategic insights. Communities used to be “light tactical” research – most people were not using communities for very strategic work. What we are finding now with some of our larger clients, who have made the effort to build out bigger panels, some as large at 175,00 members, is that they can now do more strategic work on them. As a result, we are seeing budgets move from more traditional research into platform-based panel offerings, such as ours, where clients can better leverage their own big data. 

Charlene Weisler:  So where do you see research going?

Bruce Friend: I see this model where companies tie into technology with an embedded community where you can talk to someone today, talk to them again a week from now, and on an ongoing basis. The company owns the panel asset, is building out that asset, that asset really has (data) currency to them while they can still conduct large survey studies within the panel. But it’s really about the creation of a resource that links behavioral data, attitudinal survey data, qualitative data, etc. into an ongoing relationship in an ongoing data stream. Automation will drive a lot of this, as well as will A.I. We are making big data smaller, more contextual and more understandable because we are looking at data that is in a panel and is more representative of the audience or subscriber base that the client has. 

Charlene Weisler: Sort of bringing the data science and ethnography elements of research back together.

Bruce Friend: Yes. I feel that we are coming full circle, back to where we were years ago when I first entered the industry. At that time research and big data lived harmoniously within the same insights departments, and that must happen again in my opinion. Otherwise, companies today that haven’t already moved to effectively consolidate their research and data science teams into one, and build business intelligence assets that support their entire organization, run the real risk of rapidly falling behind their competitors that have.

This article first appeared on www.MediaVillage.com

Lotame Ignite Americas Conference Focuses on the Power of Data

Data has (finally) become the force for change in our industry. Not only have many media companies devoted much of their upfronts about it, but Lotame's inaugural Ignite Americas forum last week expanded the discussion of how data can be used to improve ROI, upend legacy measurements, tell a story and totally restructure a company. It is finally revenge of the nerds. 

Data Attributes Have Evolved

There is so much data available from first, second and third party data that the parameters of what constitutes a valuable dataset have evolved. Previously, data size was one of the paramount requests. Do we have a large enough sample approaching census that will get us stable results? Now, with the complexity and sophistication of data availability, the wants have shifted from size to data quality. The components of data quality are Transparency (Do we understand the data attributes and origins?) Accuracy (Are we getting actionable successful results?) and Performance (Do the results deliver on the KPIs and/or ROI?).

Data expert Joyce Lee, Director Global Data Sales Strategy, Oath, has seen her role evolve from accumulating third party data to finding ways to "sync all first party data in the Oath family." Data stitching is like cooking, she noted, from having the right ingredients, to blending them together to serving a final, perfect result. Michelle Mirshak, Vice President Data Architecture and Platforms, Spark Foundry, noted that "Some of our more sophisticated clients are asking us to normalize data across channels to better track performance," as the data is analyzed in a more holistic way. The importance of data quality cannot be understated. According to Mirshak, whether it is third, second or first party data, "it all comes down to performance," because quality is not just a first party data attribute.

Data From Collection to Insights

Alejandro Matos, Digital Marketing Director, Omnicom Media Group, Dominican Republic and The Caribbean, explained the challenge he faced in generating insights for a local retailer who had limited data on its online consumers. “We needed to come up with a way to capture data,” he said. His work in capturing data that explained the consumer journey focused on a variety of sources - from placing beacons in various locations in the store, launching a DMP, CRM on social media and on apps, placing pixels in banner ads, gathering data on mobile, emails, websites, from multiple sources and then matched with third party data before modeling.

The result was the ability to more clearly understand the consumer journey, but it was putting the collection structure into place that made all the difference. Is every data project unique and custom? Not necessarily, according to Matos. The collection methods may be similar across clients but the insights and stories that can be crafted from the data, even from similar data sets, will differ.

A Move From Legacy Metrics to Segmentation and Attribution

With the availability of sophisticated datasets, it is time for the industry to move away from legacy age and gender demographics. According to Chris Frazier, Vice President Business Intelligence, Cadent Network, his company uses data to, “build what that audience will be” by building out a target consumer beyond age and gender, to “reach the intended target at the right time.” He added that linking to linear TV using traditional Nielsen to digital platform performance is a challenge, impeding the ability to measure and guarantee standardize-able sales deliveries across platforms. “We would like to see uniformity in how we measure impressions. Is it two seconds? Is it a minute? We want to see industry standards,” he stated. When it comes to addressable, attribution is key. “Attribution allows you to connect your exposures to where the sales are. It’s a measure of ROI to media spend.”  

The TV industry is beginning to embrace the use of consumer data in measurement in conjunction with demographic data enabling cross platform measurement. The need for a holistic, unified view of audiences and campaigns has never been greater and is essential for the evolution of the advertising industry. As data advocate, Andy Monfried, Founder and CEO, Lotame, concluded in his opening keynote, the requirements of a DMP is to unify disparate datasets to target the right audiences, extend a brand’s position to find and reach new customers and to better understand a consumer journey through greater personalization. This requires internal buy-in, retaining talent and aligning the strategic corporate vision to better understand and execute on the data insights. We are at the beginning stages. For the industry and for a company like Lotame, it should be an exciting and ground-breaking time going forward.

This article first appeared in www.MediaVillage.com

What Is a DMP? And Why Does Everyone Want One?

As the media business becomes more data driven, those working in it are creating new terms to help describe new processes, such as “DMP.” And what is a DMP? It’s a data management platform.

According to the CIMM Lexicon, DMPs are systems and data repositories to store, organize, manage, and retrieve data sets. DMPs help normalize datasets to enable audience analytics and, ideally, to optimize media buys for advertisers. DMPs are a crucial part of any successful data-driven campaign, enabling large repositories of data to be accurately integrated, compared, and analyzed.

The Local Advantage

Local TV, which for years has had to contend with small local market measurement samples, stands to greatly benefit from the big data capability of DMPs. Not only can they now collect more and disparate first- and third-party datasets related to local TV performance into larger, more stable samples, but they can also confidently expand their analyses across platforms and create audience segments (depending on what data their DMP ingests).


Read the full article on the Videa blog.