#011: These Are a Few of Our Favorite Insights

The following is a straight-up machine translation. It has not been human-reviewed or human-corrected. However, we did replace the original transcription, produced in 2017, with an updated one produced using OpenAI’s WhisperX in 2025, which, trust us, is much, much better than the original. Still, we apologize on behalf of the machines for any text that winds up being incorrect, nonsensical, or offensive. We have asked the machine to do better, but it simply responds with, “I’m sorry, Dave. I’m afraid I can’t do that.”

00:00:04.89 [Announcer]: Welcome to the Digital Analytics Power Hour. Three analytics pros and the occasional guest discussing digital analytics issues of the day. Find them on Facebook at facebook.com forward slash analytics hour. And now the Digital Analytics Power Hour.

00:00:25.26 [Michael Helbling]: Welcome to the Digital Analytics Power Hour. This is episode 11. We’re going to talk about in this show a topic near and dear to all of our hearts. It is the topic of insights and a few of our personal favorites. Over the years, we’ve managed to analyze some websites or some digital applications. So tonight, joined by my two stalwart co-hosts, Tim Wilson. Hello. And of course, joined by Jim Cain. Oh, Tim said it. I wanted to say it, too. Yeah. No, this is very good. And of course, I’m Michael Helbling. And so tonight, guys, this is exciting. We are all by historical trade analysts. We joined the web analytics ranks in the aughts or before, or the naughties, as they call them in Canada. And we all have gone through this process of creating analysis, delivering insights. And this show is going to be dedicated to some of our favorites. And here’s the encouraging thing for all the people listening. They’re probably not even that good. Like seriously, you guys can all probably do better.

00:01:38.01 [Tim Wilson]: No episode has made me face my imposter syndrome condition more squarely in the eye than the, oh, just jot down your most glorious insights ever.

00:01:48.90 [Michael Helbling]: Yeah. It is hard to talk about them as if they were your own, but in some ways, you know, in my career, I’ve always found the greatest insights and actions we’ve taken. I’ve always been part of a team. So I like to talk about it in that context as we delay getting to the topic. Let’s jump into the topic, shall we? What if we started one step back before we started into story time and we talked about what we think an insight is? Is that skipping?

00:02:19.61 [Jim Cain]: That’s exactly what I was going to say. So I couldn’t agree with you more saying the thing that I was going to say. Awesome. Tim, do you also concur?

00:02:28.13 [Tim Wilson]: I’m probably outvoted. This is where I feel like I realized I didn’t prepare by once again going back and reading the seminal work. on what is an insight from our friend to the north, Mr. Chris Berry.

00:02:43.84 [Michael Helbling]: It was exactly who I was going to reference, Christopher Berry. That definition of an insight is a great starting point. Anybody have it up on their screen can share it.

00:02:54.69 [Tim Wilson]: Well, is it OK if I read it off the tattoo on my forearm?

00:03:00.36 [Jim Cain]: I was so excited for a sec. I was like, you guys are going to compliment me for once. Oh, yeah, Chris Berry is really smart.

00:03:07.56 [Tim Wilson]: There’s some really smart Canadians. Let’s talk about one of them since he’s not here tonight. Oh. Very simple. Chris Berry, Mr. CJP Berry at CJP Berry on the Twitter for anyone who’s not familiar with him. The writer of Pithy posts of which many of them I can understand and many of them I realize I’m not smart enough to understand. But an insight is four things. New information. Executable. Causes action. profitable or in more detail and inside is a piece of information you didn’t know before which is the whole like if you stopped there it would just be you know interesting information executable and this can feasibly be executed culturally acceptable and of a scale relevant to the firm then causes action it causes a decision to be made that wouldn’t have been made otherwise and it results in profit or a sustainable competitive advantage which that last one is is the one that I think You say profitable and you’re starting to think that, oh, I can point to the income statement and say this number would have been X instead of Y. And I think there are definitely times that it is a positive business move, but not necessarily one that you can say I can directly tie to financial results.

00:04:18.39 [Michael Helbling]: Yeah, I would say that’s the only aspect of the definition that I would let people wiggle around on just a little bit, depending on where their organization is in terms of their ability to tie, especially digital analytics metrics and indicators to the actual company’s bottom line. Because when you can’t do that, no insight can really truly meet this definition because you can never prove that last part. But other than that, it’s just a super solid, well thought out definition. We are all in your debt, Mr. Barry. Perfect. Okay, now we can jump into story time. Unless, Jim, you disagree with this definition, which is technically your right to do so, since you are also Canadian.

00:04:56.58 [Jim Cain]: No, I mean, I don’t disagree at all. I mean, my notes in the show notes, which you guys, instead of reading Red Chris Berry’s blog, which I totally respect, were, you know, they’re similar-ish. Something that’s interesting is not an insight. Something that’s valuable is an insight, and kind of knowing what that is, again, needs to be tied to business value. And I think even if a piece of work meets all of Chris’s criteria, piece you said about drives action, implies it was delivered to someone who can take action. So a great piece of analysis delivered to the wrong person in the org or delivered improperly to the right person doesn’t count, right?

00:05:34.27 [Tim Wilson]: It’s actually funny because, you know, that’s definitely one of the kind of, get the bile into my throat, you know, the old actionable insights phrase, you know, that you’ll talk to a marketer who’s like, I want actionable insights and, you know, this definition of insight and then even kind of your, you know, similar, you know, same, same vein is kind of like, well, that’s, that’s like a key KPI, you know, an actionable insight. You could think of any terms of the English language that Yes, an insight wouldn’t necessarily need to be actionable, whereas the application of don’t call it an insight unless it causes action.

00:06:08.68 [Michael Helbling]: Yeah, didn’t Jim Stern write a blog post where he kind of broke apart the word actionable? And basically it’s a legal term. Basically means you could get sued with this insight or something like that. Oh, I might have missed that one. Oh, I feel like me must go look it up.

00:06:25.67 [Jim Cain]: Check that one up on the Facebook page when we publish.

00:06:28.22 [Tim Wilson]: I still sort of count as an insight if there is suspected information. Somebody says I think this is going on or if this is going on I could do something and there’s a pretty good chance that it’s going on but I don’t have any, I think confirming somebody’s suspicion or assumption such that they have the confidence to take action because they may say I bet this is what’s happening but I don’t have the data to show it. So I don’t know that you’re providing You know, new information implies that it’s just a total surprise. And I think you can have insights that are not surprising. It was somebody actually doing what analysts and marketers forget to do all the damn time, which is go navigate your own flipping website and say, oh, wow, I had a hard time finding that button. Or, gee, I can’t manage to submit the form on the first try on my own site. I bet other people are probably having this problem and then it’s not necessarily new information, it’s just quantifying and validating the yes that is the case. Now I’ve armed somebody to take action because they can say, you know, this is what’s going on.

00:07:36.53 [Jim Cain]: You know, I think a good insight presented badly that no one executes on is in an insight. And what you’re talking about is people have always kind of known this, but when I wrote it up properly and presented it appropriately, everybody got excited. I’ll bet you a shiny new tuning. that every single anecdote that comes up today will include a piece where someone goes, and then when I presented it to this person, we found together, like there won’t be, I wrote a one-pager and CC’d it to 40 people.

00:08:06.04 [Tim Wilson]: Yeah, that’s, I think, and I had to note that it was like a lot of, some of the insights that were like, I considered best insights because they led to action seemed obvious in hindsight, but You know, the data was kind of there staring people in the face, but nobody had said, oh, let me actually look at it and tease it apart. And then, yeah, to your point, in a lot of cases, it’s visualized. It communicated effectively, you know, show and stark relief. This is what’s going on. And then everybody’s like, oh, I guess that, yeah, I guess that is going on. And that’s probably not a good thing. We should change.

00:08:38.36 [Michael Helbling]: So here’s another level setting question, maybe. What if your insight is the thing that tips the balance for the business to do something that kind of knows it should do anyways, but hasn’t ever kind of pulled the trigger on? Is that still qualifying your minds as an insight if you deliver an analysis that shows and gets people excited and oriented around the course of action?

00:09:02.94 [Jim Cain]: I think so. I think that concept of drives action is one of the most critical pieces. Drive successful action is an optimal outcome. But I think Tim was saying earlier, there have been situations, everybody kind of knew it, but it took someone to go and see the numbers prove it. To me, that’s a huge insight.

00:09:21.88 [Michael Helbling]: I guess that makes sense. And it sort of is not like a new concept, but maybe the analysis is new information that is the fomentor of that change or that action. All right, good. I have some stories to share that in that case.

00:09:38.67 [Jim Cain]: Did you just steer this entire conversation? No.

00:09:42.66 [Michael Helbling]: I don’t know. I’d like one of you guys, please. We’ve danced around enough. It’s time to tell some true tales of analytics awesomeness.

00:09:52.15 [Tim Wilson]: I mean, I could definitely start with one of my favorites, because it’s kind of occurred again and again, and it’s occurred recently in kind of the more kind of extreme examples and it goes to what I, one of the core things I go to of tagging a site when I don’t have great guidance is if there’s a form, I’m going to make sure that I have tagged the forms, any forms and the form validation errors. A specific recent example was I was just checking a site trying to confirm that we were capturing registration data properly in the web analytics platform, comparing that to the CRM. And I went through the process and got this, I think the first error I got was your passwords doesn’t meet our requirements. And I was like, well, that’s interesting. You didn’t tell me what the password requirements were. And it turns out they’re really fricking strict. I bet that happens a lot. And then I went back and entered the password correctly. And I was really just trying to get to confirm what the tagging was on the confirmation page. And then I got this, a registration error has occurred. And that was the extent of the, of the error. So I was actually chasing the checking our data and realized that we didn’t have the form validation errors tagged. So we went back and tagged it. I didn’t actually say anything to the business. We just used the tag management system to go and put that in. And sure enough, it was like over half of the people who submitted the form were getting an error. And then the big no surprise was A big chunk of those were getting the, hey, your passport doesn’t meet its requirements. And then the number two thing was this mysterious registration error has occurred. And I had a hard time actually getting past that. I never could get past the error. It just seemed like it was sporadic and it wasn’t very descriptive. So that was one where I let the data run for just like a week. And then I sort of pulled it and put it in a simple visualization and said, I think I sent one note out to the team that was kind of controlling the process and said, I’ve occasionally seen this registration error has occurred. What is that? Do you know what that is? And I kind of got a, that’s some weird thing. It doesn’t happen very often. So I then turned around and went back to him and said, you guys need to really dig into this because look, when that error occurs, the percent, the conversion rate for the form goes way, way, way down. Cause people are, and they’re submitting, you know, comparing uniques to totals. I’m like, they’re, they’re getting it like three and four times when they get it. And then they’re bailing and being able to put that in front of them to say you’re driving a lot of media to try to get people to fill out this form. And then, and they’re willing, but they’re not able. So that’s, that’s definitely, and I bet in my career I’ve had probably four or five cases where actually, Outlining this is what’s really happening on your form people can’t get through it easily Who’d you deliver that to Tim? I delivered it to the well kind of simultaneously to the Brand manager and the agency that owned the the site to one person There I had like a standing bi-weekly meeting with them and so I just said hey I was kind of poking around and I found this and that sent them off to kind of research what was the issue.

00:13:11.40 [Jim Cain]: And if someone’s listening in and they’re a user experience guy, are they going to go, that was an analysis. No, I’m just curious on that one because I think it’s analysis.

00:13:19.24 [Michael Helbling]: That’s what I think was great about that story is that that’s where the analysis started was with your own user experience and assessment of what it took to do that action on the site. I find that usability and user experience is just an amazing place to start for site-side insights and optimization. especially for things that are like just blockers for people. A similar situation which was we had this weirdness in the data that there was sort of on-site search and when people used the search the call to action for using the search wasn’t super clear in that there was a button that said search and there was a form field right next to it and what was happening is people were clicking the button before entering anything And the site was conducting a search on basically the enter your search phrase.

00:14:11.09 [Tim Wilson]: Enter your search.

00:14:11.81 [Michael Helbling]: Oh, yeah, yeah. And giving back a page of like, what are you talking about? I have no results to share with you. And that little change. So we just wrote it up. We said, here’s how many users it’s impacting. Here’s what we think the business value of this is based on the revenue that we see going through search. Deliver that to leadership and move that into prioritization to get fixed and updated. And saw that update take shape. But that’s like one of those, it’s just so simple, like that, it’s not like we changed the world that day, but that was a slight user experience improvement that we could prove with data. And that’s what I like about it is, sometimes UX and usability is so subjective initially, because not enough proof for some people has been delivered, right? Because it’s, you know, oh, those users in the lab, they don’t know what they’re talking about, that, you know, that kind of thing. But then if you can back it up with, hey, 30% of our site visitors also have this same experience, now you’re talking about a pretty good size population and you’re impacting a much bigger segment at that point.

00:15:15.91 [Jim Cain]: So are you saying a UX guy would go, I think this is happening and based on our best practices, we should fix it. And an analyst will go, I’m pretty sure this is happening. And now that I see it, I’m going to see if the data validates my hypothesis. I think it’s a totally different approach.

00:15:32.64 [Michael Helbling]: Well, it basically is that step of, hey, me and the UX guy both agree that this is probably something we should fix. But to get the rest of the organization on board, we need to show them something better than a couple of videos from the usability lab.

00:15:47.84 [Tim Wilson]: It’s not going to be the UX guy who designed the experience to start with. It’s the UX guy who inherited, who’s kind of working on tweaking the site. And it’s driving him nuts because he thinks it’s such a miss. And which goes to the criticality of being, having a good relationship with multiple organizations and being able to partner with somebody in UX who says, yeah, this seems like it was a, you know, our navigation, our secondary navigation has no pause. So if you accidentally are trying to move to sub navigation and you slip off of the sub nav, it disappears. We need to put a one second delay on it. It’s like, okay, how can we, how can we figure that out?

00:16:27.81 [Jim Cain]: Let’s do a test. Yeah. We’ve never been successful at really partnering with people in the UX team. It just seems like we speak different languages. That’s maybe a skill that I can work on. My anecdote that slides into this particular theme is that we had a customer a couple years ago, and again, we do a lot of retail. And they said, our Canadian visitor conversion rate is terrible. We ship a contiguous North America, you know? So what is it? Is it search terms? Is it this? Is it that? And we spent a lot of time in the data until we finally ran a test transaction, and we realized that on the enter your shipping information page, because the Canadian postal code is a space in the middle, so it’s like three digits space, three digits, and an American postal code doesn’t have a space. And whoever wrote the form wouldn’t let you put a space in. You had to do it all as one thing, or else you’d get an error message. And the second that we tried to run a test transaction and it pissed us off, we went, I wonder how many people this happens to. And then, you know, as Tim was saying earlier, we measured the hell out of the form, clears the nose on your face. That’s where one of the biggest bottlenecks was. Yeah.

00:17:34.71 [Tim Wilson]: Yeah. That’s the Nancy Coons and I wind up having that conversation often. Like when an analyst is told to go, okay, can you go look in the data and figure out what’s going on with this? If an analyst doesn’t, the first thing you need to do is go through this and actually understand Which I think is two things. It’s one, it’s actually capturing the user experience and then it’s two and maybe this is, I don’t, I am a consultant so I work on a bunch of sites so I don’t just remember how everything is tagged on every single site. But even when I was on a single site and had done the tagging and do it intimately, I still wanted to check the tags were firing as intended and not just kind of leaping to, oh, what I see in the tool is, is what’s going on. Yeah. All these examples were saying, huh, we went and looked at it and it seemed like we basically developed a hypothesis and then went and said, how are we going to Can we validate that?

00:18:29.68 [Michael Helbling]: And I think there’s a couple of really cool things that you just alluded to, but maybe you didn’t, but I’m going to act like you did so I can bring up something.

00:18:37.70 [Tim Wilson]: If they’re really cool, I’m going to say that’s exactly what I meant.

00:18:39.86 [Michael Helbling]: Yeah, exactly. But there’s a couple of things you do after actually doing analysis or before doing analysis of the data, and you’re kind of saying, hey, here’s an important step in developing an insight, which is, Validate the underlying data in the first place, right? Is this actually worked the way we think it’s working? Which is really important. And then post analysis, it’s the delivery of that analysis and the form of delivery of that analysis. When you deliver analysis in the context of the person who’s receiving it, in other words, you speak their language, So let’s say this person is not a digital website person. They’re a marketing person. Then you don’t talk about digital metrics, or maybe you could talk about conversion rate, but tie it to metrics that map to them in there. That’s huge for getting analysis adopted. Because if you say, hey, this is the impact on EBITDA, you know, earnings before taxes depreciation and hybridization. Oh, yeah, yeah, yeah, we’re all, yeah, yeah. Yeah, OK. There might be like a brand new analytics person out there listening for the first time. That’s right. And, you know, giving that context to the analysis is a big eye-opener when you can.

00:20:03.79 [Tim Wilson]: Not everyone can sometimes you have to go for the back so you know lead generation You take you take and you that’s where you start doing a little bit of somewhat a sumptive model say wow if we kept the number of people in half We know what we’re spending on media. We know what our conversion rate is I can do the math for you and say you know what if we just got this to happen half as much then our CPA is That’d be our certified public account. I mean, our cost per acquisition, whatever rate, whatever they’re focused on, you’d say, this would actually move this in a positive for the business by lowering the value direction. So yeah, getting it to what they care about as quickly as possible, or it’s the raw volume. This would increase the number of leads for the same price, but figuring out what it is they’re more concerned about.

00:20:55.26 [Jim Cain]: So two key things for a really great insight is stakeholder relevance and opportunity cost. Oh, that’s a good point as well. Okay. Yeah. I mean, I think, I don’t even know if I’d call this an insight, but one of the biggest wins I’ve had in the last couple of years was getting the CEO of a fairly large client to stop asking for bounce rate on his executive summary, because he went, like we spent half an hour talking about what bounce rate means. And he finally went, Oh, I get it. It lives here for these reasons. Absolutely get that off my page. And it also meant that the level of questions he started to ask went through the roof. Like is really good education considered an insight or that’s just something also awesome that an analyst does?

00:21:40.84 [Michael Helbling]: Yeah, that’s just something also awesome. I guess kudos to you on the win because that’s one huge part of the process. A lot of times is elevating the conversation to the right level. both for the analyst and the stakeholder to be talking about the right things at the right time. Yeah, because if your CMO or your president is stuck talking bounce rate, unless you make money by not bouncing, you know, there’s limited utility in having that conversation. So anybody want to start with another zinger? Your turn. Your turn. Yeah, go Jim. Give us a Canadian insight.

00:22:18.70 [Jim Cain]: So there was a beaver dam in my backyard.

00:22:22.64 [Michael Helbling]: What a STEM wanger.

00:22:23.85 [Jim Cain]: Well, I mean, one of the ones that was on my list are, you know, just a couple of gimmies. Maybe one of them is on coupon code fields. So the very first time I did this, we were looking at large site retail and they wanted to figure out whether it was worthwhile having an affiliate marketing program because it was very expensive, but it looked like it was shooting the lights out. And once we started to look at it in more detail, we realized what most analysts who do a lot of retail know, which is that coupon code fields are kind of terrible and you shouldn’t have them. Moereover, a very large percent of the people who use coupon codes were driven to you by coupon stuffers like RetailMeNot. And when you start to actually look at the data, A lot of times you’re actually shooting yourself in the foot twice. I’m on the shopping cart. I see coupon code. I immediately leave that site and I Google brand name coupon code because there’s no way I’m giving up a discount. So I abandon the shopping cart. Then I come back through to the shopping cart. through a coupon affiliate. And so not only did I lose the original sale, now I have to pay arbitrarily five to 15 points on the sale to someone who doesn’t even deserve it. They did nothing, right? And once we started to build a model to show not all affiliate is bad, people like this cost you a fortune. Also maybe having a coupon code box period is not a good idea. It led to some really significant wins and it turned into something like that’s a follow through for me and I wonder if you guys agree but I think any analyst with some experience has their playbook of these are insights that I’ve generated in my career that are home run swings every time they’re my best practices like that coupon code thing I’ve applied over and over again it’s become a best practice and to me a great insight tends to become what other people would call my least favorite phrase but low hanging fruit

00:24:28.20 [Tim Wilson]: You know, well, I think I don’t know the best practice, right? There’s been there’s been lots of justified. I think the best practice can be a dirty word because it’s I think it’s back to what we were talking about earlier that you can say, you know what? I’ve seen this five times and five times. It’s been the case. It’s a best practice, but still let me.

00:24:49.65 [Michael Helbling]: Oh yeah, always going in.

00:24:51.36 [Tim Wilson]: Trust the verified type and say, does this apply? Because you don’t know when you’re going to hit the situation where it’s not. It’s a case where no, your price point, people won’t buy if they don’t get that discount. You’ve got a little bit of a pricing issue and you’d actually lose the business if people didn’t find a discount somewhere. Now you’re having to pay the affiliate a little bit of that, better just to lower your prices and not be arbitrarily paying a middleman.

00:25:17.22 [Jim Cain]: Oh, and don’t get me wrong. A best practice for an analyst is something where you go, I could save myself 10 hours in terms of running this playbook on the data and I could be right. I could do this because it always works.

00:25:29.68 [Michael Helbling]: Yeah, and this is where I think experience and intuition, coupled with data, creates really strong analysis. Because you’ve seen those things and you’re like, hey, you know what? Half of the traffic goes through on-site search. I’m going to spend some time analyzing on-site search.

00:25:44.63 [Tim Wilson]: But that can be, I mean, just to be clear, that doesn’t have to be the intuition of the analyst. It’s the, we’ve covered this before, that there’s the, oh, now we have all this data. I don’t need to think independently. And I want the intuition of the CMO and the director of marketing and the 22 year old intern, well, maybe not the 22 year old intern. Where’s our millennial crack? I think he just made it.

00:26:08.74 [Jim Cain]: That was it?

00:26:09.76 [Michael Helbling]: Yeah.

00:26:11.65 [Jim Cain]: Get off my lawn.

00:26:14.17 [Michael Helbling]: I like millennials and think they’re a valuable part of our industry. You’re still hiring. Visit us at hr.searchdiscovery.

00:26:22.45 [Tim Wilson]: Do we have them?

00:26:24.74 [Michael Helbling]: I don’t know.

00:26:25.75 [Tim Wilson]: We’ve got another one that was kind of fun. A lot of these are ones that I’ve applied them multiple times, and they are a go-to. Hey, I bet if I throw this into a little heat map, it’s going to show the same thing. And it was digital. It was social, not web analytics. But it was trying to kind of tackle time of day, day of week for social media, where people said, how often should I be posting on Facebook, in this case, specifically Facebook? But you can do it with Twitter. You can do it with Pinterest. How often? times of day, what days of the week, and consistently what I find is that because the people who are doing the posting have to have a certain volume, they get into a pattern and they sort of have already settled into a we’re gonna post twice a day at 10 a.m. and 4 p.m. and they put some logic into it saying well based on our audience this is when we think they’ll be testing Facebook and then they have these little oddball cases for one reason or another where they post it on a Wednesday at 9 p.m., and you’re taking that data and putting it in a fairly simple, say, I’m going to look at the reach and the engagement of that content. pull back and say, you realize you were in a rut, like you were treating this like email, and you were hitting the exact same time, and there are likely, and we can show when you have these other random posts, these few posts that fall in other slots, that those actually performed quite well, maybe you don’t want to be treating it like email. We send our emails at 9.30 on Tuesday mornings. and treat it more like social media that people are checking throughout the day and your followers and the people who will get exposed to it, you’re probably going to do better with the same volume of mixing it up more. And by the way, that would give us a better mix of data where maybe we do find consistently dead spots and consistently hot spots. You kind of refine when you’re posting and that’s happened multiple times when you put in from say you’ve never posted before 10 a.m. And don’t you think people check Facebook before they go to work like oh yeah I guess they guess we do and I’m thinking this is interesting dude because your two examples both of which I think are cool.

00:28:38.42 [Jim Cain]: But instead of you going, someone got me to go fishing and in the data, I found this, you’re the one forming the hypothesis of both of these scenarios.

00:28:47.42 [Tim Wilson]: No, no, in that case, there was the question that, well, the question that came in was, what’s the optimal time of day and day a week? And I went into it thinking, oh, well, you’ve been posting all over the place. So I’ll just do it. I’ll find the optimal time of day and day a week based on when you’ve been posting and realize that I had no data for certain times of day and day of the week. So really the, The first level and the first order inside I delivered was I can’t answer your question. If nothing else, you have to give me a few weeks where you do mix it up so I can actually get a data set to work with, but kind of inadvertently delivered an inside along the way that, and then I could rationalize it, which is part of the, part of the insights too. If you say, I found the data says this totally bizarre, crazy thing and I can’t explain it, then Yeah, don’t present that insight until you can actually rationalize why it’s happening. So yeah, I had to put the story behind it to actually say, oh, I don’t have the data. I understand what they’re doing. I understand how this is probably negatively impacting them and put that story in front of them. So I couldn’t even get to the question they were asking until we’d solved kind of a more base issue.

00:29:52.72 [Jim Cain]: But it sounds like, I mean, to use classic Wilson vocabulary, they pointed you down a road to the question, and you’re like, I can’t answer it. But I believe that your email send times are too structured. And if I’m right, I’m going to recommend based on data we haven’t captured yet to tell you to send at different times of like you formed a hypothesis for them and then validated or invalidated it based on their original ask. Is that the difference between a good analyst and a great analyst? Like instead of waiting for a hypothesis to come to you, if there’s not a good one there, you help formulate it.

00:30:29.18 [Tim Wilson]: I think that’s more analysts versus senior analysts. If you look at building out your skill set as a marketer, because I think it’s dangerous to say, oh, I’m just going to get senior analysts, and I’m just going to let them generate the hypotheses. I want the marketers to be generating the hypotheses. And I want these, with those little things happen, where I wind up generating it, I want that to be spawning them to say things that I can say, oh, you just stated a hypothesis, let me go validate that, because I’m most happy and successful when somebody else who’s got experience and ideas and motivation, and I’m listening to what they’re saying, and I’m translating those into hypotheses that I can validate.

00:31:09.57 [Jim Cain]: Did you like how I used your hypothesis structure, though?

00:31:12.46 [Tim Wilson]: I listened when you talk. I’m a little surprised you didn’t just take credit for it.

00:31:17.02 [Jim Cain]: Well, I mean, after all the kudos I’ve gotten for that hole, Kate doesn’t stand for a thousand routine, I do. I’ve noticed you’ve stopped saying it when you present that I ruined that for you. Probably. That’s almost worth it for me. So I can give another example. Yeah, go. And again, this is like a real tactical one. But one of the things that is kind of in my playbook is when we’re working with a major advertiser, we’re looking at everything. We’re not just looking at one type of advertising. So being able to create some insights around actual agency or campaign type ROAS tends to be full of wins. So for example, this agency is riding the hell out of your brand for paid search. And I hypothesize that if you were to dial it back 50%, you’d get almost all of that back in organic. But I don’t have the data yet, similar to Tim’s point from earlier.

00:32:11.86 [Tim Wilson]: Some of that stuff’s tough. Anything in social falls in that and the paid versus organic falls in that you don’t have the nice world of being able to A-B test, you have to say, let’s turn off our Facebook spend for a week. I had a client who did that. I kind of threw it out there and thought they would never go for it. I was like, look, the only way you’re going to figure out What kind of legs you get from this is if you just turn off your paid Facebook for a week or two, and I was more than expecting it to go anywhere. Two, if it was going to go somewhere, I would have liked to have kind of had some deliberate about that. And I got a call from our account manager. She’s like, they turned off their Facebook spin three days ago, and they’re going to keep it off for 10 days. And then you need to do some analysis and tell them what happened, which I hadn’t thought about that one in a while. But it definitely showed the total fucking scam that Facebook media is and then the disconnect between what their sales team says and reality of But I want to know what you think about Facebook paid media. I just think that I need the explicit tag on the podcast. There you go.

00:33:20.05 [Michael Helbling]: I’ve got another one in the tank, but I think we’re right at time. So instead, what I want to do is go around and have us wrap up a little bit. What did we learn here today?

00:33:32.28 [Tim Wilson]: Analyst will talk for a long time before actually describing a specific insight.

00:33:37.80 [Michael Helbling]: Well, that’s true too, and that’s interesting. But actually, you know, as much as the three of us were sort of like having a hard time getting going, I think we do like to talk about insights and things we have found in the data. And hopefully, it’ll be an inspiration to other people. But what about you two? What are some ramp-ups? Key takeaways.

00:33:59.45 [Jim Cain]: You know, I mean, I like hearing you guys’ examples, but I’d love to hear on the Facebook page, the Twitter account, other people saying like, here’s a win I got, and here’s how I did it. And to me, that was the other really big takeaway from today, which is that an insight sure as hell isn’t a one pager and Excel or an email. It’s part of a dialogue. It’s focused at a person. Like all the things we’ve been talking about before about what it takes to be a great analyst, all of those are required along with good focus and data and something that’s actionable. So that was really interesting. The packaging is as important as what’s in the package for an insight.

00:34:38.79 [Tim Wilson]: And I should probably admit that as we’re joking about the KKPIs, that it was actually Mr. Helbling a couple of years ago who pointed me to Chris Berry’s post on what is an insight. And I feel like that has come up in multiple lobby bars over the years. And it’s appropriate that it has come up It came up here as well, so as much as I jumped on it because I had successfully pulled it up, that is a credit to Mr. Helbling to find the valuable Canadian input for the podcast.

00:35:11.00 [Michael Helbling]: I’m always looking for a valuable Canadian, let me tell ya.

00:35:14.34 [Jim Cain]: And why am I always the douche, you know?

00:35:18.09 [Michael Helbling]: You are a proud co-host of this podcast.

00:35:22.63 [Tim Wilson]: I don’t, I feel like I have a big, I feel like we went, we’ve covered what I think is some critical stuff. I just, I got to mount multiple soap boxes and stand on your guys soap boxes, which are the same as my soap boxes about how to think about this stuff. I think it’s a, it’s a good thing to say, give me your list of great insights you’ve developed and to say, wow, you know, I am trying to ask myself every day when I’m talking to a client, you know, is this, Is it actionable? Is it something that is actually going to contribute to the business? And the reality is you still have to pull other stuff. But pushing back with that being kind of the goal that I want to give you something you can act on, that you will be able to, say, move the business forward. And that’s for 10 years I’ve been trying to do that. And it’s still not like I’m cranking out an actionable insight. once a day or even necessarily once a week.

00:36:17.63 [Michael Helbling]: I liked how some of the things we were talking about were, you know, discoveries of our own or came from other sources, but sometimes we’re focused on the user experience and I love that connection and I always have is the site experience or usability connected to insight generation because there’s a lot of ways to, you know, generate additional dollars. Like none of us really talked about you know, media optimization or media insights. Well, we did a little bit in terms of sort of affiliates and paid search and those kinds of things. But it’s interesting because there’s a lot of incremental value, I think, stashed in the pages of a site.

00:36:56.36 [Jim Cain]: We didn’t talk about media optimization except for every fucking one of Jim’s examples. Let’s take that part out.

00:37:04.62 [Michael Helbling]: I just realized that after I said it.

00:37:07.00 [Tim Wilson]: Well, but I don’t know. I actually think I’m a fairly hand-paid search. I mean, we didn’t talk about display media that when you say, how do you feel about media and media optimization? To me, that is blowing fruit and getting to the right person to make the obvious fucking choice on that is damn near impossible. Can we get, we’ll get our junior analyst, we’ll get our five year analyst, and we’ll get some immediate person to actually defend the way that they crunch data and just paint a rosy picture from it. Tim is not thinking about it.

00:37:41.03 [Michael Helbling]: That’s the only previous agency experience at this moment. previous current tomorrow uh last week anyway well um this has been a great episode thank you guys jam and tim for sharing some of the insights you’ve had over the years i like this conversation i think it’s a helpful conversation because it’s helpful for people to hear about The insight process and how we might have come up with an insight or been on teams that came up with insight. Obviously, if you’re listening to this show, you want to share an insight. We’d love to hear from you or you’ve got questions about this topic. We’d love to hear from you on our Facebook page or on Twitter. We’re also on Slack. There’s a measure Slack. Hit any of us up on Twitter. We can get you connected to the rest of the community. You can talk about this stuff. We’d love to hear from you. Again, thank you for listening for Tim Wilson and Jim Cain. I’m Michael Helbling, saying peace. Cheers. Get incremental now.

00:38:43.07 [Announcer]: Thanks for listening and don’t forget to join the conversation on Facebook or Twitter. We welcome your comments and questions. Facebook.com forward slash analytics hour or at analytics hour on Twitter.

Podcast: Download | Embed

Subscribe: RSS