covid, cookieless, universal id, identity, consumer data, sharethis, eyeota, covid19, consumer behavior, eyeota translate

Connecting with Consumers in a Post-Covid Cookieless World: Audience Data Sessions

By: Eyeota

Eyeota CEO and Co-Founder, Kristina Prokop sat down with Michael Gorman, SVP of Product and Marketing at ShareThis, to discuss how consumer behavior has changed since the onset of the pandemic as we approach a post-covid cookieless future and how this has influenced the way brands use data. Here is a summary of the key points touched upon.

Kristina: Why don't you start off by just telling us a little bit about ShareThis and the role that you play in the data landscape?

Michael: ShareThis represents global consumer interest. We have tools that enable users to visit websites and to share their content on social networks and elsewhere. We leverage the data from those experiences to build audiences and to provide products that support analytics.

Kristina: We work together closely in the programmatic ecosystem—are there other areas that people are using ShareThis products in today?

Michael: Yes, absolutely. One area—that has always been a factor but has gotten really important during this pandemic period—are consumer insights. We're increasingly providing consumer insights as a standard part of our business. It has been pretty fascinating to watch as the pandemic has unrolled.

Kristina: What are some of the major changes that you have seen or insights that you have put together over the last 18 months?

Michael: From late February into early April interest in health went up four times globally and travel turned off like a light switch. As we moved into lockdown mode, people were going to the grocery store and bulk buying. Interesting sub-trends came up: even though travel as a whole was down, road trips, national parks and rental cars started coming up. As we went through summer, as people were realizing outdoors was the place to be, home improvement improved generally, but backyard garden improvements were twice as popular. We could start to see real estate move as people were re-sorting where they spent their time. There were signals into the fall that the demand for suburban and rural real estate, and real estate as a whole, was not going to stop. It was not just a summertime phenomenon. Things related to climate change and electric vehicles seem to be continuing strong. Online shopping is the trend that doesn't stop. It grew, it popped quickly, and it just keeps on growing rapidly.

Kristina: The differences are massive depending on what region you're in—as a global company are you seeing huge shifts just over the last month or so?

Michael: Well—actually—I would say it's a little more gradual in the data, but we’re seeing trends that are clearly picking up. Beauty products, for example, are showing us a steady positive trend.

Kristina: You mentioned that there are ways in which clients use ShareThis data for insights: can you give us an example of clients that are using it for insight and how they apply that? Is it an overlay into their first party data? Are they looking for trends to gauge decision-making in their business?

Michael: Yes, exactly - it's more the latter. The nice thing about our position at ShareThis is that we have something interesting to tell. We can tell the client: 'These are the people that we see who are interested in your brand and here's what else they're doing'. If we then add first party data in, I think we can get into more sophisticated analyses to help deepen understanding of the customer. But I do think, Kristina, we may be over-weighting first-party data. It's great, but we may be overdoing it a little bit in how much emphasis is being placed on it. The majority of interactions, that marketers are going to have with people, are with people that they may not know at all. They may not even be sure that they are customers or have any data collected on them.

Kristina: Can you share a bit about how ShareThis is approaching identity and the cookieless landscape?

Michael: As a company, ShareThis has a lot of scale and a lot of interactions with consumers because of our presence on page. So we've always played a secondary or a supporting role in the whole identity system. We have worked closely with a lot of identity partners, either as partners, customers or suppliers, because you need to have to have a lot of interactions with consumers, or at least have access to them, to keep an identity system current. We have developed a solution we're calling the Atlas Global ID to bridge on those existing relationships and modify our identity, our way of keeping track of users, so as to not rely on the third-party cookie anymore but still preserve the ability to deliver insights about users going forward. I think it's consistent with other approaches you've talked about: agnostic identity. We will be integrated with all the major providers of universal ID solutions and our goal is to give our clients a set of choices so that they can find one that works best for them in the cookieless world.

Kristina: Where is ShareThis today in terms of live testing environments?

Michael: We're in a phase where everyone is getting their internal act together and is starting to share the products as they become ready for the testing phase. We're right in the middle of that mix ourselves. There's a part of the industry that is focusing on preserving the deterministic characteristics of the identity system and porting those over to using new identifiers and new methods. That's going to be great and it's one of the IDs we want to support, but a number of companies are also pioneering probabilistic methods. We're really supportive of that because I think that we've overemphasized knowing exactly who someone is and the promise of 'one-to-one'. And, in fact, most of the time, as marketers, we're really dealing with groups of people. Our product is based on having a very broad but fairly shallow sample of global behavior. We find that if you see people 10 times a month somewhat randomly it's enough information to categorize them. 10 observations on a billion people is in some ways better than a thousand observations on a million. I think we may get a chance to see a kind of A/B test because the United States seems to be gravitating towards the approach trying to preserve precise measurement, and on the flip side, in Europe the regulatory requirements are such that people are gravitating more towards these probabilistic and cohort-based solutions.

Kristina: Can you share a little bit about some of the new updates that we've done to our partnership so far in 2021?

Michael: We are connected around the identity solution and making sure Eyeota is one of the partners for which the bridge continues to operate so we can provide data there. Another area that I'd highlight is data validation and being able to give marketers confidence that the data is not only accurate, but that it's well-made and appropriately made. We were an early customer for Neutronian and I know Eyeota was too. We've been really happy with that result. If you look at ShareThis in Google Search at the moment, we're highlighting the fact that we worked with another partner—Polk, owned by IHS market—to have them test our data for the presence of auto buyers in our auto segments. If we say this is an auto segment for Audi, then how many people actually buy Audis with a higher propensity than the national average? We were able to validate that across 11 of our segments.
We see Eyeota focusing more on supporting marketers with analytics and offering ways to do not just targeting, but to manage their prospecting programs, if you will. We're looking to support that with products on our side. For your Enrich offering, for example, we're looking to provide segments that are tuned to support modeling as opposed to tuned to support just targeting.

Kristina: ShareThis's reach is very globally spread—are most of the products you have in-market available equally across the globe?

Michael: Absolutely—a global version of our products is available across the board. We can also tune it - for example, we have made versions for Indonesia or Portugal. This is at country level, but in the US we work regionally as well. We observed 250 different languages in content and we are doing data science on 19 languages that are being actively used as a source for the enrichment of our data assets. And we don't just do Portuguese for Portugal; the languages and geographic boundaries are all mixed up. We find that being multilingual is as important as being a multinational.

Kristina: You can also do a lot of custom work with data—is it primarily based on keywords or what do you focus on when you're building a custom audience for clients?

Michael: The unique thing about our data asset is that we preserve access to the most granular layer of data, the visit of a user to a page. Our key unit of analysis really is the page, what's on the page. We are using data science and some of the most current techniques to decode language and to characterize it mathematically so that it's essentially infinitely configurable. So yes, it's keywords, but it's also the concepts, the meaning. One of our methods these days is to say 'Just give us 10 webpages, that reflect the sort of people that you want to reach, the sort of people who are engaging with this content'. Then we'll rank up all 600 million that we saw last month about that topic and build a cohort, if you will, of people interacting with content like that. It's powerful stuff and very flexible. We love it when clients and agencies can be more specific about what they are really trying to do or what they're looking for - that allows us to excel.

Kristina: Looking at the challenges that brands and marketers are going to be facing realistically over the next couple of years, is there any specific advice or guidance that you would give to brands and marketers about using data in the future in the third-party cookieless, changing landscape that we will be living through?

Michael: Yes. The critical thing for a marketer to focus on in this time of change around shifting to a new identity system is measurement first. If you know how you're going to measure and you work through the challenges of what is going to constitute success, you can then work your way back and go to all of your key suppliers, vendors and partners and say 'This is what I'm going to do, how can you accommodate that?'. The marketing industry as a whole has shifted—not quite as fast as the world shifted with the onset of the pandemic—but very rapidly all the companies in this industry are moving to a different mode of performance. Everyone has a solution. As a marketer, you probably find that 80% of your existing suppliers are able to adapt and provide comparable (or even superior) service, depending on how they mesh with your choice of measurement. Anytime change happens it's exciting and new solutions are going to pop up. While I don't think anybody should fall in love with the latest thing, maybe this is the year to focus and say 'I can run everything on a de-identified basis when it comes to marketing; I don't even need identity, or precise identity.'


Watch the Eyeota Audience Data Sessions: Connecting with Consumers in a Post-Covid Cookieless World video in full here.

Watch the Video Here

With Eyeota, ShareThis audiences are available globally for omnichannel campaign activation on display, mobile and social channels. For more information about ShareThis, connect with an audience specialist today at