Welcome back to another episode of the PPC Show, where we interview the best and brightest in paid search and social media advertising. This week I'm joined by Sayf Sharif, Director of Analytics at Seer Interactive.
I attended his session at Hero Conf London and left with my mind blown. So I asked him to come on the show and talk about how to build advanced Google Analytics audiences.
Stay tuned as Sayf walks through how to create advanced Google Analytics audiences based on a variety of user behaviors, to target not just people that view a specific page or product, but people who have complex behaviors, like viewing more of content A than content B over 6 months, while also not subscribing to your newsletter, and having never visited your lead form, and who have never clicked on an ad before.
Listen to the Podcast
I'm a driven and fun guy, specialized in digital marketing and Google Analytics, helping organizations get the most out of their marketing dollar using data-driven strategy. Whatever tools you use for your email marketing, or your content, or your paid ads, or your social media, or your a/b tests, what interests me most is to help you make it better.
Past Speaking Engagements:
- HeroConf, London
- SMX East, New York City
- Meeting of the Minds, San Francisco
- eMetrics, Toronto
Sayf Sharif: I care more about how people are actually behaving when they're out, when they get to the site, so that we can understand their actual behavior and whatever that extra chunky content or product is, that they actually want. Then we can understand that and then we can bring that back up the food chain back to channel and say, "What campaigns are driving actual viable customers, or what organic content, landing pages are working or whatever it is."
JD Prater: Welcome back to another episode of the PPC show where we interview the best and the brightest in paid search and social media advertising. I'm your host JD Prater. This week I'm joined by Sayf Sharif, Director of Analytics at Seer Interactive.
I attended his session at Hero Conf London and left with my mind blown and so I asked him to come on the show and talk about advanced Google Analytics audiences. Stay tuned as Sayf walks us through how to create advanced Google Analytics audiences based on a variety of user behaviors to target, not just people that view a specific page or product but people who have complex behaviors like viewing more content A, than content B over the last six months while also not subscribing to your newsletter and have never even visited your lead form or have never even clicked on an ad before.
Let's get to the show.
Sayf, welcome to the show many.
Sayf Sharif: Thank you. Glad to be here.
JD Prater: We brought Sayf on to the show because we were both at Hero Conf London and he's giving this presentation around what we're going to get into, but around these Google Analytics, really defining behaviors. I went in thinking, "Oh Google Analytics stuff, I can always learn more." Thinking, "I'm probably like a seven."
I leave your session and I'm like, "I'm like a three. This guy is amazing." Thanks for coming on and talking to us. I'm really excited for this but, give the listeners a quick preview of who you are and your background of where you are, man.
Working at Seer Interactive + Background
Sayf Sharif: Sure. I just want to say, my goal in London was to make people feel terrible about themselves. I'm glad that I succeeded.
Yeah, I'm the director of Analytics at Seer Interactive. Most people are familiar with Seer, from Wil Reynolds and his talking at various conferences and things like that. We do other stuff too. I have been here for about two years and we really have been trying to leverage analytics in a way for our paid teams and our organics search teams where it's not just about audits and tracking. It's about how can we provide value for our customers. That's the whole point of this stuff.
That talk I was giving there in London was about what we're doing with, what we're calling behavior driven personas. We had called them quantitative personas but that was a little too nerdy for some people so we're saying behavior driven, because it's what are people actually doing and how do we determine who's valuable and then how can we use that to do everything from testing to targeting to personalization and the whole nine. That's where the value is and people really connect with that. I'm glad you liked it.
JD Prater: I really liked it. I've been a fan of you guys' blog for probably, jeez, probably three or four years now. I thought you guys did a really good job with SEO, the organic stuff. Then, I got to meet Gil of course. Gil Hong was on the show this last summer. Go back and check out that episode, he was really talking around shopping and stuff, getting ready for I think state of search was kind of where he was getting prepped for.
Sayf Sharif: Yeah.
JD Prater: All right man, let's jump into this. I'm actually really curious. How did you get your start into Google Analytics before we even stop going into this behavior personas? How did you get started?
Sayf Sharif: A lot of times ... I used to have a intro slides about me before my speeches or, my presentations and Wil basically convinced me that, no one wants to hear your background, no one care. Let's get right into the meet.
But, I actually went to grad school for archaeology and worked as an archeologist, which is the normal track. Anthropology is actually one of those area that does feed a lot into analysis and understanding behavior. I'd always been a tech person and I left archeology early on when the web really first came about and started off doing front end development, things like that. That really bled into needing to understand what people were doing and what was successful.
I started actually working with Urchin, pre Google Analytics way back in the day and having to do lots of fun analysis with that, where you combined data sets and step away from your computer because it would be chugging for eight hours on your local drive. It's nice now in modern days where I don't have to do that kind of stuff.
Yeah, the way I always talk about it is, what we're doing on analytics, it's just digital archeology, because archeology is, you're going somewhere, you're looking at the evidence that humans left behind, like the behavioral evidence and those data points. Then, you are trying to understand what they were doing and what they needed and what their problems were and we're doing that but with websites or mobile apps. It's just digital archeology so I'm kind of doing the same stuff I studied for in my grad classes except I'm using a computer instead of a trowel and a bull whip.
Digital Marketing Moneyball
JD Prater: Nice tie in man. I think, one of the things from the presentation that I really liked is how you prefaced it was Digital Marketing Moneyball.
Sayf Sharif: Yeah.
JD Prater: Talk to us about that. How you can find what actions are leading to this value.
Sayf Sharif: Yeah, like I said in the talk, I think I have a slide that has Brad Pitt. To have a handsome Brad Pitt picture. You can't go wrong.
JD Prater: Yeah, I know.
Sayf Sharif: Yeah, Moneyball, I'm sure probably most of your listeners are familiar with it, but it was the book by Michael Lewis and the film with Brad Pitt and they're talking about how athletics really got ... We're able to compete against people like the Yankees who had ten times as much money, by focusing on really putting their money on the most valuable players they could get for the least amount of money. Looking at what statistically made someone valuable in baseball, which was getting on base, which lead to runs, which lead to wins.
What we're trying to do when we look at things like customer lifetime value, customer centricity, behavior, things like that, is to say, if we're going to segment for audiences, the first thing we want to look for, is who, the valuable audiences are and what their behaviors are, because, those actions lead to that value.
WE're calling it Digital Marketing Moneyball partly because, you say, "Well, your page value for your URL ..." What's page value. You start talking about these kind of things but, when you say Digital Marketing Moneyball, people, I think understand more, it's about we're trying to find the valuable people for the least amount of money possible.
You know what, if we can ... Everyone that's listening to this show probably understands the more, segmented, targeted the audience, you're going to work with, you're probably going to be, it's going to be cheaper and more valuable to you, just from, we're going to re-target everybody that hit our site to, we're going the re-target the person that looked at a shoe with a shoe ad, is probably going to have a better success rate than, just targeting everyone at the site for a general level.
The more we can get into that complex user behaviors to understand, this is the kind of pattern that someone's following where they're actually valuable and how we can do things like, like, I said, targeting testing personalization, whatever that is. That's Digital Marketing Moneyball.
JD Prater: I like it. I thought it was a really good analogy. I think it's something that we can call definitely relate with. Yeah, kind of walk us through the next step. I think we're all on board that we need to be doing more of this. I definitely see it for 2018, we're all going to get more sophisticated and we need to get more sophisticated as marketers not just because we need to but it actually drives better returns. Right?
Sayf Sharif: Yeah.
Understanding User's Paths & Actions In Google Analytics
JD Prater: For our clients, for even our internal I house team. How do we get stared with some of this Digital Marketing Moneyball and understanding these paths, these actions that are really creating value?
Sayf Sharif: Yeah. I mean, so, obviously there are tools. I'm not a huge fan of tools, because I feel people like spending money on tools rather than actually doing any work. The tools don't always work that well. I will mention that there are tools that you can use to help you do this but, I like doing things for free, when you can do it for free. It's one reason like, free Google Analytics. Like, why would I pay someone to do my job for me.
I think, the first thing I've been telling people when they say, "How can I get started doing this?" Is to say, "Okay, well, first of all, you need to have your actual conversions in GA." We need, whether it's eCommerce conversions in the eCommerce revenue or it's goal completions on some sort of goal value. If you don't have a value associated with it, it's difficult.
Sometimes that means to being some sort of close loop implementation where you've Sales Force using the measurement protocol hit back to GA with the actual value often conversion or maybe just say, "F it. I'm just going to do, you can average page value of something like that." Fine, whatever, but you got to have a value.
Once you have a value and you have conversions, the first thing that I do before anything else is to just start segmenting based on that behavior. I want to segment out true bounce visitors. People that ... I'll have scroll tracking implemented so I can see where people are scrolling. I want to look at people that bounced and didn't even scroll. People that hit that page and were gone within five seconds and didn't scroll. It's almost on every site, I just say like 20% of your visitors, it's a mis-click. I want to just get rid of that first and segment that.
Then, I want to look at high value people. I want to look at the people that converted. Show me people, give me a segment of everyone that converted and then give me a segment that is neither of those. People that were engaged to some level, probably hit multiple pages, but they didn't convert. Now I have this bad traffic, good traffic, and the traffic that they were engaged but they didn't convert. Then I can start looking at what is the differences in those behaviors.
Even with that, sometimes we see, "Oh the conversion rate actually, we thought Facebook converted terribly compared to the Google organic or whatever but, when we look at the valuable customers, it's a much higher conversion rate." We actually are getting people coming from social that they're really converting high once they're, of the valuable component. How do we target those people? What are they clicking on? What are they coming through on compared to the low value people? What are they engaging with content-wise? That kind of stuff.
Then, we can start saying, "Okay, this kind of behavior ..." I think my examples, I had in the presentation, you have these examples like, it's really obvious. Someone views a video versus doesn't view a video, they're more likely to convert. Maybe it's also, if you look at this you can start saying, what content is more likely to be viewed by people who are high value, versus low value. It might not just be getting the one that gets the most pages, it might be something else. Then, you can start saying, "Okay, I want to target people specifically for those products or those pages or those services, or whatever that is."
Then, at a deeper level, you can ... That's just like the first level of analysis but then once you understand that behavior, you can start getting into the user scoring stuff that I had talked about where you're saying, "Okay, well I want to understand my commercial versus my residential customers or my do-it-yourself customers versus the people that are going to hire a contractor." Then, you can start doing things like adding additional tagging and tracking and understanding directly into the data. When you want to segment you can say, "Show me everyone that has a do-it-yourself score higher than five or something like that."
JD Prater: I really liked that part and we should definitely get into that. I want to, one thing that as you're kind of getting into this, it reminded me. It was like one of those hot takes. It was this like, "I hate personas." You're like, "What? This guy is an analytics guy. What does he mean he hates personas?"
Why don't you go ahead and explain to us, why you hate personas and then, the next part of that, that I really liked was your Prego story. I'm just going to tee you up on those two.
How Prego and Target Got It Right
Sayf Sharif: Yeah, the prego story is not fully mine. I definitely got to give credit to Malcolm Gladwell for that one.
JD Prater: Sure.
Sayf Sharif: The personas stuff, I will say that this brand personas the, Dedicated Dan, the Social Sally, that kind of stuff, I am not a fan, but I understand how in certain situations they're useful for people that are doing brand design and things like that. I understand why they exist, I just think they're over used in ways that shouldn't be.
If you start saying, "Well, we need content for the site. What kind of content would Dedicated Dan like?" I think that's a mistake, because then you're ... We can actually look and see people's behavior. We can actually do user research and people are not ... You won't fit into four or six categories. I think that I said this, "Why is it that brand personas always seem to be able to fit on one slide?" Because, is that our limitation?
I think I told the story, the Target story about how, the pregnant teenage story from Target, which I don't know if other people know but, how the dad coming into Target and said, "You're sending stuff about pregnancy to my 16-year-old daughter. It's inappropriate. Blah, blah, blah." Then, two weeks later came back in and apologized to the manager of Target because he complained about the stuff to the manager of Target of course not marketing. because his daughter was pregnant. He came and apologized. "Actually, yeah, she's pregnant."
Target identified this girl, the computers, not the people, identified this girl as pregnant because she bought cotton balls, unscented lotion and a large bag that wasn't necessarily a diaper bag but it could double as one. There was three products and the Target computers, was like, "Um someone's pregnant." Because, it looked at the patterns of those behaviors of people purchase these things, then, three months later they're purchasing diapers, therefore this person is probably pregnant. They don't have a pregnant teenage girl persona but they targeted her anyway. They targeted her based on her behavior, her patterns, what she purchased.
When we limit ourselves to these personas, that the brand person comes up with, it fits on one slide, I think we start losing value of the fat that we're complex people. We don't all fit into some single persona.
That's the persona thing. It's good to start but I'd rather go ... That's why we talk about the behavior of different personas. It's about behavior. Personas are actually based on human behavior and tracking enough stuff that we can make decision that way.
The Prego story, again, Malcolm Gladwell, he talked about it better than I ever talk about it. I don't know if there was a Ted Talk. Anyway, if you just Google Malcolm Gladwell and spaghetti you'll find it. Choice, happiness and spaghetti sauce, that's what it was called.
Basically, he was talking about Howard Moskowitz, he was the psychophysicist, which is the coolest job title back in 70s who tried to ... Pepsi hired him to figure out how sweet to make diet Pepsi. He tested a whole bunch of people and it just came back as noise. He expected a bell curve and there was a no bell curve it was just noise. It always bothered him.
Then Campbell Soup hired him for Prego and Prego was competing against Ragu at the time. This is in the 70s and everyone had Ragu. Anyone who's as old as me remembers the old commercials where Prego suddenly started competing with Ragu, and you had a commercial where you had two bowls of spaghetti sauce and Ragu, got poured on one and Prego on the other and Ragu just splattered everywhere and Prego sat right on top.
That came from Howard Moskowitz because he went around the entire country to all these fairs. He tested thousands and thousands of people. They had a hundred different attributes of spaghetti sauce, how think was it, how red was it, how sweet was it, how spicy was it, how much oregano it had? Every single aspect of spaghetti sauce. They came to the understanding through this analysis that ... They had them rate it, like, one to 10. Like how much do you rate this? They came to the understanding that, people didn't just like one kind of spaghetti sauce, just like they didn't like one kind of sweetness of Diet Pepsi. There was three main buckets they first came up with. They came up with regular, they came up with extra chunky, and they came up with extra spicy.
Prego immediately released, they started and made an extra chunky sauce, released those commercials and then captured a third of the market for spaghetti sauces, because no one else was doing it. Now, there are 72 kinds of Ragu. When ever you go to the store now and it's like, "Why is there so much everything?" It's because, well, because everyone's not you. Everyone likes things different. They know that now and they target the biggest groups they can.
You take kind of the persona stuff and you take that stuff, it's like, we all like different stuff. There's lot of horizontal segmentation. The other part of that though is that in all that testing no one ever asked for extra chunky. Right? Not a single person, all of that thousands of people that he tested with spaghetti sauce and put 10 bowls of spaghetti sauce, not a single person said they wanted chunky spaghetti in the sauce. They just rated chunky spaghetti sauce as higher and they captured the market.
Most of these personas are still based off qualitative research. "Hey, what do you want in a hand bag or what do you feel ... What kind of problems are you trying to solve." It's not that those aren't important but people don't always say what they want. They might say, "I want this." Then, you put them on a site and then they do something completely different, because people are dumb. Not, dumb but like, they're weird and you're primates and are just like strange shiny. You rely on this qualitative research.
Okay, it's a start but, I care more about how people are actually behaving when they get to the site so that we can understand their actual behavior and whatever that extra chunky content or product is that they actually want, then we can understand that and then we can bring that back up the food chain back to channel and say, "Okay, what campaigns are driving actual valuable customers." Or, "What organic content landing pages are working." Or, whatever it is.
Yeah, that's where it comes from.
JD Prater: Nice. I really like. I think everyone who's listening is now convinced that behavior driven personas are the way to go, because I like that you were talking around like, you're basically leaving a digital footprint. That's really what's more important because that's actually showing action more so than what you think you might want. You're actually doing these things.
For everyone listening, we're on board. We've got the Digital Marketing Moneyball. We've got where it's sold on digital or behavior driven personas. I'm trying to get it all, man.
Sayf Sharif: It's all right.
Mapping Digital Footprints in Google Analytics
JD Prater: Give me some next steps. Let's just say ... We talked about segmentation. Let's just say that we looked at ... We got rid of our crap traffic and now we're just really trying to diagnose, like let's get super in the weeds. Let's get real tactical. Are you saying like, let's go into the behavioral site content in GA. Is that how you would start looking for and connecting these dots. How do you start connecting the dots, I guess.
Sayf Sharif: A number of different ways. I think one thing that is one of the big features of Google Analytics that people don't use is the custom dimensions. In the free version you get 20 custom dimensions that you can put on users and sessions and hits to your hearts content. In the 360 version you can do 200. These are just different aspects of that data. I can have a page URL and ... I was actually talking to a ...
I did a finance webinar the other day and we looked at Bank of America and I showed how, from the home page of Bank of America, and the first page of results of Bank of America, checking account, I could get to close to 20 different URLs with over 30 different conversion links and buttons for a checking account. That many pages for checking account. There wasn't a checking account page of Bank of America. They had interest free this and they have whatever else, that. I don't know. I don't remember the names. Corporate checking account. They're all about checking.
There's content groups in GA and there is custom dimensions. We want to say, "Okay, I want to not just understand how this URL is doing. I want to understand how my checking account content is doing." Deeper than that some of the stuff was ... Like I said some was for personal, some was for corporate and corporations. Okay, I want to know what's checking account and I want to know whether it's personal versus corporate. Some of this stuff was targeted towards parents of teenagers. Like opening up accounts for children or things like that.
We said, "Okay, this stuff is ... This is stuff that's targeted at parents." We don't necessarily know their parents but we're probable that this is targeting for parental stuff." Maybe it's like, "How do I figure out your own savings for the future." We say, "Okay this is someone who is like proactively helping, proactively planning or financially responsible."
If I wanted to go into this site and say, "I want to understand valuable users or I want to understand who the behavioral persona is here." If I wanted to say, "I want to understand how parents of children are looking at checking accounts." To do that is possible if I just had raw data but it's a lot easier if on every single hit there I'm saying, "Okay, on this page, that's targeting parents of teens, I want to also immediately type dimensions for checking, for personal, for probable parent." Things like that.
Then I can start saying, "Okay, how does the content for parents do versus the non parent content." "How does the personal do versus the corporate?" "How does the stuff that displays urgency ...?" You might have a landing page that's displaying urgency.
The example I have for this is not financial but we have the, we work with Trex who builds the composite decking and one of the problems people have with non-composite decking, which is the wood decking is, splinters. You get splinters in your feet from your crappy wood deck. If someone's searching for content about splinters from their deck, there's an urgency there. No one in their right mind sits there with a perfectly good deck and goes, I want to Google splinters in my feet. If you understand this content has urgency ... I need to file my taxes before ... I have one week to file my taxes or whatever it is. You can say, "Okay, there's a mood." There's aspects like that.
Once you understand those things then we can even start assigning points and scores for those. We can start saying, "Okay, I want to have a frustration score. I want to have an urgency score. I want to have a parental score." Then, with GTM it's really easy to say, I want a trigger on these pages to throw up a little bit code that throws an event that says this is an urgency event, we put a ... We capture into a cookie or we score up into GA and we say, okay this shows urgency of this it's shows for my checking or whatever else. We add it to a custom dimension.
Now, that's in the data but it's also on that user so I can then look at that user and say, "You know what, I want to target people that are parents, that are probably parents that are also displaying urgency about their taxes." Or, something like that.
Then, I can look at that audience and if that audience shows that it is highly valuable, or it is highly converting, now, well, I just created an audience. I just have a Regx where I'm saying, "Okay, I'm going to go into my audience builder in GA and I'm going to make an audience for people that show one or two urgency but once they get to three urgency, that's where they start getting valuable. Maybe an audience of people with an urgency over three who are probably parents. There you go. Congratulations. You're winning.
JD Prater: Yeah, that's where I was like, "And, my mind's blown." I thought that was really good. I really wish that we could spend two hours. My ask of you is to write the most detailed blog post of all time.
Sayf Sharif: Yeah. I've been working on. I have it. I swear to God. I have been promising this for a while. I did promise it in November. I'm running out of time but, I really intend to have the whole big long deep blog post out there for you.
Negative and Positive User Scoring in Google Analytics
JD Prater: Good. Good man. Yeah. That was something that was really good. We talked about custom dimensions. We talked about content grouping and like the scoring, which I thought was really cool. The positive. Talk to me a little bit too, because I thought one thing was really cool, you talked about negative scoring and how you incorporated that. For people that weren't there at the conference, talk to us a little bit about negative scoring.
Sayf Sharif: Sure. Well, there's actions that indicate you're likely to convert in a way and there's actions that indicate you might not be. The example I gave at the talk was again for Trex and it was about do-it-yourself-ers. A large component of their customer base are do-it-yourself-ers who want to build their own decks. Other people don't want to build their own decks. They want to hire a contractor, which is my audience. I get those people. I don't understand the do-it-yourself-ers. That just makes me scared.
Anyway, but, you can do things that indicate you're probably going to be a do-it-yourself-er. You might interact with content that's specifically about, "Hey, you want to build your own deck." If you have a blog post that says, "Hey, how do I build my own deck, myself?" Well, they're probably, there's a chance that they might just be reading it, but let's give them a point for that. That's the positive scoring.
But, if they go start searching for contractors in their area, they might not be a do-it-yourself-er if they're searching for contractors. We're going to give them a negative score.
We can adjust that score and we can save it and store it over time and we can say, "Okay, what's their score now? How's that score change over time?" It can go up and down. We're not just saying, "Here's actions that indicate you want to build your own deck or you want to open a checking account, but it could be ... Because, like for a checking account example, if you look at one piece of content about a checking account and then you look at 30 pieces of content about doing something with a CD or something like that. Well, you're probably not a great checking account target.
If you have a remarketing pixel on something about do-it-yourself content or about checking account, then, everything else you do, it's like you would exclude that from an audience. You say, "Okay, if someone completes a goal for finding a contractor, we're going to remove him from the do-it-yourself audience. Until they do something like that, if they're just doing behavioral actions before a conversion, leading up to something, you still don't know you want to give them not just positive scoring but negative scoring and multidimensional scoring where you can say algorithmically, "I need someone that has at least twice as much points in this category than that one or, if someone has a negative score here, I want to just eliminate them from the audience entirely.
Remarketing Lists Created From Google Analytics Audiences
JD Prater: Yeah, I mean, that's where I was like, "Man, I've got to get on this." That is such a powerful remarketing list. We're talking display, we're also talking like, just like the RLSA say, so, all of you search marketers out there listening, just think about how fine tuned these audiences are and how you can really tailor a message or tailor some ad creative for this type of person.
It's no longer just like a catch all bucket that you probably have running. It really is some really advanced GA audiences that I doubt were just brilliant because I had never even really thought about the scoring aspect. In your presentation you gave access to your script. We can include that in the show notes. You guys can understand the script that you can throw into GTM in order to kind of get this information on to your website. They'll start tracking it.
Sayf Sharif: Yeah, yeah. It's a really ... It's a strip down basic example of the script where it's just adding a point to something, and then being able to track it into a cookie using GTM and GA. It's stuff where, with that script, if you know Google Analytics and Google Tag Manager, you could have that up and running on your site within a half hour.
JD Prater: Well, that's a perfect time for us to leave a good cliffhanger. If you're listening go check it out. We'll make sure to include it in the show notes. Let's transition into some lightning round here where I'm going to ask you a couple of questions and you've got 60 seconds to answer each one.
You're ready to go?
Sayf Sharif: Okay.
JD Prater: All right we'll do a softball question first. It's the question I always like to start off with and it's ... Let's just say tomorrow afternoon, you've got zero meeting. No internal, no external. You've got three hours to yourself and you're like, "Man, I need to get caught up. I haven't caught up on industry news. I need to see what's happening." What are some sources, some people, podcasts, what content do you consume in order to keep up.
Sayf Sharif: My normal content is podcasts, these days, because I can do it on my commute. I really like half hour podcasts because I have about a 25 minute drive so it's really consumable on my cell.
JD Prater: Let's go ahead and wrap this up,. No, I'm kidding.
Sayf Sharif: Yeah, yeah. I also like ... Like I mentioned before, the digital analytics power hour is a really good one because, I really like that one. I find that one keeps my mind active. I would say that's the ... Anything else I can find out. There's not a lot but, that's the number one I would say, I would always go to.
JD Prater: Perfect. No, I like it. I'm not familiar with that one, so I will make sure to add it to my keyword podcast. I'm with you as well, I've got about a 30 minute commute. That 30 minute podcast, it's just perfect for me. Fantastic.
All right, question number two that I always like to lead into. It's really around, for you, you talked about Google Analytics. What I want to ask around the free versus the paid, is the paid version, like the Analytics 360, is it really worth the price?
Sayf Sharif: It can be. I mean, I wouldn't recommend it to everybody. Full disclosure. Seer just became a GA-360 reseller so we can't ... I will admit that I do have skin in the game.
JD Prater: Got you.
Sayf Sharif: Yeah, it's not for everybody. If you're doing for instance a lot of double click stuff, those integrations seeing the view threw in the multichannel funnels, having the DBM, DCM integration, can be incredibly valuable. If you have so much data and you're doing so many different things that you're hitting un-sampled data and its' just unusable ...
I've seen people that have so much un-sampled data and 'required cardinality' that they were having problems looking at anything over that data. Even in a day, having that un-sampled data, having the big query integration, it can be worth it.
If you have less than a half million, if you have less than, a quarter million sessions in a month and you're not really interested in data driven attribution, you're not using double click, you're not doing anything that it's going to provide you, then, no. You definitely shouldn't buy it. I mean, I've had plenty of people before where it's like ...
Like I said before, I'm not about tools. I'll use tools. GA-360 is just a tool. It can be valuable but for a lot of people it's like, "Look, you have a lot of better ways you can be spending your money right now, than 360. Let's fix a lot of other stuff up and if you get to that point, then we'll help you get 360." It just depends. For the right people it's extremely valuable.
JD Prater: Got you. All right. Let's keep it into GA. Underutilized feature that you're like, "I can't believe people aren't doing this."
Sayf Sharif: I mentioned before custom dimensions. It always, it's one of the things we look at when we look at when we do a critical audit when we start projects. How many people just don't use any custom dimensions what so ever. They're not adding anything to their data. I think that's number one.
I think even just GA Events, I think are extremely underutilized. People use them here and there but a lot of times also will go in and it's just they don't really have many or they're just, they're not structured in a way that's being useful or a way that's implemented to add information. I think a lot of stuff we do with measurement strategy is to really understand how can we really leverage thing like, the additional events, the additional dimensions of data to really squeeze it for as much information as possible. People don't always do that.
JD Prater: Nice. One of the things that you mentioned to me when we were in London was that you're building out this team. How many people are you up to now, how many analysts?
Sayf Sharif: 15, now.
JD Prater: Yeah. 15 analysts. That's crazy. That's a lot. What are some characteristics that you look for in a future analyst. If you were to hire number 16, tomorrow and you're doing some interviews today, what are you looking for.
Sayf Sharif: We compartmentalize it to several different categories. There is no analytics unicorn. No one can really do all the stuff. They don't exist, except for me, hair flip. No, I'm not ...
There is like the strategy analysts, who understands business, understands behavior, understands what do we need to study, where can we actually get value from. There's people that are really good at visualization and displaying that and communicating value and what needs to be acted on and things like that, telling a story. A lot of time that's the same person but sometimes ... It's a Venn Diagram. They overlap a little bit but not all the time.
Other people are more developers. They're really good at implementing. They aspire to be CMO of or they live on the GTM blogs and want to develop stuff. Other people are data engineers. They want to do, "How can I really create and crunch and play with the data and get it into a Hadoop cluster and do all this aggregation, automation, things like that." Other people are data scientist, who want to use R and Python. These are all different people.
When I look for the next person, the first thing I have to ask is, what do I need right now? Where are my needs? Sometimes that's a data scientist and sometimes that's an analyst and sometimes that's someone who's really good at visualization. Once I know that, then I can look for that person. But, in general, I would say, the biggest circle is really that first part, which is understanding human behavior and people. If you can't do that, you're going to struggle with everything else that we're trying to do, except maybe if you're a developer.
In general even there, you've got to have a desire to understand people and also have a desire to like, you enjoy solving problems and creating solutions for them.
JD Prater: Nice. Are you guys hiring anyone right now too, you go ahead and plug it.
Sayf Sharif: We're always hiring. We are definitely looking for more people at the support level right now, but it comes and goes. Sometimes we're looking for people with more experience. The right person comes along, anytime, I'm going to hire him, but, yeah, we're expecting to hire ... I don't remember what the prediction has been. Over the next year I'm going to need to hire a lot of people. Keep your eyes on our website and apply and if you're good, we'll talk to you.
JD Prater: Yeah. Make sure you start following Sayf here on Twitter as well.
Sayf Sharif: Yes.
JD Prater: Got to have the followers man.
Sayf Sharif: Gotta have the followers.
JD Prater: All right. Last question and then I'm going to let you go. All right, let's just say, Google Analytics shutters. The tool is gone tomorrow. You're like, "Man, I need a job now." All right. Let's just pretend that was the only thing that you did and now you don't have a job. What's a fall back job that you think about?
Sayf Sharif: Like a legitimate one or like a ...
JD Prater: Like a legitimate fall back job, yeah.
Sayf Sharif: Man. I feel like I'm pretty secure.
JD Prater: Yeah. I agree.
Sayf Sharif: Fall back jobs. I feel like, definitely, if GA were shutter, I'll be like, "Okay, Adobe. Yay."
JD Prater: Yeah.
Sayf Sharif: "Coremetrics." maybe look for another job. No, I'm kidding.
JD Prater: Got you.
Sayf Sharif: I haven't really thought about that. I mean, I think, I probably honestly would go into something completely different and creative. I make board games and card games and things like that in my free time myself, that I never really share with anyone so I'd probably just say, "wow, let me kickstart some of the stuff and see what happens." That probably would be where my instinct would be to go, would be to start creating some game Kickstarters and seeing if I could just start my own company that way.
JD Prater: Well, that's cool. Yeah, that's probably the most unique thing anyone's ever said. That's really cool. I like it.
Do you want to give us any samples of this kind of game, you just let us know we'll put it up there in the show notes and you can have people up-vote or down-vote it.
Sayf Sharif: All right. I haven't published anything yet, so we'll have to do it ... Everyone will have to sign an NDA, and then we'll do it.
JD Prater: There you go, there you go. I like it. Awesome.
Well, thanks again.
Sayf Sharif: Yeah.
JD Prater: Enjoy your Thanksgiving.
Sayf Sharif: You too.
JD Prater: I really look forward to having you back on the show sometime in 2018.
Sayf Sharif: Yeah, anytime.
JD Prater: All right. Have a good one.
Sayf Sharif: All right. You too. Thanks.
JD Prater: Bye.
Sayf Sharif: Bye.