The PPC Show

[PPC Show] PPC Automation and The Great Attribution Myth

This week on The PPC Show, Julie Friedman Bacchini, President at Neptune Moon, shares her dinosaur perspective on the future of PPC automation in paid search and the great attribution myth. You can connect with her on Twitter.

Julie's been in marketing for over 18 years so tune in to hear:

  1. How advertisers can take baby steps into paid search automation as well as the best used cases for applying automation.
  2. Then she'll explain why she thinks attribution is kind of a joke.


More About Julie

Fun Fact: If there were no other constraints then Julie would want to do something working with animals.

Julie Friedman Bacchini is a professional speaker who has presented on paid search and general marketing and web topics at local, regional and national events – including SMX East and PPC Hero Conference. Her easy going style makes her a popular presenter, along with her passion for her topics and accessibility to event participants.

Able to present at a highly technical or expert level or to more general business audiences, Julie really enjoys sharing knowledge and breaking down internet marketing topics so that they are easy to understand.

Show Notes & Transcript

Julie Bacchini:                      ... and those things are unknowable in the sense of tracking of attribution. I don't know how they ever are going to be knowable unless we all become microchipped and decide that privacy really is just no longer a thing.

JD Prater:                                Welcome to the PPC Show, where we interview the best and brightest in paid search and paid social advertising. This is episode 59, and I'm your host, JD Prater. This week, I'm joined by Julie Friedman Bacchini, self-proclaimed digital dinosaur, to discuss PPC automation and the great attribution myth. Julie's been in marketing for over 18 years, and on this episode, she's gonna be sharing with us how advertisers can take baby steps to paid search automation as well as the best used cases for automation. Then, finally, she'll explain why attribution is kind of a joke. So stay tuned and let's get to the show. Hey Julie, welcome to the show.

Julie Bacchini:                      Thanks. I'm really excited about this.

JD Prater:                                As am I. Thank you so much for coming on. I think we're gonna have a really good one. This is a really good one to actually dive in to talk about. So we're gonna be talking about some PPC automation and the great attribution myth, but before we get started, why don't you give us a little background of who Julie Friedman Bacchini is?

Julie Bacchini:                      Sure. I am a bit of a digital dinosaur, as I like to say. I have been working in marketing in general, but specifically in digital, for about 18 years now, in the digital space. I have been here since the beginning, and I have seen everything from websites just beginning to develop to platforms beginning to develop, and I've worked in pretty much every aspect of the online space, with a particular specialty over the last 8 to 10 years in paid search.

JD Prater:                                Wow, that's impressive. How did you ...

Julie Bacchini:                      Well, thanks.

JD Prater:                                ... get started with Neptune Moon, then?

Julie Bacchini:                      I actually had a full-time marketing position with a health system, and I'd been doing some work, freelance work, on the side, thinking about going out on my own. Then, my department got eliminated, and I thought, well, I guess today's the day, and the rest is history.

JD Prater:                                Wow, bold. You know, I've actually run into a few people with a very similar type of story. You know what? Thank you to you type of people who are so bold to go out there on a limb, start their own agency, and fulfill the work, do the work. I am always impressed with that kind of story. So kudos to you.

Julie Bacchini:                      That's very nice. Thank you.

JD Prater:                                Well, let's have some fun here. Let's jump into some PPC automation. Tons of announcements coming out, especially in the last year, and since you have such a really great view ... You've talked about 18 years. Talk to me about what you kind of see on the horizons. I think if we would have talked a year ago, we would have seen paid search specifically. People that are like, "Eh, I love the idea of scripts, but I'm scared of them." There really seems to be a shifting of tides, and I just want to get your thoughts around PPC automation, specifically within paid search.

Julie Bacchini:                      I do think that things are shifting. It does seem that the pace of development ... A lot of the automation-related technologies really seem to have taken a pretty big leap forward in the past 12 to 18 months. I think that that's definitely true, because you're right. In the past, the idea of trying to do a script on your own or even just doing anything more than creating just a basic rule within an account of notify me if this happens, or if this happens, then go ahead and do that. That was sort of the beginning and the end, unless you were using a pretty sophisticated third-party tool for options as far as automation went. Now, I mean, I feel like the term machine learning, we're a little oversaturated with that at the moment, but I think that's just starting, and it encompasses such a wide range of technology.

It's a little unclear exactly which pieces of that are going to take off, but I think, as a whole, it definitely has a lot of potential to take some of the things that have to be done on a regular basis when you're managing paid search and just be able to create logic and create ways of doing things that free up managers to kind of dig a little bit deeper into pieces that machines can't really understand, like where the nuance is, where you need to be making connections, where you want to take numbers and then figure out what does that mean in the broader sense. I don't think machines are great at that. I think people are. If a lot of that can be brought to the table via automation so that you can more easily do some of that type of work, I think that that's really exciting.

JD Prater:                                Yeah, I agree with you on that. Let's kind of dive into it. What are some of the tasks that you think should be automated or that you are automating, and then, what kind of impact have you seen automation have, even within your own campaign management?

Julie Bacchini:                      I've been dabbling in scripts. I mean, I've done some low-level programming for websites and other things like that over the years, so I finally got over my fear of scripts in the last couple of years. I have been using scripts to sort of automate. I really like automation for monitoring of things, especially when you get to scale. I think it's sort of one thing for you to be manually monitoring performance or bids, even if there's a problem that comes up, to make sure that all of your ads are pointing to URLs that aren't, in fact, broken. I think there are a lot of those things that when you're working on a small scale, maybe you don't necessarily need to automate, but when you get to a certain critical mass, it does kind of become a little bit of a waste of time, I think, for somebody who has a high level of expertise to be spending time double-checking that kind of stuff.

I think being able to utilize scripts that just automatically run and they will send you a report or they'll make a notation, they'll add a label, they'll make a notation right in the account so that you can then more efficiently, I think, just take a look at everything and know, yup, everything's okay or these three things need my attention. I think that that really makes a difference in the day-to-day management and what you're doing and where you're spending your time.

JD Prater:                                Yeah, I definitely agree on that one. It's one of the things that me coming into paid search, I was like, oh, yeah, scripts. Cool. You've got like Daniel Gilbert, you've got like Frederick Vallaeys writing these amazing scripts and giving them away for free on Search Engine Land. I was like, I'm just going to put those in my accounts. Then, I would talk to ... I'll call out Matt Umbro here. Umbro's like, "Nah, I don't really use them." I'm like, "Wait. So am I the weird one? Why aren't people doing this more?" I was just like, "This is great. It's going to help me do my job, and it's going to make me faster, and it's going to make me better." Then, that's when I kind of found this rift within automation and the control freaks of the world. Moving forward within that idea of control freaks, we finally have got to this point where it seems like paid search ... I'm calling paid search people control freaks, which is ...

Julie Bacchini:                      Fair.

JD Prater:                                ... in a position to paid social people that are just like, "Yeah, Facebook, do whatever you want."

Julie Bacchini:                      Yeah, for sure.

JD Prater:                                Where do you think the next year's going to take us? Let's say you and I ... We talked about a year ago, we probably wouldn't be even having this conversation. At the speed at which things are moving ... and if you've seen this new interface within AdWords, we can already tell where it looks like Google wants to go with things, which seems like more control. Where do you think, in a year from now, if we have you back on the show, where are we headed?

Julie Bacchini:                      That's a tough question. Part of it has to do with what becomes universally accepted and adopted. To go back to what you were saying about this rift between people who like automation in general and who don't, I think as things become more accessible and they become more baked into ... I mean, the AdWords interface, the new one, is not completely rolled out, so it'll be interesting to see if there are more automation features that are baked in. I think that when things are simpler to implement ... A script, it really depends on your background, and it depends on a lot of different factors. I think if using a script is something that when you hear the idea, you're like, "Oh, that's cool. I really want to try that," or it makes you hyperventilate ... You know what I mean?

JD Prater:                                Yeah.

Julie Bacchini:                      People are kind of anywhere on that spectrum. I think they're all sort of justified. It's a little bit a product of your own experience. I think that more things are going to become more accessible so that you don't have to feel like you're somebody who knows how to code in JavaScript to be able to take a script that some brilliant person has created and then actually put it in your account and have it be working and know what to do with it. I feel like that's going to be so much more accessible a year from now even than it is now.

JD Prater:                                Got you. This is my last follow-up question here on automation is there's always this kind of undercurrent around people talking that automation is going to steal our jobs. Within the PPC world, do you think that's really going to happen? I guess let me say is that going to happen, and what do you think that's actually going to look like? You touched on it kind of briefly, but I'd love to hear your thoughts on fleshing that out a little bit more.

Julie Bacchini:                      The dinosaur's perspective.

JD Prater:                                There you go.

Julie Bacchini:                      I will say this, I think that, inevitably, in every industry, there really are not very many exceptions, the longer it's around, the more that at least a portion of the industry becomes what I would call commoditized. I think that we've already seen that to a certain degree in paid search. Less so in social, because social has not really been around as long as traditional search advertising. I think that there's a certain amount of commoditization that happens. What I mean by that is there's a certain amount of baseline work that becomes a lot less specialized. In other words, you don't necessarily need the same degree of technical skill that perhaps you needed years prior as something was beginning to develop.

I think, to a certain extent, that that has become true in paid search. I think the limitation, at least in my opinion, as far as how far can automation go as far as replacing somebody like you or replacing somebody like me, I don't think that it can in a lot of instances. For basic needs, for super simple accounts, maybe, but for people who are really interested in having something that can be all that it can be and have maximum reach and maximum return and all of that, I don't know that there's a substitute for the human brain when it comes to really being able to not just aggregate and interpret and analyze a lot of information, but I think the one place that automation really falls down is nuance. I use the example when you look at close variance.

JD Prater:                                That's a good one.

Julie Bacchini:                      To me, that's sort of the first brush a lot of us have had with machine learning. It's got a long way to go to replace a human. It doesn't understand nuance. It doesn't even really understand all the time the same word can have more than one meaning depending on the context. If you think about it from that point of view, I do feel like we are a long way away from automation being a serious threat to people's career longevity.

JD Prater:                                Exactly. I agree with that one, too. I'll even throw in there are responsive ads that they will build for you. I was like, "Oh, you almost got the brand color, but not quite. That's not quite right." There was definitely some nuances. I give them credit for getting really close and doing a really good job, but I think, in the end, one thing I've always thought about ... Have you seen the movie Hidden Figures?

Julie Bacchini:                      No, not yet.

JD Prater:                                It's really good. I don't necessarily want to give it away, but there is a part where IBM's got their big computer coming into NASA, and it's either you can get replaced by the computer or you can learn how to program the computer. They took the necessary steps to make sure that they had jobs. I think that's also given rise, maybe, to this rise of the technical marketer. Have you seen this? Do you have any thoughts about that?

Julie Bacchini:                      Yes. I think it's smart. I mean, you have to sort of figure out where things are going. I say this as someone who's been doing it for a long time. You have to figure out what people want from their providers. There's a combination of, on an emotional level, understanding what it is that people are wanting and what they're needing from their providers and positioning yourself to be able to provide that. I think the other piece is, and one of the greatest values that professional paid search managers have, is staying on top of things that are happening. We are more apt to know about, learn about, and not be afraid to implement something that is sort of technical in nature, so whether it's scripts or it's using schema or it's diving into the whole world of AMP. Is that a good thing? Is it a bad thing? It's kind of both.

It's having the knowledge to be able to see those things and then weed out what's worth doing and what's not worth pursuing. Then, learning about the things that you think are worth pursuing and making them part of your process and part of your value proposition.

JD Prater:                                Got you. Man, so good. If you are listening and you're new to the PPC world, take that. That is some really good advice. All right, so let's shift gears here away from automation and talk about the great attribution myth. You just wrote this fantastic article around attribution. You went on to say, "Attribution is kind of a joke." What did you mean by that?

Julie Bacchini:                      I did say that. I am not one to shy away from bold statements. A lot of my blog posts, to be honest, happen when something just sort of strikes me, and then I have to get it all out, and this post was no exception. I think this is sort of the dirty, little secret that doesn't really get talked about. Maybe people don't want to talk about it, I don't know, but I feel like it's a really important topic. The way that we generally talk about attribution, I think we treat it like it's a mature, fully developed, really reliable system, being able to accurately know exactly what a customer's journey was from that initial spark in their head of, hm, I might need something, to when they do all of their looking, researching, going to physical stores, all the things they might do. Then, they eventually reach that point where they purchase/don't purchase. I think we treat that as a process that we have a complete view and access to and all of it is trackable and attributable, and I just really don't think with today's technology that it is.

JD Prater:                                100% agree. I couldn't agree more. This is me. This is me coming from agency side and now being internal. I would love to say, "Yeah, I'm looking at the customer journey all the time and I fully understand it," but it's crazy. Let's face it. One, customer journeys are crazy. Two, the technology is not there. Let's kind of break down some of that technology and the attribution. Why do you think it's not there yet? What are some of the things that I know that you pointed out in this article that are lacking?

Julie Bacchini:                      I think, first and foremost, there are certain things that human beings do in the course of a decision-making process that I call them unknowable. From a real, practical standpoint, if you are trying to track different things, there are definitely processes and things that people do that just are unknowable to a marketer. You have no idea ... If I'm seeking some kind of product or service, you have no idea, as the marketer, for instance, if I have a conversation with my neighbor about it, if I am seeking recommendations on Facebook, if I'm doing different things. There's just no way for you to know, but one of those things could turn out to be the tipping point or the most influencing factor for me as to whether or not I decide to become a customer. Those things are unknowable in the sense of tracking of attribution, and I don't know how they ever are going to be knowable, unless we all become microchipped and decide that privacy really is just no longer a thing.

JD Prater:                                Then, should we just give up on attribution?

Julie Bacchini:                      No, I don't think that we should give up on attribution, but I think that we should contextualize it better. I feel like the way that we've been talking about it really makes it seem like it is more powerful and complete than it actually is. I really think that if we start to talk about it for what it is, then we can be using it more appropriately and we can be using those pieces of data to kind of create a feedback loop for the different parts of the journey that we are able to see. Then, we can take some of that information and we can extrapolate it out when we're looking at other types of data that you might not be able to directly correlate, but you can make a reasonable assumption that there is at least a correlation, possibly a causation, if you're seeing a bump in foot traffic, if you're seeing a bump in inquiries, if you're seeing a bump in different types of activity. Looking at things more on like a timeline factor based on things that you were doing.

I think it's kind of changing the way that we think about it and changing the weight that we give to today's current attribution I think puts it in a better perspective and a better context, and it makes it more useful.

JD Prater:                                Yeah, I agree. How do you go about handling ... Typically, what is trackable and what is attributable gets the most amount of budget. At least, that's what it feels like. We know that we're still kind of stuck in this last click mindset, but how do you go about giving more budget to maybe more top-of-funnel efforts, maybe like a Facebook, and how do you talk about that with clients?

Julie Bacchini:                      I think one of the worst things that has happened as we have had access to more data is that, again, we have given that data more power than it actually deserves at this point, which I know is also a pretty bold statement. What I mean by that is we are treating data that we have access to as sort of the be all, end all, when it's really a portion of the picture. I think that we need to return a little bit to how things were before digital. You remember, at least I remember because I was doing this before there was digital, that we didn't have any of this kind of data. Everything that we did, we had very, very minimal data as far as what were people actually doing, how were they specifically interacting with the types of outreach or the types of advertising or things that you were doing to try to move the needle to get more inquiries, get more purchases, get more foot traffic, that type of thing.

I think that we could probably rein it in a little bit as far as treating the data that we have as being so absolute. I think the whole last click piece is a really great example of that, because the absolute data that we have really kind of is last click in a lot of instances. We know for sure that the time that you visited the website and you actually decided that you were going to purchase the product, you came in via an organic search or you came in because we sent you an email with a coupon offer, and that moved you from thinking about it to doing it. We have a pretty clear understanding of what that last thing was that we did that got you from maybe to yes, but I think that we need to treat that as we know what the last step was, so we know that this can be effective in taking from thinking about it to yes. Start thinking of it in that way instead of giving it more ... We don't want to give it more power.

I mean, I think we've probably all seen that cartoon that shows, oh, all this came from mobile. Oh my gosh, we need to put all of our money in mobile, because everything is showing up as the last click is the mobile or the last click is coming from social. That's nuts, because we know it's not. You can think about your own behavior. I mean, really, I've said this for years. One of the most important things you can do, I think, sometimes, is just think about your own behavior. What does your customer journey look like when you're looking for different types of local places to go, products or services, or things like that? Even as a digital marketer, what does your journey look like? For me, it depends on what I'm looking for, how urgent. It could look eight different ways. I think that that's typical for most people.

We have to start thinking about how do we move the needle, and how do we do it in a way that ... I mean, the bottom line is clients want to move the needle. I think we're doing them a bit of a disservice by acting like we can definitively put everything precisely, exactly, and completely in each bucket. We need to be working with them for a full-spectrum approach, because people are looking at stuff in 100 different ways. We need to be trying to do our best to narrow down where we think we're most likely to reach them, and then we need to spend time seeing if we're pushing stuff into column A, are we seeing any difference on the bottom-line metric that we're trying to measure? Then, kind of working from there.

JD Prater:                                I think it's really good advice. It's really good advice. Even here in San Francisco, I know there's a lot of startups really trying to crack that attribution code. There's a couple of companies that are doing it fairly well. I mean, still, they're not cross-device. I think that's the huge limitation that I'm seeing. I was just thinking about this yesterday, kind of prepping for our conversation. I have my work computer. I have my phone. I go home, and then I have my personal computer that's my computer. My wife has her work computer and her personal computer and her phone. We have six different devices that, potentially, a conversion could ever happen on. To think that how Google Analytics tracks, really it's by device is how they're fingerprinting you, but we would come in as six different people. We're really just the same. We're two different people, but in the same household whenever we're doing some of this stuff. That was crazy for me to even wrap my head around is making that decision. I think you're right, with those unknowable pieces, and you're never going to know all those. Good.

Julie Bacchini:                      But it's okay.

JD Prater:                                It is.

Julie Bacchini:                      I think what I really want to get across is there's been such a gigantic push for big data and big data. If you're not using it, you're not doing it right. That has been such a tremendous push over the last, what, like three to four years. I'm tired of hearing about big data. You need the important data. You don't need gigantic data. Just taking a little bit of a deep breath and saying, "Okay, look, yeah, data's awesome, but data can lead you down the wrong path, too." If you're not looking at the right information or you're misapplying it or you're misattributing it, then it's not that helpful. I think that there's always going to be ... To your point about the multi-devices, I mean, we're the same way in that we have a tremendous number of devices here in this house.

There's always going to be, at least I think for the foreseeable future, this tension between people wanting to have the best advantage they can have as far as brands knowing them and being able to personalize things, but there is a hard limit on that. That tension between marketers want data, customers want personalized offers, but nobody wants to cross that plateau where, suddenly, the brand knows exactly who I am, they know exactly which device ... Like, they know everything about what I'm doing. That's a bridge that pretty far from where we are right now, and I don't know that it's a bridge that ever gets crossed or should get crossed, really.

JD Prater:                                We can just leave it right there, then. We can transition into the rapid fire round, where I'm going to ask you-

Julie Bacchini:                      The lightning round.

JD Prater:                                The lightning round, that's right. I'm going to ask you a couple of questions here, give you about 60 seconds or so to answer each one. You ready to go?

Julie Bacchini:                      Okay. I'm ready.

JD Prater:                                All right. Start you off with an easy one here. You got an afternoon free, no meetings, no prep work, no optimizations, no management, and you're like, "I need to do some reading and get caught up on the industry." What are some of your favorite blogs and some of the places that you'd turn to to keep up with the industry and also within PPC news?

Julie Bacchini:                      The first thing I would do would be I would go to Twitter. The PPC chat hashtag, I have a column open for that all the time, because that queues me into things that are happening right now. I find that that's really helpful. Some of the sites that I read regularly, I read blogs regularly from a few of the participants in PPC chat, but more generally, Search Engine Land and Marketing Land, I think, have really good coverage of what's going on. Then, I get a couple of other newsletters. I like Street Fight. I don't know if other people are familiar with Street Fight Magazine. They have just a ton of information about all kinds of things marketing related. I get a couple of emails called Smart Briefs. I get one on social, and I get one on interactive advertising. I find a lot of great information through those sources.

JD Prater:                                Cool. No, I never heard of Street Fight. I'll definitely take a look at that and leave a link in the show notes. Thanks for that tip. Let's say Neptune Moon, you guys are going to be hiring a new PPC account manager in the next month or two. What are some of the skills that you look for in this position?

Julie Bacchini:                      That's a great question. I would say certainly a strong base in the fundamentals. It's really hard to start somebody off who doesn't have a grasp of completing basic tasks inside of the different platforms. Beyond that, the piece that I look for the most is somebody who asks questions, someone who has ideas about things. I really encourage that. I think that no matter what level you're on, taking ownership of what you're working on and really trying to immerse yourself in it and really think about it from the client's point of view or applying experience you had from other things to say, "Hey, we did this over here. I'm wondering if this might work there." I look for that type of personality and that type of thinking, because I think that has the greatest potential for growth and development.

JD Prater:                                Got you. Cool. All right, good answer. Good answer. I like it. I got two more for you here.

Julie Bacchini:                      Okay.

JD Prater:                                You've been in the industry for a while, 18 years, and then really focusing in within paid search. What's something that you wish you knew then that you know now?

Julie Bacchini:                      Wow.

JD Prater:                                That's a tough one.

Julie Bacchini:                      I know, that is a hard one. I feel like I feel like that every three months, that there's something I wish I knew. I don't know. I think just knowing sooner that search was going to actually become so much of a thing and knowing how important mobile was going to be. I feel like I could be further along in helping clients to create better mobile experiences. I could have started pushing for that a lot sooner. Some are there, some aren't, but that is not slowing down any time soon. I feel like that is going to become more and more and more and more critical as we move forward. Knowing when that was actually going to hit that tipping point would have been helpful.

JD Prater:                                Yeah, I wish I could have definitely foreseen that coming up the way that it did, 100%. That one hit quick. That one hit quick. Actually, I really thank Google for helping push that one along with, hey, we're doing these mobile optimized sites, and you need to make sure your site is ready. That at least pushed a lot of my clients to really think mobile first and thinking about that. Good one. All right, last question for you that I always like to ask all of our guests is let's say tomorrow, Neptune Moon ceases to exist. It's not going to happen. We know you guys are going to be around for a while. What's a fallback job? What's something that you would start to take up, let's say, by next week?

Julie Bacchini:                      If I couldn't do this anymore?

JD Prater:                                If you couldn't do it anymore. What do you think?

Julie Bacchini:                      I don't know. If I could do anything, if there were no other constraints on me, I would want to do something working with animals.

JD Prater:                                I like that. I like that.

Julie Bacchini:                      I probably wouldn't make any money doing that, but it would feel good and it would be a great thing to do in the world.

JD Prater:                                Yeah, definitely. Definitely. Well, you have survived. Congratulations.

Julie Bacchini:                      Thank you. It was fun.

JD Prater:                                Yeah. Thanks for coming on the PPC Show, really talking through automation and attribution. These are really hot topics right now within the industry and sharing your bold statements and your bold thoughts. I really appreciate your time and for you coming on and sharing those with us.

Julie Bacchini:                      Well, thank you so much. I really appreciate you having me on. It's a lot of fun. That really flew by. It was just like sitting and having a cup of coffee.

AdStage Team