Name:
Data Trends: The Alchemy of Turning Your Disparate Data into Gold
Description:
Data Trends: The Alchemy of Turning Your Disparate Data into Gold
Thumbnail URL:
https://cadmoremediastorage.blob.core.windows.net/6cd0f52e-3135-47b9-9dc6-29d145c6dad5/thumbnails/6cd0f52e-3135-47b9-9dc6-29d145c6dad5.png
Duration:
T00H56M05S
Embed URL:
https://stream.cadmore.media/player/6cd0f52e-3135-47b9-9dc6-29d145c6dad5
Content URL:
https://cadmoreoriginalmedia.blob.core.windows.net/6cd0f52e-3135-47b9-9dc6-29d145c6dad5/PS22 September - Data Trends.mov?sv=2019-02-02&sr=c&sig=jFXqy5R1gKXrCc03oC8dsdvhcPKU3RmwK%2FA%2FyUP0eBk%3D&st=2024-09-17T22%3A27%3A25Z&se=2024-09-18T00%3A32%3A25Z&sp=r
Upload Date:
2022-09-09T00:00:00.0000000
Transcript:
Language: EN.
Segment:0 .
STEPHANIE: And with that, I am happy to pass it over to today's moderator, Dustin Smith, the co-founder and president of Hum.
DUSTIN SMITH: Sounds good. Thanks, Steph. Appreciate it. And thank you very much to our panelists and our guests who we're going to have a conversation with. This isn't going to be sort of ping pong around. We're really going to try to engage in discussion, and please do drop things in the chat and participate in the polls. I really want this to be involved.
DUSTIN SMITH: And I'll introduce panelists in a second and, basically, tee them up to introduce themselves. And I wanted to bring back to the actual title of this webinar, so "The Alchemy of Turning Your Disparate Data into Gold." And I'm actually going to read something very briefly. This is from the Outsell STM report that came out yesterday. And this is authored by Lettie Conrad. "The STM industry is awash with high-quality data of various kinds that presents a tremendous opportunity to improve decision making and fuel further growth in the industry.
DUSTIN SMITH: The two areas where Outsell sees strong potential relate to customer data and research data. By leveraging customer data, content providers can grow their audiences and improve engagement and retention while better meeting customer needs." So ultimately, what we're going to be talking about today is going to lean a little bit more towards the customer side of things, what is called zero- and first-party data, the sort of data that comes from the interaction of people and content.
DUSTIN SMITH: And that's really where the bulk of the data, the sort of gold that publishers sit on. And in some cases, it's 99.9% or greater. There are many people who are coming in interacting with content. And so ultimately, those are some of the things that we're going to dig into today. And so we have Melanie from SSP, Michael from PNAS, Margarita from IWA Publishing, and Joe from CyberRisk Alliance.
DUSTIN SMITH: And so why don't we go in that order? And so it'd be great for you to introduce yourself and your org and talk a little bit about where you are in your data-unification journey. That's one of the things that is very crucial in order to make best use of that zero- and first-party data. It's really pulling it together and making it actionable. So if you could talk a little bit about your org and where you are in your data-unification journey.
DUSTIN SMITH: We can start with Melanie and then go Michael, Margarita, and Joe.
MELANIE DOLECHEK: All right. Thank you so much, Dustin and Stephanie. I'm Melanie Dolechek. I'm the executive director for the Society for Scholarly Publishing. We are about a year into our journey of trying to pull everything together. We have quite a few different data sources. We are a membership organization, so a lot of what we do centers around our AMS, or Association Management System, which is very similar to the CRM.
MELANIE DOLECHEK: And we also have a very popular blog called The Scholarly Kitchen, where we're putting out new content daily. And so those are probably the two most critical data sources that we have. But we have others, right? So we have an online community, which is powered by Higher Logic. We have our marketing system, which is part of Higher Logic.
MELANIE DOLECHEK: It's called Informz. We have an on-demand library. We have a business directory. So we have all of these different sources, and we definitely had to work through a lot of that. But so we're very, very early still in connecting everything because that's something we'll talk about a little bit later as an interesting challenge. But that is where we're at in the process.
MELANIE DOLECHEK:
DUSTIN SMITH: Great. Michael?
MICHAEL HARDESTY: Great. Thanks, Dustin, Stephanie. I'm the digital product manager for PNAS, which publishes the flagship PNAS journal, as well as the new open-access journal, PNAS Nexus. I'm really excited to be here today and look forward to discussion with my fellow panelists. I even remembered to wear my Talk Data to Me socks that I picked up at SSP earlier this year. Yeah.
MICHAEL HARDESTY: So PNAS-- we are in a ideation phase of our data-unification journey. By that, I mean we're aware of the problems, but we don't currently have a solution in place to stitch together all of our different data points and really amplify the value of that data. We are planning to implement a CRM, and we've been exploring a data lake for our first-party data. And my role in that process is really to identify the outcomes that will positively impact our program and help articulate the vision for how we should be leveraging our data to reach our objectives.
MICHAEL HARDESTY: And whether that's supporting open-access models or using web metrics to drive product development, I'm really looking closely at how we can get the most out of our data.
DUSTIN SMITH: Phenomenal. Before we go to Margarita, Steph, why don't you launch the poll? We're going to ask, how many data sources does your organization have? So this is how many places are you pulling data from as part of your data-unification journey? So, Margarita, why don't you go and, then Joe can pick it up from there.
MARGARITA LYGIZOU: Hello, everyone. I am Margarita Lygizou. I'm head of Sales and Marketing for IWA Publishing. We do books and journals on water. And we started investing in data quite a few-- maybe three years ago. And when I'm saying investing, time, resources-- from human resources to our own time. We are a small team. And we knew that out of the, I think, 26 different data sources, we can get gold.
MARGARITA LYGIZOU: We needed to find out-- we had questions, and we needed answers, basically. So what we did, we used the data-warehouse-type solution, a funnel we put everything in. And now, at this point, we're in a position where we have action-oriented dashboards. It is brilliant-- brilliant to see. For example, we have an editorial dashboard that informs the progress of the editorial office.
MARGARITA LYGIZOU: We have a different one that we call it the Health Dashboard-- the Journals Health Dashboard-- where we see downloads. We see different aspects of the customer journey. And this is it.
DUSTIN SMITH: We're going to talk more about highest and best use of data and some of the key goals in a little bit. But good to sort of set the table here. Joe, you want to go next?
JOE HADDOCK: Sure. Hi, I'm Joe haddock, and I'm the Chief Digital Officer for CyberRisk Alliance. And we like to think of ourselves as a data-driven business-intelligence platform serving the cybersecurity community. We're a pretty new company. And so in many respects, I think we're like a startup. In terms of media, we have publications, podcasts, physical events, and also community that tends to circle around virtual and physical meetings for those sort of thought-leadership-type events that we do.
JOE HADDOCK: And so I'm responsible, basically, for all of our public-facing technologies and unifying that data, which has actually been a goal from the very beginning of the business. And so I think that we're actually fairly far along. We've been working with Hum a partner for about a year now. And I think that our primary behavioral data systems have now been combined.
JOE HADDOCK: And it's more about teasing out what are the actionable items on that. Obviously, I have very strong opinions about what some of the actionable items are. And then I think that we have discovery additionally, as we look at the data.
DUSTIN SMITH: We'll look forward to hearing those strong opinions, Joe. You did have a slide. Is now the time you want to show your fancy slide?
JOE HADDOCK: Oh.
DUSTIN SMITH: Or would you like to do that--
JOE HADDOCK: Yeah, sure. So when I was looking at the questions today before the call-- here I'll share my screen-- I was thinking that it would help to give a little context about the way I'm thinking about this. Maybe other people are thinking about it in other ways. Sorry, I'm trying to make this full screen here.
JOE HADDOCK: Here we go. So these are the systems that we are combining. WordPress represents our websites. We happen to-- all of the businesses that we have acquired have been on WordPress, so we are using WordPress as our CMS. It's a-- I guess it's a perfectly good CMS. It's quite popular. Everyone knows it.
JOE HADDOCK: Intrado is our virtual-event system. Auth0 is what we're using for our single sign on. Marketo is our email marketing automation systems. Swoogo is our personal events system. So this is where we actually do our larger virtual events and then our physical events. Convertr's where we actually validate and normalize and distribute all our third-party leads. Snowflake is, basically, our data warehouse where we put data for analysis.
JOE HADDOCK: And we're using Power BI as an analysis tool. And then Hum is, basically, sitting in the center and unifying all of our data for us. And that was it. I just wanted to give context about the systems. There are other systems. We're just not seeing them as essential for understanding the user behavior.
JOE HADDOCK: And I think this just gives a little context at least to how I'm thinking about the data. Do you have any questions or anything else, Dustin?
DUSTIN SMITH: Others may have questions, and you can feel free to pull this up if folks want to poke at it. But maybe you could drop the slide for now, and we could pull up the poll results. In the meantime, Joe, how many-- do you have a tally of how many systems?
JOE HADDOCK: Yeah, six. Six systems. Yep.
DUSTIN SMITH: OK. So the poll results-- looks like pretty evenly spread with quite a few in the sort of 5 to 9 and 10-plus. I have no idea. That's the biggest category, so very interesting. And, of course, one of the questions we might ask is, what are the sort of categories of systems which people are looking to integrate or where the most valuable sort of data lies.
DUSTIN SMITH: Steph, would you mind launching the next poll? And so this is talking about use cases for leveraging your data. And so this is seeing-- there are obviously a limited set of options here. So pick the one that you find strongest or most compelling. But I'm wondering if we can move to, actually, talking about use cases. What are the biggest problems or opportunities that your org see as, as you actually unify your data and look to make the best and highest use of it?
DUSTIN SMITH: So why don't we go-- Margarita, do you want to take a leadoff? You've been at this a little while. And where are you looking to take your data?
MARGARITA LYGIZOU: Yes, absolutely, I can. With us, the latest thing that we are really looking into-- last year, we became an open-access publisher through a new model, the Subscribe to Open model. And internally, we had the information that we needed. Everything was absolutely fine. I have to say here that, for books and journals, the information is completely different. So you have to have a different mindset. I'm going to be talking about journals for now. .
MARGARITA LYGIZOU: So being open access, it opens up different types of business questions. So we knew what we knew. So we knew our customers. And what we wanted to find out was, basically, what other people were using the journals. And they were not paying us any money, so this was a lead generation exercise. And we use [INAUDIBLE] for that.
MARGARITA LYGIZOU: And we have been able to find out people that basically use our journals quite a lot. They were not our usual customers, and we have been able to go and say to their libraries in a very, very traditional way. You're looking at their journals. Why don't we work together? So that's one of the main differences that happened once we became open access.
MARGARITA LYGIZOU: Pre-open access, what we were looking at was basically how many points a customer is coming into contact with, the different-- we have books, as I said. We have journals. We have IWA, the membership organization, the parent company. And we wanted to find out how many of these people are-- they're buying books.
MARGARITA LYGIZOU: They're also authors. They're in a university that has a subscription. They're members. This sort of thing. And it was quite eye opening. I don't know if other panelists can say the same information, but for us, it was like a completely new world opened up. And there were so many things that we started thinking, oh, we can resolve this, and we can answer this other question.
MARGARITA LYGIZOU: And that was the point, where we have to curate the questions, basically, because you can-- especially if you're a smaller team-- and we are a very small team-- we had to build on top of what-- start with something, and build on top of that.
DUSTIN SMITH: And just to pin some of that down, some of what you're talking about is how do we drive more product sales? How do we drive--
MARGARITA LYGIZOU: Yes.
DUSTIN SMITH: --the sort of membership funnel? How do we take people who are doing one thing and make the connection to potentially put something else in front of them and--
MARGARITA LYGIZOU: Absolutely. The first thing for us was to find out whether this was happening already because we had the information. It was just that we had to export so many spreadsheets-- it was spreadsheet galore-- and put all the information together. And it was taking time. So we had the information, but it was taking a lot, a lot of time.
MARGARITA LYGIZOU: So where we are now is that we open the dashboard. We're using Tableau, and we have all the information in there, which is good. So now, we are at the point where we say, OK, how do we have these people that do x to also do y? So the very beginning, depending on where we are, the other publishers are in their journey or other companies, you have to start with, basically, having a good idea, good base of what you have, and then build on that.
DUSTIN SMITH: Joe, do you want to talk about right message, right person, right time? I think that is one of the key things that you were thinking about at CRA.
JOE HADDOCK: Yeah, sure. The thing that I--
DUSTIN SMITH: Say whatever you want, Joe.
JOE HADDOCK: No, no.
DUSTIN SMITH: I don't mean to paint you into a corner.
JOE HADDOCK: No, no, no, no, no. The thing--
DUSTIN SMITH: But I think you do have some things to say about it.
JOE HADDOCK: The thing that I want to say is that I just think that-- I'm completely convinced that this is the evolution of the digital platform, is being able to harness this behavioral data to, as your question is, Dustin, give people the right content at the right time. So if you don't actually understand where they are, by being able to sort of generally bring this data together and provide them with that-- Margarita was just talking about this, right-- how are you going to lead them to the next thing or engage them more or-- sorry.
JOE HADDOCK: I'm getting my own way here. But just take a look at all the digital platforms out there right now, the primary platforms. This is what's going on. Whether it's YouTube or Facebook or Google, they're already doing this. And we have to move into this space of leveraging the behavioral data to present people with the information that actually helps them through the journey, right?
JOE HADDOCK: By helping them through the journey, we are going to actually engage them more. We can actually affect outcomes in terms of what their awareness is or what they're actually consuming. So, yeah, right content at the right time, Dustin, I think is really-- that is where this is going, right?
JOE HADDOCK: And trying to identify what systems are going to give you those signals. What is actually actionable behavior? And what are the biggest actions? What is going to actually influence things the most?
DUSTIN SMITH: Could you give an example, Joe, of the actual right content at the right time? How does that happen.
JOE HADDOCK: I set up a digital platform at HIMSS before coming to CyberRisk Alliance, and I have some really great examples. We're still pretty early with CyberRisk Alliance, although we're getting to that phase. But I would just-- I mean, as a really obvious example, at HIMSS, there are 3,000 CTOs in the health care space, and we had them all. They all came to our events.
JOE HADDOCK: They were all opening our emails. We were really well saturated in the marketplace. And we felt that we really knew the CTOs for the industry. But we would send them an email specifically crafted for them, and we would get a 10% open rate-- 10%. And that is saying that we actually know what they want. Whereas when we would send out [INAUDIBLE] to 3,000 people, 10,000, a 10% open rate.
JOE HADDOCK: But on a database of 2 million people, when we sent out the same emails behaviorally, we'd get 65%, 70% open rate. It's because people-- if you're actually looking at the behavior, if you're looking at what people are engaging with, it's transformative to your communication. Whatever you're messaging, whatever the thing is that you're trying to connect with them, if you can lean into those behaviors, it transforms your-- not just your marketing but your entire-- the digital platform is the communication platform.
JOE HADDOCK: I mean, I still love books, but most of my consumption is digital now.
MICHAEL HARDESTY: Just to jump on that point because I think it's interesting-- and maybe this is just overly simplistic and obvious to everyone else-- but when you're thinking about the users that have given you that zero- or first-party data-- your point, Dustin, about that interaction actually giving it to you-- they do it for a reason. And that's because it's an exchange for the value that your product provides them. So I think that the users that you-- those known users, they are more engaged.
MICHAEL HARDESTY: They're more likely to be loyal to your product over time. And just tying into the right message, right time, if they've given you that data, use it, and make that experience personalized or contextual, and don't ask them for the same thing again, right? So it's improving that experience and, hopefully, over time, growing that segment of your audience even more.
DUSTIN SMITH: Melanie, you got anything you want to jump in on?
MELANIE DOLECHEK: Yeah, I think our motivation was slightly different. I mean, obviously, driving revenue and getting the right user to the right content is a part of it, for sure. But I think we took it even back a step further to say, what do they need? What are they interested in? And we want to develop our programming around those interests.
MELANIE DOLECHEK: So, for instance, if someone is reading The Scholarly Kitchen, and we can see what topics are getting the most interest. And we have lots of ways we can see that. But Hum really brings it together for us because then we can develop something, and then we know who to push it back out to, right? So there's the program development and making sure that we're delivering what our members and our community wants to learn about, wants to read about, where they need education, and so forth, and then being able to turn around and target that right back to them and say, hey, we know you're interested in peer review, so here's some content on peer review.
MELANIE DOLECHEK: And some of that might be revenue generating through webinars and things like that. And some of it might just be more, hey, we're going to go out and find authors that can write about this topic because it's clearly something that our membership needs. So it's very mission oriented for us in terms of understanding your audience-- first of all, what their needs are, and then, from there tailor creating what they need and then marketing it back to them.
MELANIE DOLECHEK: And in essence, then, we're growing our audience in the process. So it isn't all just about marketing for us. In the end, that's part of it, but it is also about what programs do we need to serve our membership?
DUSTIN SMITH: Can you talk a little bit about the difference between just having sort of aggregate-level insights and knowing exactly a group of people who are interested in a particular thing, so you can develop-- if you develop a new, say, product or educational product or webinar, the ability to target those folks with messaging that kind of puts it in front of them.
MELANIE DOLECHEK: Yeah, absolutely. So we actually have a case study recently. We developed a workshop on open access. And we're promoting that to our normal list, but we didn't really have behavioral data to go with that. And so we were able to use data from The Scholarly Kitchen and marry that up with our marketing system and our AMS system and be able to target the folks that specifically have been looking at and doing a lot of reading specifically on open access.
MELANIE DOLECHEK: And then, in turn, many of those folks then were able to register for the workshop and get the content-- again, back to the right time, right content piece together. And so we wouldn't have been able to do that-- I mean, we would have known, yes. We knew people were reading open-access content on Scholarly Kitchen by looking at our Google Analytics and our article metrics and so forth through WordPress.
MELANIE DOLECHEK: But we didn't necessarily know who they were, and so we were taking a very blanket approach. And this allowed us to be much more specific and targeted with getting that content in front of the right people.
DUSTIN SMITH: Great. Steph, do you want to throw up the poll results? Nice mix. Pretty much even. So good stuff all the way across. Anything anybody on the panel wants to pick up on, on any of these, before we dig into some of the challenges, given folks want to get started on stuff and may need to wrestle with some similar thing?
DUSTIN SMITH: No? OK. And don't feel as if we have to go all the way around here. But one of the things-- especially Joe and Margarita, you've been at this game quite some time, with a few battle scars, I'm sure. But thinking about the biggest challenges in terms of unifying data, and certainly the folks who are in the category of we don't know how many systems, that's probably challenge number one, to find and identify them.
DUSTIN SMITH: But maybe you can talk a little bit about some of your challenges, how you've overcome them, and we can take it from there.
MARGARITA LYGIZOU: I can talk a little bit about challenges. Joe, do you want to go first? No. I'll go first. So can I?
JOE HADDOCK: Yeah, please.
DUSTIN SMITH: Please go. Oh, yeah.
MARGARITA LYGIZOU: OK, good. So the first challenge we had was resources and time. We knew we had a lot of places to get information from. And we knew that because every time we wanted to get a report for something, we had to go to all these different places to download the information and then put it together. But because we are-- and I keep on saying we're a small team because it's a different thing to have resources allocated in a bigger organization and a different thing having resources allocated to a very, very small team.
MARGARITA LYGIZOU: What we started doing was putting all the information in a funnel-- everything. We just put everything in. We wanted to have single-customer view as a first step. And we had a lot of cleaning to do. So for everyone that is attempting to do that, I would say allocate resources for cleaning.
MARGARITA LYGIZOU: Data health is very important. What we found was that we were putting information together. We thought we had cleaned everything, and then we were looking at the information and thinking, oh, wait a minute. That's definitely not right. And then we were going back to the source data. We were cleaning again, putting the information in. The second thing I would say is a challenge was, once we put everything together, and it was in a good nick, we had-- it was like a new world opened up in front of us.
MARGARITA LYGIZOU: And we're saying, oh, my god, this is amazing. We never had this information before, and all this information-- all the other information. So what we had to do, again, as a small team was to put steps as to what we needed to do first. Primary goal for us was revenue increase, as I'm guessing is for everyone or most of us in here. So we needed to make sure we are clear what we need answers for and what we should go first.
MARGARITA LYGIZOU: So I would say these were the main challenges, wherein I'm not saying that we are in-- we know everything now. We don't. There are many, many things that we still try to resolve. But we are in a much, much better position than us, you know, pulling our hair out. Oh, my god, what is happening? Why do we have over 20 sources of data?
MARGARITA LYGIZOU: What happened? Again, though, I have to say, because we have books, and we have loads of people selling books in different parts of the world, we had to go in and export information from different places for a smaller agent that sells in India in the same way as we had to go and export information from-- I don't know-- EBSCO was big, and it's for journals.
DUSTIN SMITH: Melanie-- we can jump in line in front of Joe-- do you want to talk about small-team challenges? Because I think, if Margarita is small, then you probably have a smaller team to tout.
MELANIE DOLECHEK: [LAUGHS] We do. We ONLY have four full-time employees and a couple of part-time contractors, so we are pretty small. And for sure, we would not have been able to pull data together without a unifying tool to do this for us and without Hum's support in doing all the integrations and all of those pieces. And so, I mean, I think without that, it wouldn't have been possible.
MELANIE DOLECHEK: And even, we had our reservations about how specific we were going to-- how much time were we even going to have to use the data in an effective way? Because we couldn't do custom marketing campaigns every time we wanted to send something out. And so it's still a challenge for us, is picking and choosing what we want to focus on, even though there's all this data there, there's the work of implementing it and then the work of actually putting it into practice and making the use cases for it.
MELANIE DOLECHEK: Probably one of the biggest challenges for us, though, actually had to do with privacy. So we were collecting a lot of data in all these different places, but we were doing it largely in an anonymous way. And so by pulling all of this data together, all of the sudden, it's not anonymous anymore. And privacy was a real concern. We have 250 volunteers that work on our committees, and they have questions that they wanted to have answered with the data.
MELANIE DOLECHEK: But the first thing out of everybody's mouth is, well, what's the privacy plan? How are we protecting people and their data? And so forth. And so beyond just the compliance issue of cookies-- so we ended up having to implement a whole cookie-compliant system across all of our systems because we didn't have that. We weren't doing that much collection of specific data.
MELANIE DOLECHEK: But beyond that, there was sort of the creepy factor, right? Our members were a little concerned about, what are you collecting about me? And who can see it? And are the volunteers on these committees going to be able to see it? And so we actually had to kind of come up with what our strategy was around privacy and making sure that that was part of the plan.
MELANIE DOLECHEK: And so we didn't even want to integrate systems until we had that piece down. So that was definitely a different perspective that we hadn't necessarily thought we were going to have to have so much sort of discussion around and planning.
DUSTIN SMITH: Joe, you can finally go. Thank you, Melanie. This was great.
JOE HADDOCK: Challenges-- One is actually, when we start talking about unifying the data, everybody has all of these systems that they want to get in there. And I don't believe that all of them actually belong in there or are actionable. So part of it is just sorting out what actually is giving us the type of signals that we want. And then the other part of this is that, to Marguerite's comment, there's seas of data.
JOE HADDOCK: And sort of understanding what is actionable? Why are we looking at this? Because I've sat on meetings where we name a lot of numbers, but I don't see them as actually being informing how we're going to work. And whether that's writing more of this type of content or a different take on this type of content, I find that-- one of the things that we've been observing this year is that we had a really strong engagement around cloud as a topic in cybersecurity.
JOE HADDOCK: And it's waning. We're not holding that-- even though we're writing as much about cloud, the interest is waning. And so, to me, the thing is that I think that everybody that was really interested a year ago has moved forward. And now they're in the cloud, and they're not interested the same way. Now you have to talk about the difficulties of the cloud or how to fix the things in the cloud, opposed to moving to the cloud.
JOE HADDOCK: And I think it's things like data elements like that that help you sharpen your content and understand, oh, our audience has moved past this. We need to write something new. But anyway, I guess it's actionable, and what are the right sources?
DUSTIN SMITH: Steph, do you want to launch the last poll? And Joe, you can keep going.
JOE HADDOCK: All right, I have one other thing on this. And that is that I find, a lot of times, my team doesn't want to use the data because we don't have everything 100%. And to me, even when you start connecting just the first parts of data, you have information that you can start using. And every time we'd come up with something new that we want to do, everyone wants to stop and wait for that. But it doesn't make any sense to me.
JOE HADDOCK: If you have the data, you can start using it right away. And in fact, some of our really early successes were on really very simple ways of how we abstract the data. And then we could see-- so for example, we were just using basic taxonomy on our content and being able to look at the data-- at the engagement through that lens of an abstract taxonomy.
JOE HADDOCK: So we weren't looking at the content anymore. We were just looking at it through the lens of the taxonomy. It had a really big effect on our engagement and our content, just that simple step right there. So that's the other big challenge, is that people want it all. They want everything completed, opposed to starting to work with what we have.
DUSTIN SMITH: Good. Steph, do we have poll results?
MARGARITA LYGIZOU: Can I add something to that--
DUSTIN SMITH: Sure.
MARGARITA LYGIZOU: --to what Joe said about-- we had a similar issue when we had the first part of the information. We were saying, OK, great. Now we can use it. But what we found-- and we did-- what we found out later, though, is that mapping was one of the most important things, which is, again, data health. To give you an example, we were putting information from different places that had university names for us.
MARGARITA LYGIZOU: And because we-- in some databases, it was done very manually. And University College London was University College London and then, in a different database, was UCL. It was not able to connect the two. So we had two lines, in theory, two different universities. They were not two different universities. There was one. So we had to go again and again and again and clean and clean, clean, clean, in order to have good, quality information, actionable information, to use.
MARGARITA LYGIZOU: So we did have the same thing. We didn't start with the first round of having the data because this is when we realized, oh, that must have some issues. And we had to go back, clean again, and clean again.
DUSTIN SMITH: Yeah, tying, Joe and Margarita, your points together, one of the things that we see is author- and editor-applied keywords sort of all over the place. And with water, you can have lowercase, uppercase water.
MARGARITA LYGIZOU: Exactly. Exactly.
DUSTIN SMITH: And so that's four different-- and you're basically spreading your signal among all of this dirty data.
MARGARITA LYGIZOU: Absolutely.
DUSTIN SMITH: And it's amazing the signal boost you can get once you actually have clean data. Though, to Joe's point, no reason to not do anything until everything is perfect.
JOE HADDOCK: Right.
DUSTIN SMITH: Data cleaning-- I think Joe would attest-- data cleaning is a lifelong endeavor and not something that you do, and it's done and dusted.
MARGARITA LYGIZOU: I wish it was that done, but no.
JOE HADDOCK: Never done.
MARGARITA LYGIZOU: [LAUGHS] Never done.
DUSTIN SMITH: Indeed. Michael, you have anything to add on the sort of challenges? Maybe the challenge to go from planning to action, if you're at that juncture right now?
MICHAEL HARDESTY: Yeah, I mean, everything that the panelists have said, I would echo. It sounds spot on. On the topic of data hygiene, that's just always a problem for us. And we found that implementing PIDs, like RORIDs or Ringgold, can really smooth out that process a little bit, particularly when you're going across systems, and you want to get that 360-degree view of all of your data, and unify it that way.
MICHAEL HARDESTY: So that's really positive. And on the point about signals and noise, I try to be really thoughtful about prioritizing the right data points and mapping those to our outcomes and coming up with those OKRs or those KPIs because otherwise, I imagine you just run the risk of just drinking from the fire hose of data, so to speak. And you do just miss a signal for that noise.
MICHAEL HARDESTY: So that's something that, starting out, trying to be really intentional about prioritizing. Maybe we do have over 10 or who knows how many data sources. But what are those sources that are really going to make a difference? And where does-- start there and then build out. And use the learnings you're getting from that process to inform where you go in the future.
DUSTIN SMITH: Yeah. One of the things we see is that there's sort of a core set of systems that you can randomly build out and make it more use-case driven, as opposed to, say, we're just going to wrap our arms around everything or suck in all of the data, which just takes time and attention and actually dilutes in some ways, to Joe's point of which ones actually have high-value actionable data.
DUSTIN SMITH: We're about to get to the question hour. Margarita has a supplemental point, but I do want to encourage people to drop questions in the chat. And then we can open it up to anybody who wants to--
MARGARITA LYGIZOU: I actually saw a comment in the chats. Paul, I think, made this comment, and I completely agree, about data literacy. It's basically the ability to see five different people-- for us, for instance, we have the dashboard. What we wanted was that whoever opens up the dashboard to have the exact same information out of it. So, basically, we spent quite a bit of time in visualization, making sure that what we are saying, I can understand it.
MARGARITA LYGIZOU: And then someone that is not working with the sales data all the time gets exactly the same information out of it. We have deeper views where you can have the information that would make sense for people that use-- I don't know-- sales-- all the time sales data or marketing data. But those that-- they're open to everyone. We really needed to make sure that everyone understands and everyone gets the same information from what they're seeing.
MARGARITA LYGIZOU:
DUSTIN SMITH: Phenomenal. So the question that just popped in-- can the panel share their experience in working within their org to identify and agree what they want from their data? And what questions are the most important to answer first? This is getting to the best questions that you might want to answer with, with your data. Anybody can feel free to jump in.
MELANIE DOLECHEK: So we're a very small staff, so our staff had the questions that they wanted. But we actually went out to our committees and created a task force with a representative from each committee to bring their questions to the table. So what did their committee need to do their work better to better serve the members? And we put all that in a spreadsheet and started prioritizing it. And I'm not going to say it was the best way to do it, but it worked to the extent of we're actually pulling some data out and doing some things with it.
MELANIE DOLECHEK: But I think that trying to understand what the questions are was kind of the hardest part because people didn't really understand what data was available, like, if you weren't working with it all the time. And I think that's just one of the challenges of a small-staff, volunteer-heavy organization, is trying to sort some of that out.
DUSTIN SMITH: Anybody else puzzled through that?
MARGARITA LYGIZOU: We are a bit bigger than Melanie. I was saying small, small, but we are a bit bigger. We are 15, all of us. [LAUGHS] But we followed a similar path, to make the we want, make things available. We always wanted to sustainably move to open access. And so over the last quite a few years, actually, we were making things available more and more, up until last year, when we became fully open access.
MARGARITA LYGIZOU: And then we wanted to make it sustainably. So income was a big aspect. So we took these two, and we said, how can we achieve this? And then we started putting, OK, we need x and y in order to achieve it. How do we get information about this x and y? And this is how we funneled, and we found the core of what we needed answers for, basically.
DUSTIN SMITH: We'll get to Hannah's question in a second, but a quick follow on this. For those who have run this a few times and aren't right at the beginning, are there places in which people have seen something work and have a spark for new ideas once they've seen certain types of data? Or something work in the wild that sparked a new question to answer or initiative or use case?
DUSTIN SMITH: Light-bulb moments. And if not, we can ask the follow-up question in the darkness.
JOE HADDOCK: I think that we've-- I don't know if I can name a particular data set off the top of my head, but we've definitely done a lot of product innovation that is based off of the data-- so looking at what's going on with the data, and then building products that actually wouldn't work without the data behind it. And even the lighter element products that we have or traditional media products that we have, we're able to enrich them and make them better by having the relevant data associated with them, which I think is an inevitable outcome.
JOE HADDOCK: You need to be able to provide the data behind the product.
DUSTIN SMITH: Can you go a half level deeper and just maybe provide an example of what-- something a product, or a media product, with data behind it or data underpinning it?
JOE HADDOCK: Yeah. So we have-- one of the things that we do is we'll take a number of media products that we have. So we sell webcasts. And we sell display advertising. And we sell custom content and different pieces like this, right? And what we would do is, actually, we would put together a program that leveraged all of these assets.
JOE HADDOCK: And we would meet with the client about their content. And often, we would do a content audit to understand what content-- what the journey is of the client is actually presenting through the content pieces and how they would tie together. And then we try and suggest, if it has holes in the journey, if we don't feel that it works, creating new content for that journey, to stitch it all together.
JOE HADDOCK: And then as we run the campaign, we actually meet with the client. And we're looking at the data and showing them, this is where the engagement is happening. This is where people are, and understanding these people are more in an awareness state, where this other group is actually doing research or ready to buy. So we are setting up a buyer cycle across the campaign so that they-- is this the kind of thing you want, that you asked for, Dustin?
JOE HADDOCK: I don't know if I'm-- [LAUGHS] But none of this would actually be-- stitching it all together and making sense of it is the data. And we have been able to meet with our customers and actually shown them they're going after the wrong people. Some of this stuff is crazy transformational, when you're actually convincing the client that they have a different customer than they think they do.
JOE HADDOCK: People often have very set minds about who they're selling to. And we can actually show them there's a different audience that wants your product.
DUSTIN SMITH: No, that's good, Joe. You did phenomenal. So, Paul, who is in the contention for MVP audience member with all the comments-- so thanks, Paul-- do the panel see publisher data as having the potential as a commercial asset in its own right, or is it best as a means of providing insights into how products and strategies are realized?
DUSTIN SMITH: Anyone want to jump in there?
MELANIE DOLECHEK: Can I go?
DUSTIN SMITH: Hum talks about data assets all the time. But go ahead, Melanie.
MELANIE DOLECHEK: Yeah. I mean, from our perspective, going back to the privacy piece, I don't think that it's something that we would try to commercialize or monetize necessarily. For us, it really is about serving our mission, meeting our members with what they need. So from our perspective, no. I mean, I don't think it's an asset that we would then try to monetize in any way.
MARGARITA LYGIZOU: Absolutely, the same for us, by the way. It is an asset for us internally but not something that we would monetize in a different way. But I--
DUSTIN SMITH: You may have a-- I just want to make a quick distinction here because when you think of data as an asset, the first thing you might think of is selling the data, so just basically wholesale giving it to somebody else for money. And I think one of the things Joe's-- I mean Joe's world class at this. But it's basically using the data to power new experiences or new ways to run your business.
DUSTIN SMITH: So it's not just the insights, but it's how you adapt, how you wrap that into the digital experience, which is maybe a little nuanced [INAUDIBLE]..
MARGARITA LYGIZOU: This is an asset for us then, internally, yes, because there are companies that basically get the information. And then they sell not the actual data but the intelligence from the data. We would not do that. But we would use the intelligence to inform our business internally, yes, absolutely. That's the beauty of it, isn't it? And one of the things I saw from someone else asking about machine learning-- we are starting, and we want to work together for this, the machine-learning tool.
MARGARITA LYGIZOU: We are interested in the technology that engages new communities. We have members-- the membership organization, the parent company. And there are 50 specialist groups. And we would very much like to have them, basically, serve the community, each one of these 50 micro groups with exactly what they want to read. But, yes, definitely an asset for internal use.
MARGARITA LYGIZOU: Absolutely.
JOE HADDOCK: I think that the data actually does have a value by itself and that there are pure data products that can be built. And I want you to know that I really value people's privacy. I'm not selling their privacy, and I don't ever sell anything without somebody's consent and permission. So we're very careful about that. We're GDPR-, CCPA-, CASL-, CAN-SPAM-- you name it, we're compliant.
JOE HADDOCK: We're not selling people's data. That's not the thing that-- but I do think that within those confines still, there is an opportunity to create pure publisher data products. And I think there's actually some really exciting data products that the industry needs.
JOE HADDOCK: When we publish, we're creating a community based on content. It's a conversation. And that conversation is telling us things about the marketplace. Those things are valuable. And I guess it's partly on how you're packaging it and how you're thinking about it. But, yes, to your question, Dustin, I think there are definitely data products.
JOE HADDOCK: Could do a whole call about that.
DUSTIN SMITH: I know. We could do a whole call. And maybe we will. Maybe we'll have you back for the roundtable. I'm going to talk a little bit about what's going to come. And are we seeing the working paper?
MARGARITA LYGIZOU: Yes.
DUSTIN SMITH: So a handful of things to come out of this session. It's nice to sustain community when you do convene it. And so one of the things that you'll be seeing in the next month, six weeks, two months, is a Silverchair-- a white paper on privacy and GDPR. So we're very much looking forward to that. What Hum is doing is releasing a preview copy of a white paper on first-party data and customer-data platforms and all the use cases.
DUSTIN SMITH: This can certainly be helpful in your data journey. We're getting the preview copy out to this select group because we want feedback-- what is missing? What do you want to know more about? And so, actually, I'm going to drop this into the chat in a moment here. And, Steph, if you have other sort of mechanics on follow up and some of the things regarding convening this group in the future, then feel free to do that.
SPEAKER: Yes. Fantastic. Well, thank you, everybody, for joining today. And thank you to all our speakers for lending your insights. Like I mentioned, we will be having a follow-up roundtable. There will be information in the follow-up email that you'll receive from this event. And you can also email strategies@silverchair.com if you'd like an invite.
SPEAKER: Meanwhile, the final event in the 2022 series is going to be held on November 9 and is going to cover content trends, giving digital meetings content the first-class treatment. So we hope to see you with that. And in the meantime, have a great rest of your day. Thanks so much for joining us today.
DUSTIN SMITH: Thanks, everyone.
JOE HADDOCK: Thanks.
MICHAEL HARDESTY: Thank you.