SparkToro is Audience Research in Seconds Ft. Rand Fishkin

Rand Fishkin:

Hard-to measure-channels tend to outperform easy-to-measure ones because competition is lighter, because people tend to pay attention to those channels, and you can’t just pay to have your message sent through them. The things that have moved the needle the most for SparkToro over time are very popular podcasts and YouTube channels, email newsletters, events featuring our work. I can’t pay for it, I can barely track it, I can only really see it in lift over time. What I’ll see is, hey, this thing came out, I see what are called vanity metrics by marketers now. Vanity metrics like views and impressions. And then I will see the impact of it slowly over time, and I’ll get some emails from customers, and they’ll be like, I heard about you here, I checked you out here, we have this need. Oftentimes it’s months or even years later, someone will be like, ah, what was that tool where the guy talked about being able to see behaviors of an online audience? I need that now.

Adam Kerpelman:

Hey, everybody it’s the data-driven marketer, I’m Adam.

Mark Richardson:

I’m Mark.

Rand Fishkin:

And I’m Rand.

Adam Kerpelman:

Welcome back for another hang in the data basement. Thanks for joining us everybody. And special thanks to our guest this week, Rand Fishkin, who is the CEO of SparkToro and a repeat guest. So, it’s part two, which means we’ll link to that previous episode if you want to go get the long version of the bio and all that stuff. But otherwise, Rand, do you want to give us your background real quick for people that probably won’t go listen to [inaudible 00:01:42].

Rand Fishkin:

Sure.

Adam Kerpelman:

[inaudible 00:01:43] about how web traffic works?

Rand Fishkin:

Yeah, yeah. No problem, Adam. For folks who don’t know, I started a company called Moz many years ago in 2003, that was originally a SEO consulting business and then became a software company, and for a little while was a market leader in SEO software. I left that company in 2018 and started SparkToro the next day. SparkToro makes audience research software, I also wrote a book called Lost and Founder, Painfully Honest Field Guide to the Startup World, which a lot of folks in entrepreneurial and in tech and marketing worlds have read and said nice things about. Pretty much all of them except Scott Adams, who I throw under the bus in the book. But he deserved it. And on reflection, that decision looks better every day. I am also, oddly enough, the creative director for a video game, an upcoming video game title to be announced probably early next year. Working with the studio down in Los Angeles, and that’s been a really fun and interesting project. I’m moderately internet famous for being Geraldine DeRuiter’s husband. She is a famous food, travel, humorist writer so I get a lot of people recognizing me. Even at the grocery store, people will be like, “Oh yeah, you were in that article about that restaurant in Italy.” So there you go.

Adam Kerpelman:

I didn’t even know that. That’s an exhibition of the purity of our [inaudible 00:03:22] over here. I’ll say the video game part doesn’t surprise me, I don’t think it would surprise anyone else that’s familiar with the creative thread that runs through Moz and now I can see it in SparkToro, the same sort of… like a mascot, that wasn’t necessarily… believe it or not, that’s still not an intuitive thing to make for your brand in a lot of places, particularly B2B.

Rand Fishkin:

I have great news, Adam, fantastic news for you. My co-founder, Casey’s eldest daughter, decided that SparkToro needed a mascot of its own. I think she’s 10 or 11, and she both drew up and then made in clay this little dragon in our logo colors. Adorable little bulbus dragon. We’re now having an illustrator work on turning that into a digital asset that we can put on T-shirts and mugs and our website. So there you go, Sparky, the dragon is on its way.

Adam Kerpelman:

Sparky, the dragon.

Mark Richardson:

Were you ever a Homestar Runner fan? Do you remember Trogdor?

Rand Fishkin:

Oh, yeah.

Mark Richardson:

Okay.

Adam Kerpelman:

[Inaudible 00:04:36].

Mark Richardson:

[inaudible 00:04:37] The countryside.

Adam Kerpelman:

So yeah, before we started recording, we talked about a whole bunch of different topics. The easiest segue for my money is through the mascot lens in talking about things like Reddit. And we talked about some SparkToro features and things like that that we’ll kind of get to later from a practical lens.

But more broadly, I’m curious to your take on the evolution of social media and all that kind of… When you mentioned Reddit, I had a reaction, as a member of a lot of these communities, that was very like, Oh okay, that’s a place to hang out more than some of the other, I’m not on Pinterest, although I know it’s a signal, its a-

Mark Richardson:

I’m not on Facebook.

Adam Kerpelman:

There’s activity there that I know as a marketer, but Reddit I hang out on and the conversations are completely different. And so that sort of got me to the broader exercise of if you’re…. Or I guess maybe we should start from what SparkToro does and then that’ll be the context for why I’m about to ask you about what it’s like to scoop data in from all of the social networks.

Rand Fishkin:

Yeah, yeah.

Mark Richardson:

So tee up this platform.

Rand Fishkin:

So SparkToro, I mentioned it does audience research software, but I think that can mean a lot of things to a lot of people. So basic story broadly speaking is SparkToro crawls the social web, tries to build profiles of people across as many social profiles as we can get at social networks as we can get, and then anonymize and aggregate that data so that we can tell you lots of information about any group you search for.

So, for example, you might say, I want to learn about people whose bio includes interior design and who are in New York. And so we could say, okay, well we have 612 profiles that we found who are interior designers in New York City and here are attributes about them. Here’s demographics, gender and age and job title and education and geography and politics, and here’s behaviors. They tend to use these hashtags, they tend to use these words and phrases, they tend to follow these sources of influence, they follow these social accounts, they subscribe to these YouTube channels. Soon, very near future, we don’t have this live yet, but we’ll show you which subreddits they subscribe to and content that gets talked about in those subreddits. And so it’s essentially like I want to understand the behaviors and demographics of an online audience, I don’t want to have to crawl hundreds or thousands of profiles myself, can’t someone else do it? And SparkToro is a very inexpensive solution for that.

There are more upmarket folks, so you can do this with something like a brand watch if you have a very large budget. And a lot of enterprises do that, large agencies do that. SparkToro is kind of for the every person.

Mark Richardson:

Who doesn’t want to spend hours upon hours in Excel running queries.

Rand Fishkin:

I’m not even sure how… I guess Excel is what you do after you build a web crawler to go crawl 5,000 LinkedIn profiles and try and connect those up to their Twitter profiles and then try to connect as many up to their YouTube profiles and their Reddit profiles, trying to build this infrastructure. Essentially, what we’re doing is by no means capturing the entire web, but what we’re doing is capturing public activity on public networks in a quantity that is robust enough and representative enough of broader behavior that you can make some really good bets. So we have 40% of interior designers in New York City engage with this particular YouTube channel or that particular social account or this particular website, you can intuit that if you want to reach New York interior designers for your show at the Javits Center and try to market to them, these sources are going to be very influential. That’s essentially what we do.

Mark Richardson:

You’re creating a affinity qualifier. Its the way I like to think of it.

Adam Kerpelman:

The interesting thing that I think lives at this sort of paradigm shift space where we like to play here, is the idea that you can say, okay, it’s squishier than a statistician would like, but within a reasonable margin of error for this stuff that we’re trying to do, we can know with great certainty that this is the case.

Rand Fishkin:

Yeah, I think a statistician could reasonably criticize the fact that, for example, SparkToro is capturing people whose profiles are active on these networks, public profiles only and we are missing activity that happens, even for public profiles, that happens behind the scenes. So we obviously can’t see their WhatsApp messages, you can’t see private Facebook groups, we can only crawl what Google can crawl. But I think any reasonable statistician would say, oh my God, this is a million times better than a survey.

Because if you were to ask people in a survey, if you went and found interior designers and you sent them a survey and you ask them which podcast they listen to and what YouTube channels they subscribe to and which social accounts they follow and what websites they visit, you would get crap information. It’s just, human beings are not good at answering those kinds of questions in statistically rigorous, intelligent ways. And even if you could get good answers, you’d never be able to get more than the top three to five. So how would you find hidden gems and what’s changing over time and asking people what hashtags they post on Instagram? No way. You just can’t get this data. So yes, SparkToro’s methodology is imperfect, is it, in our opinion that the nearly the best thing that you could possibly get from the public social web at this point? Yeah.

Adam Kerpelman:

I never, on this show never give up the opportunity to talk about this stupid early project because it was so formative for me. But I did a thing with the CTO of the company we work for now, who’s the one who recruited me.

Mark Richardson:

And the guy you met last time, Brian.

Adam Kerpelman:

Yeah Brian, he’s the other cohost of podcast. We did an early project we call the Love Hate Index, and it was pulling in trend data off of Google and then sort trying to wait it for an analysis of… it was a sentiment analysis exercise, turns out to be way harder than we imagined at the time. But even the rudimentary stuff we did got really interesting data. The thing that got the best Love Hate score was the Counting Crows, I think. And Chewbacca scored remarkably poorly and so-

Mark Richardson:

Surprising.

Adam Kerpelman:

And then after the fact then you could look at your test and go, okay, but everybody loves Chewy, so what happened there? It was the first time that… Counting Crows I realized is right in that pocket where the people that dislike it, don’t dislike the band so much that they go talk about it on the internet. And the people that do love it, get out there and talk about it. And so it was one of the first times I was sort of like, okay, this is different from just scooping up all this information and trying to make something of it. And the signals are slightly different if you know how to parse them. And in the world of weird super massive statistics, when we start to look at the kind of numbers that Google has for queries, you get this weird phenomenon where you’re like, hey look, it’s a 20% chance that we’re wrong, but based on people searching for cold medicine, we think there’s like a flu outbreak happening in Atlanta right now.

20% is pretty squishy, right? The CDC… Maybe it’s more than that, but it’s like, CDC might not move on that, but it’s real, for sure real, the signal that you’re getting.

Rand Fishkin:

The thing I like about search data is, it’s great for identifying spikes in need or interest proactively, but it really can’t tell you passively where you can reach people. All that Google’s data can tell you is, you should advertise on or optimize pages for Google for these keywords. But people’s behavior is influenced far more by passive sources that they read, watch, consume. Right now the massively overwhelming majority of time spent online is spent on one of approximately eight networks. And it’s not Google. Google is where we go to sort of, oh okay, I heard about this, I saw this thing on Reddit, I’m going to go search for it, or I need this thing right now, I need a landscaper, I’m going to go to Reddit. I need an interior designer, I am looking for hard water processing for my chemical engineering plan, I’m going to go to Google. But that is not where we get ideas and information, where ideas are catalyzed. I think that comes from our consumption. And so being able to know… There’s no way to ask Google, hey Google, all these people who have the flu in Atlanta, what are they binge watching right now?

Mark Richardson:

Yeah.

Adam Kerpelman:

Some people know that, but it’s not… That’s what Netflix is hiding behind their doors.

Mark Richardson:

You’re in the process of launching something that addresses this need. We talked about a beta that SparkToro is rolling out called Tracked Audiences, which I think is really cool. And it basically hones in or puts crosshairs around that type of use case. If you want to know, I’m still learning the tool, but the way I understand it is, basically it applies Google trends to various content affinities. So you’re able to track spikes in interest around a particular topic or an influencer, maybe a particular episode of LeBron James’ Barbershop podcast had had a really big reveal or some gossip that was dropped and marketers might find it beneficial to talk about that or using it. So can you talk a little bit more about that tool? I’m probably butchering it right now.

Rand Fishkin:

Sure. So basically right now when you search SparkToro it’s, here are the behaviors and demographics of an audience at this point in time over the last, essentially 120 days. So we basically are updating it, well we’re updating every day, but the data that you get is across the last three, four months. And one of the things that we saw was that a lot of marketers said, I’m interested in this audience all the time. I’m interested in the people who, for example, follow my business’s social account online across any of these platforms and I want to know what’s changing with our audience.

Are they listening to new podcasts? Are they subscribing to new YouTube channels? Have they started visiting a new website? Is a website that previously was popular with them kind of declining in popularity? Are they talking about new topics? Are they using new hashtags? Over the course of a year how are the demographics changing? Are there more non-binary folks that are coming into our audience? Or maybe there’s more people with less education who are coming into our field, whatever the case, you are deeply interested in not just the one point in time, what’s going on right now, but rather over time trends. And that’s what audience tracking lets you do, lets you say, basically, I want to track chemical engineers in the U.K, or I want to track interior designers in New York, or I want to track people who are obsessed with Adventure time the TV show, all these painting-

Mark Richardson:

Lovely art, yes. Amazing. I would fall into that bucket.

Rand Fishkin:

I’m obsessed with Dungeons and Dragons, whatever it is. SparkToro can essentially take any audience, track it over time and show you what’s changing about it every week.

Adam Kerpelman:

I think the cool thing across the board with the tool, but also this one, well I’ll speak to our use case, it gives us signal on… We are a data company, we sell data. Finding the audience, targeting the audience, not a problem at the top level if you know who you’re looking for. But then knowing which channels to use and a cleaner picture on who you’re looking for gets to be the hard part. And it changes so fast, therefore the overtime component… It’s like new channels blow up on TikTok so fast and then fade so fast, that if turnaround is too slow, you’ve missed the cheap real estate in terms of attention economy if you’re trying to scrap your way into, how do we get some attention out of this, away from the other places?

Mark Richardson:

And with B2B, the funnel might look different based on your audience. It may be more appropriate to reach a more junior member of a marketing team or an agency or a creative team with TikTok and Instagram, more top of funnel, whereas with someone who’s the VP or the CMO, maybe that Instagram interaction is more of a closer. And it’s helpful to know and be able to segment those cohorts to understand whether it’s demographics, Rand, like you were saying, if geo targeting plays a role in it, or if age, which oftentimes I found age is a huge determining factor in consideration and buying behavior. So really being able to more accurately map an expected customer journey based on those past behaviors is really powerful. And aligning content with that, knowing what to say, maybe not in the right time, everyone wants to say the right thing at the right time, but say a better thing at a better time, it’s the way I like to think about it.

Rand Fishkin:

I would love to tell you, Adam that SparkToro can do what you’re talking about and of identify a TikTok channel that’s rising or falling, but we don’t actually touch TikTok at all, intentionally. So we avoid it like the plague. And that is for three reasons. One is, I am very nervous about TikTok’s owner and ownership and generally nervous about extracting data from them. I think that there’s just high risk there.

Mark Richardson:

Fair, can’t disagree with that.

Rand Fishkin:

Second, I think it is quite plausible that in the near term, the United States and other western governments will actually ban TikTok from the app stores. And I don’t know that that’s a terrible move, I think that it is. I do not think highly of Mark Zuckerberg, I don’t have a lot of faith in Facebook’s ability to regulate disinformation or do a good job of shutting down false narratives that happen on the platform or prevent things like what we saw with the Alex Jones case and the Sandy Hook Elementary deniers and all that kind of conspiracy crap. But Mark Zuckerberg’s incentives are to keep the United States functioning as a democracy. And TikTok has the opposite incentive and goal, so I’m not super thrilled about that. And even though, yeah, weirdly, I find myself agreeing with Ted Cruz on this, anyway…

Mark Richardson:

Strange [inaudible 00:20:17] these before and after times.

Rand Fishkin:

TikTok is an incredibly entertaining platform. I think that one of the genius things that they figured out is that there’s lots of content that you probably would enjoy, but would feel bad clicking on and wouldn’t go seek out. And so they’re just going to present it to you without you having to click on it, without you having to choose and select it. And then the algorithm just pays attention to what you go away from.

Adam Kerpelman:

Or without it being tied explicitly to a social graph.

Rand Fishkin:

Yeah, yeah, exactly not tied there-

Adam Kerpelman:

Because your friends liked it.

Rand Fishkin:

Yeah, I didn’t say I wanted to follow this person, I think they’re terrible, I think everything they do is awful, but that one video from them was really entertaining, so show me that.

Mark Richardson:

How do you get validation on one piece of content versus an entire account?

Adam Kerpelman:

Yeah, so I should have said short form video.

Rand Fishkin:

The third reason, and this is more toward how SparkToro functionally works, one of the big challenges with TikTok is that it’s a place where you do not express your professional and personal interests in a structured kind of way, the way you would on every other historic social network. Unlike Twitter, Twitter for example, I go and I follow the people and publications and sources where I’m like, yeah, I’m regularly interested in the things that this person posts. Whereas TikTok it is very much kind of a trends thing, it’s a lot of entertainment. Very frankly, if you think about this Netflix for example, there’s not a ton that a marketer can do. There’s some small things, but there’s not a ton that a marketer can do if I tell you, hey, these people watch these six shows about Japanese children who go on their first errand. You’re like, all right, yeah, that’s an entertaining show, I don’t know if I can do anything with that. There’s not a lot I can do.

But on Twitter, on Instagram, on LinkedIn, especially in B2B, even on Reddit, subscribing to a subreddit, you subscribe to the landscaping subreddit, I know what I can do with you as a marketer. You subscribe to a YouTube channel about Dungeons and Dragons, I know what I can do with you as a marketer. You subscribe to half naked dancing person on TikTok, I don’t know what to do with that.

Adam Kerpelman:

Yeah, the way we end up talking about TikTok and really short form video broadly is, it’s way top of the funnel. Its a replacement for cable TV. It’s some rough demographics and that’s where people’s brains are at the time and you end up associated and maybe you catch them, and because it’s digital, the funnel’s different after you catch them, but it’s not… Well for one, we don’t have access to the data, but if we could get that data back, it would mostly just be demographics and interests for a particular thing. And again, for what? To know that we’re targeting them correctly in the right channel?

Rand Fishkin:

I mean, I think that it-

Adam Kerpelman:

Makes sense, I guess is what I’m saying.

Rand Fishkin:

The interest data can be really useful. And the challenge is just that TikTok has this weird, well weird compared to historical thing, where they show you things that they know you’re interested in, even though you don’t subscribe to them. And a lot of people talk about how TikTok just gets me, I don’t have to tell TikTok, hey I want to follow this person, or hey I’m interested in this, they just sort of get me. And so SparkToro obviously could only see what you actually subscribe to, which is not the… that doesn’t tell us what content you’re being served.

Adam Kerpelman:

Yeah, because likes are private.

Mark Richardson:

And comments, you wouldn’t be able to call those pieces of data.

Adam Kerpelman:

A lot of the TikTok rabbit hole though, the idea of the cheap real estate is the thing that’s more so chasing there, if you want to catch a trend out of that conversation, you don’t necessarily have three month turnaround in a lot of cases. And in B2B stuff probably moves, but if you’re trying to play an influencer game and you realize, Oh, attrition’s happening here, or this is a hot podcast, look, we developed the podcasting muscle so we can go be on other podcasts in addition to doing our own.

Mark Richardson:

Doing a really good one, hopefully for ourselves.

Adam Kerpelman:

Knowing which ones to target is part of the game of like, okay, let’s go be in the right places to grow this community that we’re trying to grow here.

Rand Fishkin:

Ultimately.

Mark Richardson:

If nothing else, just the R&D capabilities of SparkToro in themselves for media planning, for partnership alignments, for integrated deals, and potentially sales enablement, the applications are really myriad. So I think it’s kind of a question of Swiss Army knife, how do you want to use it? What’s the best application for your particular team?

Rand Fishkin:

Yeah, I think this is really… It’s a research tool, so it is designed to say, if you have the problem of, I’m running a YouTube bad campaign, I really need to find out what these audiences that I need to reach subscribe to, and I don’t trust Google to tell me and just do the targeting automatically, which you should not. Let me go figure that out. If you have questions like, hey which social accounts are really popular with my audience because I should invite those people, the people behind those accounts to my podcast because that’ll bring the right audience to me, SparkToro is great for that. Same with, for an event manager or an organizer to say, which speakers should I bring to my event? Same for a media planner to say, hey where should I go with my spend. For someone who’s in PR to say, hey which sources of influence… where should I write a guest editorial?where should I pitch a joint research project? Who should I do a co-marketing event with? When you have those questions, you need a, this percent of your audience pays attention to this place.

Adam Kerpelman:

So then the next thing I’m interested to tie it back to for the purposes of the marketer is the attribution versus lift thing. So again, to make it personal, on our side we are locked in many projects to try to clean up attribution for our digital leads and signals and all that kind of stuff. And it’s this game of like, the tech can’t get better, faster than either new platforms that are emerging or regulation or cookies going away, there’s just so many conspiring forces that are ruining the smorgasbord we’ve just had for 10 years of programmatic ads based on cookies. [inaudible 00:27:06] cookie list for a reason and Dun & Bradstreet bought us in part because we said, hey cookies don’t seem like the right solution for the happiness of humanity, and so they probably go away eventually.

Mark Richardson:

But the great thing is, now we have that to ride off of and just propagandize. We have Google’s…

Adam Kerpelman:

Well that’s the part of us marketing to marketers.

Mark Richardson:

We have the-

Adam Kerpelman:

To shift in attribution-

Mark Richardson:

[inaudible 00:27:31] leverage.

Adam Kerpelman:

So yeah, talk about that a little bit, what do you mean when you say that in terms of what you’re bringing forward to leadership? If you’re in a position like mine saying, okay, we’re not going to be able to say that was a marketing sourced lead with the clarity that we used to have, but what we can say is, when we did X, Y, and Z, numbers went up.

Rand Fishkin:

So lots of thoughts around all these things you guys are talking about, when it comes to attribution and lift, first off, we do have massive change that is driven by Google eventually getting rid of cookies, which they keep delaying, but it will happen. The Apple and Facebook fight, which means that Facebook’s algorithm can’t include as much data from iOS devices, and a lot of marketers have noticed that. And then you have all of the ad networks, whether we’re talking about Amazon or Etsy or Google or Facebook or all of the third parties that can’t get data as high quality as they used to because of things like GDPR and the California Privacy Act and Canada’s new laws. So there’s just a whole bunch of changes. Almost all of these are driven by big tech lobbying to build protective walls. I think there’s a misnomer that ordinary internet consumers care a ton about privacy and don’t like seeing personalized advertising. I don’t believe that, I think that any of us would happily trade all of our personal data on all of our devices for 5 cents. And in fact, more personalized advertising is generally seen as a better experience, not a worse one.

Mark Richardson:

It’s a better web experience, I think it’s a comfier web experience. You feel less bombarded when you see things that reinforce your existing affinities or interest or shopping.

Rand Fishkin:

It’s creepy, but also we’re all used to it. I think there’s a small and vocal percentage of internet users who are like, this is terrible, my internet privacy, you should be worried about it, yada, yada, yada. And Facebook and Google are like, yeah, yeah, yeah, worry about privacy, worry about privacy. Because basically that means we’ll never have competition. Because first party data, zero party data that the big networks can collect means that they have an advantage that no one else can ever take away from them. GDPR is the one to me that just shocks me. It’s essentially Europe saying, hey, let’s pass a law saying no European company will ever compete with the big American tech giants. Does that sound good? Yeah, yeah, yeah, we don’t want to innovate. Good, right? Do it. It’s just bizarre. So what this means for marketers though, is that a ton of what you currently see in all of these advertising platforms and even some organic stuff, is essentially what the platforms want you to see.

So, Google, for example, might say, we know that the people that you’re targeting visit these websites, search for these things, go to these places, we’re not going to tell you what all those are, but we are going to show your ads to them. We know they were already going to buy from you, in fact, like 80% of the people that we show your ads to, they were already going to convert. But we are thrilled to take credit for that and we’ll show it to you in your reporting, and you can bring that to your CMO who can bring it to your CEO, who can bring it to your board of directors and say, look, our advertising efforts with Facebook, Google, Amazon, are paying off. That’s exactly the goal. Their incentive is not incremental lift in conversions, it is take credit for all the sales that would’ve happened and some of the ones that wouldn’t have happened.

And so the only way as a marketer to truly know whether you’re advertising is giving you incremental lift and incremental sales on top of what you would’ve gotten anyway is to shut it off, that’s it. There’s no other choice, you basically have to shut it off for a significant portion of time to see, hey, how much were we really getting from these Instagram ads, these Facebook ads, these TikTok ads, these Twitter ads, these Reddit ads. And yes, from the experience that I’ve seen, especially for midsize and larger companies who shut off these ads, is there’s usually a massive amount of wasted spent. 50%, 70%, 80%. You look at the Airbnb case study, the Uber case study from a couple years ago, the Chase Manhattan case study that was written about in the New York Times, the eBay case study with branded search ads from a few years ago, all of these case studies that get written about and the hundreds of thousands that don’t are telling us the same story, which is that these networks are taking credit for a massive amount of sales that they did not create.

Mark Richardson:

Its like you were spying on us, we just kind of shuttered our Facebook program for that exact reason.

Adam Kerpelman:

Yeah, we’ve been going through a cycle of turn things off for a little while lately just to see what’s what. It’s really satisfying when it does what you expect.

Rand Fishkin:

Oh, yep, yep. That killed exactly the traffic I expected. The frustrating part about all this is that, on the flip side, if you were to say, fine ran okay, I’m going to shut this off, what do you think I should do? What should I actually do? And that the hard answer that infuriates people, but that is true, is that you have to invest in hard to measure or impossible to measure serendipitous marketing channels. Including a lot of organic stuff where there’s no way, no way, you will ever be able to say, because I went on the Data Driven Marketer podcast and talked to Adam and Mark and we talked about SparkToro, three people went to SparkToro’s website and signed up for a paid account. I’ll never be able to see it, I’ll never be able to know, I will never be able to attribute it, and I should probably be on this podcast. And the same thing is true for a whole ton of activities that you can and should be doing. And that a lot of times your CMO your board of directors is not letting you invest in because you can’t perfectly attribute it.

Mark Richardson:

And that’s like the omnipresent, ongoing battle between marketing, because marketing is essentially, it is… Traditional marketing, I’ll say is almost impossible to get true ROI on a an out of home campaign or a Super Bowl ad. How do you really quantify the ROI? We had a guest two weeks ago who was intimately involved building, I think it was the Doritos, the latest Doritos Super Bowl ad. And it’s just like, you spend the money to get views, that’s it. And it’s not about did they buy the bag of chips? You’re never going to go back and quantify, unless you look at lift over several months from exposure to the Super Bowl campaign, annotate, when did it run? And who’s your audience? What did they… Did you do a focus group? Did you do a sentiment analysis? Maybe you can dial it in with that, but sales tell the tale, and it’s hard to know exactly which piece of content drove the sale.

Rand Fishkin:

Absolutely.

Mark Richardson:

But it’s a great way to start at some cottage industries in the digital marketing universe.

Adam Kerpelman:

Well, the funny thing it makes me think of is, what you were saying, but also to Mark’s point, somebody on Slack shared a video, it was Rory Sutherland from Ogilvy talking about just making cool things. And the quote that ended up being the conversation in Slack was, “Meaning tends to be conveyed by things that are slightly gratuitous.” That can also include… Nobody needs the Super Bowl commercial, it’s not a critical place to be if you want to sell any of the stuff that you’re trying to sell there, Chevy will move just as many trucks in a year without their 30 second spot with…

Mark Richardson:

But will Coinbase get people investing without that little QR code?

Adam Kerpelman:

I bet there was subset of people that were like, I don’t understand what that QR code was. But the people that get it, appreciate the gratuitous little extra bit of like, well, that was a little extra.

Mark Richardson:

It was out there, it was weird. Sometimes the weird… I think the video Adam’s referring to, he talks about the San Pellegrino, the fact that they have those little, The foil cap.

Adam Kerpelman:

Oh yeah, the little foil cap.

Mark Richardson:

And [inaudible 00:35:52] is saying, I’m sure every year there’s a line item that finance director is like, we got to get rid of that foil because it’s costing us, our unit cost is through the roof.

Adam Kerpelman:

But non-trivial and a lot of money.

Mark Richardson:

Non-trivial, but it’s part of the brand. It’s part the experience of San Pellegrino and it differentiates. And that’s part of the enchantment that I think Sutherland’s getting at, is this, this gratuitous is sometimes necessary to just provide that extra level of [foreign language 00:36:24]

Rand Fishkin:

I mean, hard to measure channels tend to outperform easy to measure ones because competition is lighter, because people tend to pay attention to those channels, and you can’t just pay to have your message sent through them. So the things that have moved the needle the most for SparkToro over time are very popular podcasts and YouTube channels, email newsletters, events featuring our work. And I can’t pay for it, I can barely track it, I can only really see it in lift over time. What I’ll see is, hey, this thing came out, I see what are called vanity metrics by marketers now, vanity metrics like views and impressions. And then I will see the impact of it slowly over time. And I’ll get some emails from customers and they’ll be like, I heard about you here, I checked you out here, we have this need. And oftentimes it’s months or even years later, someone will be like, Ah, what was that tool where the guy talked about being able to see behaviors of an online audience? I need that now, where did I hear about that? And then they’ll search for Data Driven Marketer podcast, they’re like, Yeah, okay. Oh, here it was, SparkToro, and then they’ll go to Google and search for SparkToro and Google’s like, we caused the conversion, it was us Google, we’re the ones.

Mark Richardson:

Narcissistic Google.

Rand Fishkin:

They’re smart, right? That’s a smart way to take credit for-

Mark Richardson:

They know how to take credit.

Rand Fishkin:

The internet transactions by owning so much of the traffic stream.

Adam Kerpelman:

So past that, I have many thoughts, but we’re out of time for me to share them and continue. Thanks again.

Rand Fishkin:

Oh my gosh, my pleasure guys. Anytime, thrilled to be here.

Adam Kerpelman:

Once again, we have plenty of material for another episode, thanks to everybody for the listening. This has been another Data Driven Marketer, I’m Adam.

Mark Richardson:

I’m Mark.

Rand Fishkin:

I’m Rand. Take care.

Adam Kerpelman:

Take it easy to everybody.

Mark Richardson:

Thanks for listening to the Data-Driven Marketer. Our show is produced by Jessica Jacobson and Dan Salcius. This episode was edited by Steve Kosh. The data driven marketer is sponsored by NetWise, a Dun and Bradstreet company Any views or opinions expressed in this episode do not represent the views or opinions of NetWise or Dun and Bradstreet.