At its Google I/O event on May 7th, Google announced that it would take steps over the coming months to improve consumers’ control of cookies in its Chrome browser, and to block fingerprinting technologies. Details around the nature and timing of their plans have not yet been released. Our preliminary understanding is that cookie controls will not be turned on by default, limiting the impact of this change on our audiences.
At Dstillery, we have always held ourselves to the highest standards of transparency and privacy, and we support giving consumers more control over their online experience. That is why from the outset we engineered our audience building platform to use diverse sources of anonymized consumer events, and to not be reliant on any one type of signal.
As the industry evolves, we are confident that this diversity of data will enable us to continue to provide highly performing custom audience solutions for agencies and brands. As more specifics on Google’s plans emerge, we will provide updated views on any potential impact to our offerings.
Watch the recording of Chief Data Scientist Melinda Han William’s presentation from the 2019 Strata Data Conference
You’ll learn
Discover a new path to unparalleled consumer understanding based on fine-grained digital behavior using neural networks and unsupervised learning
Learn a novel approach to clustering in a sparse, high-dimensional feature space using semantic embeddings
Explore a three-dimensional visual mapping of the internet and the consumer’s digital pathways
Brands, marketers, and product designers need to understand their customers. Traditionally, market research was driven by surveys and focus groups of limited scale. Today, digital signals like web browsing behavior can provide a stream of observed behavioral data that is rich with information about a user’s interests, needs, and preferences. This type of data, coupled with machine learning techniques, holds the promise of freeing market research from the constraints of self-reported customer data.
In its raw state, web browsing data is both too detailed and too sparse to be comprehensible, let alone actionable. Melinda Han Williams explores semantic embeddings as a novel approach for understanding observed digital consumer behavior and details how to use a semantic embedding of web browsing behavior to drive unsupervised clustering for customer segmentation. You’ll learn how Dstillery has trained a neural network on 15 billion behavioral interactions. The resulting model can be seen as a much lower dimensional embedding of the internet and, if projected into two or three dimensions, as an interactive map. This taxonomy of internet behavior can be used as the foundation for a number of applications, providing unparalleled insights into consumer behavior and needs.
Back from the MIT Sloan Sports Analytics Conference, Caroline Allen, Peter Lenz, Peter Ibarra and Matt Sabban discuss how sports teams and advertisers can use audience insights to find a new generation of season ticket holders.
CAROLINE ALLEN: Welcome back to the DS Without the BS podcast, where we help demystify data science, AI, and machine learning for marketers. I’m Caroline Allen, associate marketing director here at Dstillery, and host of our channel. For our first episode of 2019, I’m joined by two Dstillers who you are already familiar with by now: Peter Lenz, senior geospatial analyst, and Peter Ibarra, associate director of insight solutions, as well as someone new to the podcast: Matt Sabban, our director of solutions engineering. They have just returned from the MIT Sloan Sports Analytics Conference, where they presented new research around fantasy sports, eSports, and recently legalized sports gambling. With March Madness, the MLB season opener around the corner, and even looking to the future with the upcoming 2020 Olympics, where should sports teams and advertisers invest their dollars and resources to find new audiences? So, welcome back, I know it’s been a while since we’ve been recording, but…
PETER IBARRA: It’s been a little bit. A little break.
CAROLINE: It has, it has, but I think the Sloan Sports Conference is the perfect time to re-launch the podcast and talk about what’s been going on here at Dstillery, but also what’s been going on in the world of sports, so, I guess to start it off, for our listeners who aren’t familiar with the MIT Sloan Sports Analytics Conference, can you tell us a little bit about the conference, and what is the significance of getting your paper accepted?
IBARRA: Sure. So the MIT Sloan Sports Conference, its thirteenth year I believe, really started out as a way for various organizations and teams and students to come together and learn about the latest happenings in analytics within sports, traditionally for on-field performance, everyone probably thinks of Moneyball. It’s really a conference that was centered around that. It’s grown in the last few years, I think this year they had over 3,000 attendees, where they’re talking about a lot of things, not just with the things that go on, on the field for any of these sports, but for the business of sports. How can they use data to have a better understanding of their audiences, how can they grow their fan base, what are the things that they can do to basically start creating more unique experiences for the fans?
And that’s kind of what I think this conference was really centered on, is the idea of how these organizations are building out their data infrastructures, and how they’re doing this in a way so that they can basically have a better connection with the people that follow their teams, their sports, or the leagues as a whole.
PETER LENZ: It’s the only conference in the world where you’re going to find the head coach of a major team sitting right next to a data scientist.
CAROLINE: And this isn’t the first time that you guys have been at a conference. So, a couple of years ago, which we actually talked about in a previous podcast, you guys went and talked about using location signals at MLB stadiums to validate sponsorship dollars. So, this time around, we mentioned that you were talking more about looking at the new avenues of sports gambling when it comes to eSports, fantasy sports, and sports betting, and your paper, I’ll read the title of your paper, it’s titled, “The Horizon of Sports is Digital: Using Fantasy Sports, eSports and Electronic Gambling to Find Next Generation of Ticket Buyers.” So what was your hypothesis around the paper, what were you looking to discover.
IBARRA: Yeah, really what we kind of wanted to do was use location audiences, similar to what we did last time, for MLB and NFL stadium attendees, and then from that, finding the new industries or channels that all these sports entities are starting to invest in. Fantasy sports, eSports, and gambling are kind of the three primary ones we identified, and understanding the behaviors between those groups and the people that we see going to games to get a better understanding of who the best people are to drive future ticket purchases. Matt here was kind of our expert in identifying those categories, especially the eSports one, I know he’s a big fan of that. So, yeah, Matt.
MATT SABAN: Yes. So, we chose gambling most likely because gambling is getting a lot of traction with the legalization in New Jersey. ESports because, if you look around the eSports world… eSports is basically playing competitive video games for those listeners who aren’t fans of video games, but it’s become a big thing over the past twenty years now, and one of the biggest eSports, League of Legends, garners, for their Super Bowl event, they have about 22 million concurrent viewers around the world, so that was another audience we really wanted to focus on. And lastly, we selected fantasy sports because that’s something that even non-sports fans can really get into.
LENZ: One thing that’s important to note here is that, when we talk about sports, we’re not talking about just the big, major league sports like baseball, football, basketball, we’re talking about something much bigger than that. This is a definition of sports that includes things like fencing and even poker or eSports. This is really inclusive of anything where people are paying money to be entertained by people who are very good at competing against each other. It’s entertainment business, it’s exactly the same as going to the movies, it’s exactly the same as going out, who’s going to buy what books. But here we’re doing it in term, the product is watching people do things.
CAROLINE: So you’ve talked about the audiences, but what was the actual experiment that you ran?
LENZ: Certainly. So we did what we do best every day: we created a set of audiences to study behaviors. So in this case, we designed three very special audiences. One around people who are around engaged in eSports, Matt is a big eSports fan and understands what people who are into eSports are doing online, so we used his knowledge to design an audience to collect data reflective of eSports as a very broad topic. League of Legends…
SABBAN: StarCraft, Overwatch, DOTA, something like that.
LENZ: Anything that someone’s going to be going out there and watching on YouTube as opposed to playing, we recorded data on that. We also had two audiences already in our system collecting data, one on e-gambling, and one on fantasy sports. In fact, we had multiple audiences for those. We had to select, of the audiences we’ve already built and are collecting data for our clients, which ones worked the best for this experiment. We took those three pools of audiences. To that we added people who showed up at MLB games and NFL games using our geodata. We know where all the stadiums are, we can see devices that show up in those stadiums.
So we took all of this together, and we smushed it together into one big pile of devices. 50,000 devices altogether. 50,000 because that’s what fits inside of our subpopulator. Subpopulator is a technology we have at Dstillery that lets us take a group of devices and say, “What do they have in common?” So normally, we say, “Here’s a behavior, like gambling, find me the people who are gamblers.” With subpopulator we say, “Here’s a set of devices, what are these devices doing?” So it’s a clustering technology, it tells us what people are doing. It’s up to humans to figure out what those behaviors actually are. It’s a way to let the internet tell us what people are interested in.
We took our 50,000 devices, ran it through subpopulator, and got a cluster of behaviors, ten clusters of behaviors. And this is the cool part, we went back to our original audiences, the three behavioral and two location audiences and said, “What percentage of each of these source behaviors is present in each cluster?” We then could tell you a story, “Oh, this cluster has a lot of gambling behavior and the people who are in it also went and did blah.” So we could ascribe back commonalities between the audiences.
CAROLINE: Say I am hosting the U.S. National Fencing Championship. What should I be doing as the marketer or as the company itself who is running this sporting event? How can I attract new customers, how could I find new audiences for this form of entertainment?
IBARRA: Yeah, no, that’s actually a really good question. It was a common theme that we saw at the conference, whether it was fencing or any other sport, where, really, these organizations are starting to build out their internal data infrastructures. They want to have a more direct connection with their fans so they can basically do exactly what you’re talking about: market to them, message them, develop content that’s specific to their needs. And that’s kind of what this whole thing is about, is using the available digital channels that a lot of brands are using every day and really bringing it to the sports world to develop those direct connections, and to develop and make more intelligent decisions, whether it’s for fencing or any other sport, really.
LENZ: One thing that these teams and organizations are learning, though, is that this data is a flood, and they’re not ready for that flood of data. They’re just massive-scale, when you’re collecting all of this different information about all these different people who are engaging with you. And that requires specialty knowledge. These organizations are very good at managing their players, they’re very good at figuring out the strategies for their teams, they know how to engage with traditional marketing methods because that’s what they’ve been doing for the last 50 years. But the new data sources are overwhelming to them, and one of the key learnings here is that they need help. Some of them have gone out and built out the expertise, they’ve gone and hired data scientists and technical people to be able to work with this data, but that’s expensive and difficult. You could instead go to a Dstillery, who has all that skill, has all that knowledge, has all of the data, and use that as a force multiplier for their own organization.
SABBAN: Right, and just to piggyback off of that, mining and gathering data is easy. Filtering out the noise is the hard part. And from the different brands that we saw out there, such as the NFL, the MLB… I like to use the NFL as an example, they’re putting microchips in the pads of all their football players, and I think they get about thirteen petabytes worth of data per game. Okay, so you have a lot of data, cool. How do you make sense of it? And, especially with this new context of all these different data points, you need a specialist there, and that’s just talking about the aspect of playing the game. How about marketing your brand? You have all those data points related to that, you need a different kind of specialist for specifically marketing or growing your brand, growing the fan experience, and that’s something that I think here at Dstillery we do pretty well, we’ve been doing for the past several years.
IBARRA: Yeah, I completely agree with that, and I think that’s where, as we were at the conference, one of the things that we felt we were actually pretty uniquely positioned as, is because we’re kind of a company that’s data agnostic. We can take in any sort of data, we can do an output of it, that can help brands in whatever way they want to go and utilize that. And I think that, as we see sports make these transitions over, there is a good opportunity for companies like us to come in and say, well let’s utilize all the things that traditional marketers have been doing, and let’s help you guys put together that infrastructure that doesn’t just take in the data and clean it, but brings out the meaningful things from it, so that they can start having those connections and really, the more important thing is monetizing all this stuff.
SABBAN: They need to draft us for their team.
CAROLINE: So, looking through these, you looked at three very specific audiences when you were doing this study, right? You looked at fantasy sports audience, eSports, and sports gambling. What are some of the unique characteristics of these different audiences that you found?
IBARRA: Sure. So, I’ll talk about two of them. Fantasy sports and electronic gambling, in our study, those were the two channels that we identified as being the best for future growth. Basically converting fans to that traditional ticket buyer.
LENZ: It should be stressed: that was what our goal for this experiment was. We were interested in exactly that, how to drive people into that traditional sports experience.
IBARRA: Exactly.
SABBAN: Traditional sports being Major League Baseball, or NFL.
IBARRA: Exactly.
CAROLINE: And how to get them into the stadium seats.
IBARRA: Exactly.
SABBAN: It’s putting butts in the seats.
IBARRA: And that’s what, as we were looking at this, how do we create those new fans, how do we use these channels to do that? And those were the two best ones. We saw things like, sports gambling had a really high overlap with golf. Or fantasy sports, you can kind of make sense that that had the strong overlap that it does with the NFL. And there were really interesting opportunities where they can understand the behaviors of the people that play them, whether they follow a specific sport, whether they actually go to these events, to create that unique content. And I think the one interesting thing, and Matt can expand on this, is how unique of a behavior profile that eSports had.
SABBAN: Yeah, so eSports was definitely the wild card. The whole impetus of this experiment was my favorite sports team, the New York Mets, their owners bought an eSports team for the game Overwatch. And I remember at the time thinking, “Man, the Wilpons are wasting their money again, oh my goodness. Why don’t they spend that money on an actual baseball players rather than a video game team?” But thinking about it, I was saying, “Hey, what if they’re actually trying to grow the Mets brand by buying an eSports team? What if they’re trying to get those fans to become fans of the New York Mets?” So we ran the experiment and it turns out that that’s not the case. ESports… So yes, the Wilpons are wasting their money again. It turns out that the eSports audience is a very distinct audience compared to fantasy sports fans or sports gamblers. They are really into the video games that they play. Another characteristic is they’re into, for example, Japanese anime and Japanese comics, or comic books, sci-fi… I don’t want to use the term “nerd” because it can be derogatory, but I’m a nerd.
LENZ: We’re nerds.
SABBAN: It’s okay. I’m a nerd, too.
IBARRA: We’re all nerds.
SABBAN: We’re all nerds. But they really like their games, and they really don’t care about traditional sports, or “sportsball”, as they like to say. They are not conducive to being ticket-buyers for the New York Mets, unfortunately.
LENZ: That said, it is a very good investment if they want to invest in where the overall world of sports is headed.
IBARRA: Yep.
LENZ: It’s not a legacy play, it doesn’t play into their core business of the New York Mets, but it helps them be in a place for a post-baseball future.
IBARRA: Yeah, I completely agree with that, actually. When we were looking at this from the angle…
LENZ: I hope we agree, we wrote the paper together.
IBARRA: That’s true. But I had the dissenting opinion. No, but I agree with that in the sense that, when you’re looking at this from the sense of, “Are they good for being a future purchaser of New York Mets?”, no. But when we shifted the question to, “Is this an industry that’s worth developing into the next Major League Baseball?” If you’re trying to compete on that level, there’s actually a really good case to be made that it is, and I think Matt touched on a lot of those unique behaviors, but they’re unique behaviors that also have a very large audience, and so if they can figure out a way to essentially make this into the next up-and-coming sports league, there’s a lot of interesting opportunities for them to do so.
SABBAN: Yeah. I just see a lot of untapped potential.
LENZ: And what’s more, these are behaviors that are young. When you take a look at the other two, fantasy sports and e-gambling, you see a lot of signals that read “older”. These are people who have been in the sports for a very long time, they’re already hooked, they’re already spending their money, but they are going to, over time, go away. And if you want to be where the growth is, you need to be with these eSports leagues.
CAROLINE: And that’s a conversation we were having earlier, too. I think one of the key takeaways that you brought Sloan Sports Conference was, when looking at the future of sports, and the future of marketing different, whether it’s baseball, or basketball, or football, sports teams and brands need to understand what value they’re providing and how to really harness on that, right? Why don’t you dive a little bit into the WWE and some of the things that we were talking about with that?
IBARRA: Yeah, that’s one thing that we, when we all walked away from the conference, and we completely agreed with. The WWE and how forward-thinking they are, and I know most people probably don’t think of them as being on a level of Major League Baseball or NFL, but they really have a good understanding that a lot of what their job is is to create content. It’s to develop new forms of entertainment so that they can have more direct connections to their fans, and so they can make more curated content for their fans and gain a higher level of engagement. And that’s really what it’s all about. And I think one of the most interesting things from the conference is, we were listening to a panel with all the leagues, and the WWE was there. NFL, MLB, they’re trying to figure out ways to create maybe one or two pieces of, a viral moment within a game. And they’re trying to figure out, “How can we create more of those things?”, right? And then, on the flip side, the WWE was talking about how they’ve developed basically a content creation or content marketing school that all of their performers go to down in Orlando, and they…
LENZ: Sports people, not performers.
SABBAN: No, that’s the term that they use.
IBARRA: Well that’s the term that they use, yeah.
SABBAN: The WWE considers their employees to be performers.
IBARRA: And the reason why is because they understood that, if we can teach our performers, our athletes to use social media, to create content that they can become their own entities within the league itself, and it can only be used to further the brand, further their engagement, and find the next niche consumer that may not be a part of the core brand but is a follower of a specific person.
SABBAN: Right. Content is king, that’s the message that I got from the WWE talk. And, on top of that, if they’re creating that personal connection, if you’re using an app like Facebook or, I don’t know if Periscope is still a thing…
IBARRA: (laughs) I don’t know.
SABBAN: But they are talking directly to their fans, and fans want more of it. They’re buying it up.
CAROLINE: And that’s something I think brand marketers are already very aware of, or they’re all jumping in that content game, whether it’s understanding who your key players in your companies are and building them up as thought leaders and knowing that people are going to be more trustworthy of your brand if they have a personal connection to them, if they feel like you have expertise in a certain field, if you’re a data company or a tech company. And as sports teams continue to evolve and grow on that marketing presence, they really need to harness what brands already do know about varying their marketing.
IBARRA: Yes. And I think that’s completely true, and it’s not like the other leagues don’t have a social media presence. There’s a lot of talk of athletes and how much they use Twitter, how much they use social media, and I think that the other leagues are trying to figure out how they can use those things to essentially further grow, whether its the NBA or the NFL. And I think, again, what was so impressive about the WWE is how they’re not thinking about that, they’re thinking about how they can turn this into an industry within itself, and it really is years ahead of what other people are doing, and it was just really impressive.
CAROLINE: Yeah. And also, understanding your audience and understanding who your speaking to is also going to then feed back into whatever content, whatever marketing you’re creating. So that’s just another, I think, key takeaway of all of this, and just knowing exactly who your audience is, knowing how to grow that audience, and knowing how to speak to them.
SABBAN: And that is the number one problem that, for me at least, speaking to eSports, because I sat in on in all the eSports panels, they want to know who their addressable audience is, they want to know who their fans are, they want to know how to make non-fans into fans. And ultimately, they want to create an experience where eSports, where playing a game becomes a tradition that’s passed on from parent to child, which I found fascinating because I’m a big gamer nerd, I never really thought of video games like that, but I’m thinking, “Wait, that’s actually pretty cool.” That’d be cool, if my kids played the same games I did, just like I learned baseball from my dad.
CAROLINE: Yeah, just like going out, having those memories of going to the stadium with your dad and you’re having your hot dog. Being able to sit down and share your team and share your team, and share your eSports team with your kids, I can imagine that.
SABBAN: So they’re trying to craft that experience now, and this is where I think that Dstillery can really help, especially with finding out who your audience is. We have the data for it, we collect more than 150 billion signals a day, we have a very extensive device graph, and although our experiment was just focused on answering this one question of, “Who are the best audiences to put butts in the seats?”, we could still use the same process and the same technology to figure out, “Who are the types of people that you can reach out to further grow your brand and to further enhance your fan experience?”
CAROLINE: So we have March Madness coming up, it’s in a few weeks, it kicks off. What, if any, do you have any predictions or any ways that you think that the NCAA is going to be harnessing, whether it’s sports betting or fantasy sports?
IBARRA: I don’t have any predictions on the actual tournament, that thing’s a complete, you know, it’s a complete mess every year which is why it’s so great. The one prediction would have and what I would love to see is how, this is the first year where there’s going to be a few states that have legalized sports gambling, and it’s going to be great to see the data that comes in from that event, because that is a huge gambling event on the calendar. And being able to use our systems to analyze that, maybe we could use that for another research paper.
SABBAN: We could. That’s actually not a bad idea.
LENZ: A metric bleep-ton of money is going to change hands, and it’s the very first time that it’s above the board for that to happen. And I know we’re going to have the entire experience instrumented in our system to collect a lot of data and learn from it.
SABBAN: Here’s a prediction right here: I’m going to be in New Jersey when March Madness is going on and betting on the games.
IBARRA: Yeah, just so you can place the bets, that’s it.
SABBAN: For research purposes, of course.
IBARRA: That’s it. Not for entertainment purposes only.
SABBAN: Right, right.
IBARRA: Yeah.
LENZ: I bet you could get that expense.
SABBAN: Yeah, maybe. I’ll try.
CAROLINE: We’ll have to get this podcast out and live before it’s out, it’ll be above the board. Alright, so, listen, I think the main takeaway from all of this is that, whether you’re a sports team, whether you’re a brand, whether your an advertiser or marketer, understanding more about the data that you have and the data that’s available to you that you might not have in your system already but you can have access to is only going to strengthen your relationship with your customers and your clients, it’s going to help convert new fans, new ticket sellers. But at the same time, it’s still a new world for a lot of people.
IBARRA: Yeah, no, it absolutely is. And this isn’t the only industry that’s happening in. We can see this with CPG, where more direct-to-consumer brands are coming up and challenging some of the bigger CPG brands that don’t have the direct connections, and so you’re starting to see some of those brands try to get more of those direct connections and develop things that do that. So yeah, this is something that’s going to be relevant across the board, and I know Lenz would agree with this, data is probably like the new oil, it’s going to be the commodity that everyone’s looking for, and so the question right after that is, “What do we do with it?”
CAROLINE: I think, as usual, you guys are more than happy to have these conversations offline if any of our listeners are interested or if anything has sparked their curiosity, you can just shoot Pete, Pete, and Matt a note, their email is analytics@dstillery.com. And if you guys want more information on the work we’re doing here at Dstillery, if you’re interested in seeing how our audiences can help you make more strategic business decisions, whether you are a brand or if you’re the NFL, WWE, thing same, give us a call. You can follow us on Facebook, facebook.com/dstillery.intelligence, or hit us up on Twitter and Instagram @Dstillery. Don’t forget, that’s Dstillery without the “I”, D-S-T-I-L-L-E-R-Y. Talk to you soon.
If you google “Women in Tech,” you’ll find countless articles, interviews, and conferences dedicated to the contributions of women in the fields of computer science. You’ll also find statistics highlighting the gender gap between professionals in these careers: women only occupy 25% of tech jobs. While the numbers are startling regardless of how often you hear them, perhaps we’ve created too limited a view on who belongs in this community.
As a woman at a data science company, I’ve seen how women from marketing to finance and engineering (and everything in between) impact the technology world. We may not be writing code, but we are teaching marketers and advertisers how to make sense of behavioral data and how to use technology to discover new audiences. We understand the pain points of digital advertising and market research, and work with product and development to create new ways for technology to solve these problems.
In celebration of Women’s History Month and an effort to redefine “Women in Tech”, I’ve asked the women of Dstillery — DstillHers — to share their unique perspectives as women in a technology company. What have been the keys to their success? What advice would they give to young women who are interested in tech? How has their mentor helped shape their career?
Being a woman in tech can be a lonely journey in a male dominated world, but it is a wonderful one. Believe in yourself, trust your voice, and know you can make a difference to the world of bits and bytes.
My advice is to embrace a winding career path. The job I have now wasn’t on my radar when I started my career. I changed directions by doubling down on work I found interesting and my willingness to learn something new opened doors. A computer science degree isn’t the only qualification for a career in tech. All work experience is relevant and unique to your growth as a professional.
The really exciting thing about working in the tech industry right now is that it is always evolving and changing. By nature, women tend to be very nimble and this presents us with a tremendous opportunity in our space. Be willing to roll up your sleeves, learn something new, ask questions and challenge the norm (even if it falls outside the scope of your day to day position, get curious!) and you will set yourself up for great success.
The one piece of advice I’d give to a woman looking to join a tech company is not to be afraid. Get involved in that project that’s piqued your interest, share your new idea with your teammates, or simply ask that question that’s been on your mind today. Speaking up and putting yourself out there will lead to success, not only for you as an individual, but for your entire company.
I believe that intuitive intelligence and thoroughness that a woman brings with her, has always helped my approach at work. I am also grateful to have mentors in my life who not only taught me to believe in myself, but also showed me how to be more assertive. I would urge all young women out there to believe that they have the power to achieve great heights and make a difference.
I encourage all women to explore the tech industry and all that it has to offer, even if you aren’t a “techie.” The tech industry is ever growing, jobs have nearly doubled in the last 20 years. As a woman, we need to be aware of the career opportunities from this leading growth industry.
My advice to young women who are interested in tech: ask questions. Allowing yourself to admit you don’t know something and being willing to learn will snowball into growth. I was previously our Technical Account Manager, writing queries and scripts. My team and department were male-dominated and I felt overwhelmed…until I didn’t. You’ll get there and you certainly don’t need me to tell you, the future is female!
One key to my success has been never accepting mediocrity and always pushing myself harder than the previous day. My mentor is Pat Summitt who was the coach of the Tennessee Women’s basketball team that I worked for during my time at UT. She inspired me through her hard work, dedication & leadership…all things I aspire to do every day in all parts of my life.
Some of the keys to my success at Dstillery have been raising my hand, reminding myself that if I’m comfortable, I’m not growing, and realizing that it is better to ask a million questions and get it right rather than being embarrassed that I’m asking a “stupid question”. And also remembering to never be intimidated – a girl can do ANYTHING a guy can do, so don’t let them always take the seats at the front of the room in a meeting!
Tech is always evolving and I lucked out early on in finding a mentor to always run any important decision by throughout my career. With any important decision I was faced with, I’ve always used her as my go-to and she’s helped lead me to where I am today. I’ve been forever grateful and try to give back as much as possible so other aspiring Women in Tech can feel like they have someone to act as their go to. Finding a mentor of your own can be a huge key to success, so ask questions to the people you meet and remember, we all started somewhere.
I grew up in the equestrian sport where men and women are on an equal competitive playing field. A skilled equestrian requires focus, physical and mental grit, dedication and the ability to guide their teammate (horse) to complete the required task with ease. These are traits that I bring to the table every single day. My advice to young women in tech is to speak up. Not everything you say has to be a profound statement, but say something. Ask questions. Show your interest with body language. Make your presence in the room known. You are listening and contributing in your own style.
The biggest key to my success lately has been embracing the unknown as an opportunity to raise my hand and put my name on something brand new. I like to think that we’re navigating uncharted territory at this company and in this industry. If you sit and wait until you have all the information you *think* you need before taking the next step forward, you’ll stand still forever.
In my career I’ve followed what’s called a “greedy algorithm”. A greedy algorithm means that at each step, you make the move that optimizes your objective function. The objective function I’ve chosen to maximize is learning. As a result, what I’ve learned has depended heavily on the opportunity that presented itself, whether that means low-dimensional electron transport, machine learning at scale, managing tech teams, or developing business strategies. Following a greedy algorithm does not guarantee that you’ll end up with a globally optimal result, but with the right objective function, it does guarantee that you’ll have fun along the way.
Brands have traditionally understood their customers using survey data and focus groups. In an era where those customers leave behind a digital record of their interests and motivations throughout the day, machine learning offers a route to a richer understanding of those people and their relationships with the brand
Chief Data Scientist Melinda Han Williams recently joined the UX + Data Meetup to share Dstillery’s approach to customer segmentation using digital observed behavioral data, in particular web browsing behavior across millions of sites. A key challenge to working with such high dimensional data is finding a representation and visualization that allows us to effectively make use of and intuitively understand the data. She explores how Dstillery trains a neural network on billions of behavioral interactions to create a semantic embedding of web browsing behavior. The result is a map of the internet, which they use to drive unsupervised machine learning for customer segmentation. Watch the session below.
The General Data Protection Regulation (GDPR) is a new comprehensive data protection law that updates existing EU laws to strengthen the protection of personal data. It replaces the current patchwork of national data protection laws with a single set of rules, directly enforceable in each EU member state. The GDPR took effect on May 25, 2018.
Dstillery’s Response
In response to the GDPR, Dstillery immediately blocks any information that arrives from a device originating from the EU. No EU user data is persisted in Dstillery servers, and Dstillery does not attempt to do any OBA to EU based devices. Dstillery’s general approach is to not collect data from EU data subjects. We block it prior to ingestion to our platform. Prior to May of 2018, Dstillery was not particularly active in the EU and didn’t derive significant revenue from the EU. As such, the recent passing of the GDPR has not had and will not have a financial impact to our business.
FAQs
1. Which, if any, of your location data collection methods are impacted by GDPR?
Dstillery only collects location data from devices located in the US and Canada.
2. What data collection policy changes are you making in response to GDPR?
All data that is determined to have originated from the EU is prevented from entering our systems.
3. Does this policy differ within the US and Canada? If so, how?
Dstillery’s US and Canada data collection and use policies have not changed and remain in accordance with the company’s privacy policy, applicable law and industry self-regulatory codes.
4. Are there any types of data that you are no longer able to collect within the US and Canada? If so, please explain how this will impact/change the way you build your audience segments.
GDPR does not affect Dstillery’s data intake within the US and Canada and therefore has no effect on the company’s audience building capabilities.
5. How do you anticipate GDPR impacting audience sizes? Please provide % estimate.
GDPR does not have an impact on US and Canada audience sizes.
6. How are you collecting data tracking approval from consumers in response to GDPR? Is this a new process? Is this being done in the US and Canada?
The GDPR imposes a strict opt-in consent standard for processing precise location data. We don’t anticipate a similar ruleset to be released in the U.S. at any time in the near future. Under the Digital Advertising Alliance Code in the U.S. and Canada, companies that seek to use precise location for ad targeting purposes must obtain reasonable assurances that upstream partners have obtained user consent. Dstillery obtain such assurances from our upstream partners in accordance with the DAA Codes.
7. If you have a measurement offering, please explain how GDPR will impact that capability.
Not applicable to Dstillery. We will not collect data on EU data subjects on the platform. All of Dstillery’s offerings are unaffected by the GDPR, as Dstillery does not leverage any data determined to have originated from devices located in the EU.
8. If we syndicate European audiences to another delivery platform does the onus shift to the delivery platform or is Dstillery still liable as they built the audiences?
There is potential liability, so we are avoiding any situations where we could potentially capture/use known EU data or be seen in a position to question whether we are. We do not do any manner of audience building using data from devices in Europe.
9. Can Dstillery deliver in Europe if we use a 3rd party platform and 3rd party audiences?
No.
10. Are EU visitors to the US and Canada or EU citizens living in the US or Canada impacted by GDPR?
EU data subjects who happen to be visiting the US and Canada are generally subject to US and Canada law. As a practical matter, the only way we can ascertain jurisdiction is by an IP address lookup and/or lat/long passed to us via partners.
If you have additional questions, contact us here.
