What is user-centered design and why should we care about it? User-centered design is an approach to understanding your customer and designing an experience from their perspective.
Today, we have Dana DiTomaso on the podcast to share her in-depth process for creating a user-centered design. In this episode, you’ll learn how to discover the key problems in your user experience and what steps you can take to fix them.
Listen to this Episode:
- What user-centered design means for marketers
- Why achieving user-centered design starts at the top
- How to use analytics to understand what’s going wrong
- Why you must examine the “content consumption” metric
- How to approach your customer research
- Why most buyer personas fail (and what to do instead)
- The reason why you should focus on three key metrics
- The difference between reporting vs. monitoring data
- How to Use Emotional Targeting to Acquire Happier Customers
- How to Improve Website Conversion Without Sacrificing SEO or Speed
- Kick Point
- Claire Suellentrop on How to Use Jobs to be Done
- Jobs to be done framework
- Content Consumption: Go Beyond Pageviews
- Why Personas Are in Important in Marketing
- Full Contact
- Dana on User Personas: Learn Inbound 2017
- Adele Revella on Buyer Persona
- Follow Dana DiTomaso on Twitter
Louis: Bonjour bonjour! Welcome to another episode of EveryoneHatesMarketers.com, the marketing broadcast for marketers, founders, and tech people who are just sick of shady, aggressive marketing. I’m your host, Louis Grenier. In today’s episode, you’ll learn how to use user-centered design so you can stop guessing and start providing the right thing at the right time to the right people.
My guest today has typed her first line of code six years before I was born. She specializes in digital-first marketing. She’s a prodigy entrepreneur. She founded many websites, web design, and dev shop, SEO, digital marketing agencies, and she’s now the President and partner at Kick Point, a digital marketing agency based in Alberta, Canada.
She’s been nominated at the SEO Speaker of the Year in 2018, by Search Engine Land. She lives with her wife and three cats, who have 2.5 tails between them, and that’s something I need to know more about. It’s funny because, in her spare time, she yells at American football players while I would be yelling at rugby players. We have a lot in common it seems.
So Dana DiTomaso, let’s hear what you got and let’s get started.
Dana: Thanks for having me. That’s horrifying to know that I was coding before you were born. That’s a great way to start that off. Laughs.
Louis: Yeah. It’s a great way to make you feel good, right?
Dana: Totally, I feel awesome. I’ll go run a marathon later. It’s great.
Louis: Right. That shows me as well that I always try to code. When I was younger, I remember trying to code, like trying to design in Photoshop, and trying to do this kind of stuff, and never managed to do it, ever. I use photo tutorials and then never did it properly. You see. The thing you are much better at doing than me. And you did that six years before I was even here.
Thanks for spending the time to talk to me today. We are going to talk about user-centered design. It feels like this term, like user-centered design, or commercial reorganization, or growth hacking, or all of those kinds of terms, are like thrown out everywhere, right, in the marketing field.
It doesn’t seem like no one really understands what it truly means. From your definition, what does it mean, user-centered design? What is it about?
Dana: User-centered design … and I don’t even like the term user. I prefer to call it customer or client-centered, depending upon what you call people you’re selling stuff to. Is the idea that you are providing the right experience at the right time to the right people.
That requires more than just growth hacking, which, in particular, I really don’t like the phrase growth hacking. I feel like that’s just marketing without a strategy. You’re just trying stuff and seeing if it works, and I find it… I know. Go ahead and tweet at me about it.
Louis: No. No.
Dana: That’s fine.
Louis: I agree with you. I don’t like it as well.
Dana: Yeah, I find it really frustrating. Actually, I remember, if we are talking about history, I remember when growth hacking was first invented as a term and the blog post.
I remember reading it in my office and saying, “This is ridiculous. This is just marketing. This is what we’ve been doing all along, and now you’re going to call it growth hacking?” Ever since then, I’ve been annoyed at it. I feel like it’s a way to dress up a term, which is just put in the work, and just do the work, and there is no silver bullets.
User-centered design is a lot of work. It’s not something where you flip a switch or you buy a tool and then magically you’re providing better experiences for your customers. It means you actually have to listen, and start responding to people.
In particular, you have to respond to people who aren’t like you. That is a really difficult thing for a lot of people to put themselves in someone else’s shoes, not just in an “Oh, well, I guess if I was a 60-year-old woman, this is what I would want to do,” but really understanding what your product or service is going to for those people.
What sort of great concept of jobs-to-be-done, which Claire Suellentrop talks a lot about in terms of: When you’re thinking about your personas and what are the jobs they need to have done. Then when the sale is done, they still have jobs.
It isn’t just serving them in the pre-sales cycle, but it’s during the sale, after the sale, even if they leave you as a customer. What are you doing to make sure you’re serving them at every stage, and it is definitely a service type. You’re providing a service to people in many different ways and your marketing should also serve people and their needs.
I think that is a phrase we often talk about where marketing serves people, but we don’t really think about what that word “serves” means. Like you are at a restaurant and you order something, and the waiter comes back with something totally different. Surprise, that’s growth hacking. Nobody would want that.
Another great example that I give of user-centered design. For example, you look up on Google Maps how to get somewhere. You’re in an unfamiliar city, and Google Maps starts giving you directions. Then halfway through it says this trip has been sponsored by the San Francisco Department of Tourism and they take you past the Golden Gate Bridge. They take you past Coit Tower, all these different landmarks.
Then you are at your destination, but you had to go through all those different steps to get there. Obviously, that was a terrible user experience. Google doesn’t do that. I’m sure, at some point, someone offered them a whole bunch of money to pay extra to make sure all routes went past a McDonalds, for example, and would play the ad at that time. But Google hasn’t done that because that wouldn’t be serving their customers.
Sometimes it means making the hard decisions and maybe turning away from some money to make sure that you’re providing the right product. Really, Google Maps’ dominance in the marketplace shows that they have the right data, but they are doing the right things with it.
Louis: I can’t seem to understand how else companies and marketers and designers, all of those people are supposed to build experiences or do marketing any other way than what you just described, right?
Maybe because I’m very biased, I have strong beliefs about marketing as you might guess. I’ve talked to many smart marketers, like yourself, who believe in the same thing. So I struggle to understand. What else are they supposed to do providing the right thing to the right people at the right time?
We talk about growth hacking a bit as something that annoys you as a concept, which I completely agree with. What alternatives are people currently doing that is not user-centered design? What other stuff are they doing?
Dana: I find that a lot of people, particularly larger organizations, the bigger the organization is the more layers of crap you have to get through in order to get to the top. That means that everyone is in their own little silos. So one department is doing something and not talking to another department.
For example, we were just speaking with a client who is a home builder. They have a huge situation where they have a department that’s post-sale. They’re called the warranty department–not the customer care department–so it’s already not the great phrase for it.
There is only two people and this company sells way more houses than two people can handle. And you know what, in new home builds stuff goes wrong, that’s just how it is. Not every house is going to be perfect, that’s okay. We accept that, but you should go fix the problem.
Because this team is wildly understaffed they aren’t able to make any decisions. They have to check with the General Manager before they do everything, so there’s all this institutional just crud that’s built up that prevents these people from doing a good job.
So this department has huge turnover. Nobody stays, so nobody knows what’s going on. You have to train new people all the time. It’s basically a money losing part of the organization, which means that their reviews are tanking.
Now marketing is in trouble because they have bad reviews on Google. Well, marketing isn’t allowed to comment about anything that is going on in the warranty department because that’s the warranty department’s problem, right?
This is a perfect example of where everybody just sat in a room and agreed that we’re just going to cut out the political crap and we’re going to actually deal with this problem. If you get there, but at many organizations, that’s really difficult, particularly if there is a founder who’s still CEO.
They may have a lot of emotional investment on how things are done, because maybe that’s how they set it up when there was ten of them in a room. Now they’re a 500 person company and that stuff just doesn’t scale in the same way.
It’s important to take a step back. Or people who have been in the role a long time will think “I know my customer, and this is what they want.” That may not be what they want. Or there could be a flaw at some part of the process where, we’ve seen this many times when you say one thing in your marketing, and then the actual experience is something else, but it looks really good in your click-through rate, so they’re going to keep doing that.
Meanwhile, they’re not thinking about lifetime value down the road. If you can do something in your marketing, that means that people are going to stay with you six months longer. Or maybe it means that your click-through rate is going to go down, or your sign-up rate will go down a little bit.
But in the long-term, you’re making more money because you’re spending less on ads, because your click-rate is lower, and those people are staying around longer because you’re giving the right message.
That’s where all those different pieces really do need to talk together. Everybody has to be at the same table with this.
Louis: I think this is a very common concept to talk about and to repeat. It sounds so simple, right. It’s all about let’s start with the people you’re selling it to. Then you make a decision based on their best interests.
Whatever problems they have, you try to find a solution that suits them, that you can make money off, but that you’re not trying to take advantage of. It just blows my mind every time I talk about this things to people like you, who know their stuff. I really don’t understand how we’re supposed to do this.
In my small career so far, I can see how easily you can drift from beliefs that yes, you need to give a shit about your users to being siloed into your own small department forgetting about people altogether because the only thing you are doing is looking at screens every day and history report.
You completely forget that at the end of the day, those dots on your history report are actually people. So there is a big difference between what people believe and what they actually do, and the type of companies they work for.
Let’s do a small exercise together. I mean, it’s not going to be very small actually. It’s quite a big one. Let’s try to understand and go through a step-by-step process to implement–or to start thinking about user-centered design. Let’s not talk so now you have the results of digital marketing. I think we can pick something that is digital world that would be a bit simpler.
Let’s say you work with a client. You really realize that they are siloed, as you mentioned before. They don’t focus on the user. They focus on anything but the user. How do you go about implementing this user-centered design in their company? What is the first step?
Dana: You have to go right to the top. You have to start with the CEO, the President, the Board, the GM, whatever that leadership group is. You have to get buy-in from all those people because this cannot be a bottom-up process.
This has to be a top-down process. Because there is a reason why things are flawed, and unfortunately that means that somebody probably in a leadership role isn’t doing what they should be doing. At some point, they made a bad decision, and here we are. It’s really difficult for people to step back and say, “Yeah, I really screwed up. Or we really missed the opportunity on this.” Because it’s a really difficult conversation to have.
I find that’s why when we as an agency go in, we have a lot of freedom where we can say to the President, this is really awful. Here are 12 of the reasons why. Here’s the things that you need to fix. Are you willing to fix this?
At that point, sometimes we aren’t even giving them a quote. We’re just telling them this is what the problem is, are you ready to actually commit to fixing it. Sometimes in that meeting, they’ll say, “Oh yeah, totally, we’re really committed to it.” And then they’ll drag their heels on actually doing anything. It’s like, well nevermind, and we’ll walk away.
This is actually part of the reason why we like being a small agency, is that it’s okay to go into a meeting and just be really honest with people, and blunt with them, and tell them this is what’s bad, and this is what needs to be fixed.
If they don’t get their backs up about it, great. Then we can actually start to fix things. Sometimes, things need to get really bad before they’re willing to actually take a step back and say, “alright, maybe I don’t know what I’m doing.” For a lot of people, that is a really hard thing to do.
So you have to start there. You have to be willing to have those difficult conversations. If you are a marketing assistant number six, and there are four levels above you. Then if you’re sitting in the office trying to figure that out, then that’s harder. Because you have to start and you do have to work your way up to the President.
In that sort of situation, you do have to start pulling data and doing little tests and seeing how, “If we did this thing differently,” or “Let’s do this cohort analysis,” which is also a great thing to start with. Figuring out, where we’re doing a disservice to our customers?” Really just keep pushing that stuff up.
Eventually, something’s going to happen, but then you would have to get buy-in from your boss, who would then have to work on your behalf to get buy-in until you get buy-in from the top, and then it needs to flow down from there.
Because this isn’t just a marketing thing. This is an everything thing. Marketing is 1% of everyone’s job, sometimes more. If you are the person answering the phone, it’s more than 1% of your job. You are the face of the organization. I think that is something that has to come down from the top, where the CEO has to say, “Okay, look, this is what we’re doing, and this is why it’s important.” And everyone has to be there for that. That’s the first step: is the buy-in.
Louis: It sounds like this isn’t the first step, it’s the second step, because you mentioned gathering data, and you mentioned: “I would go talk to the leadership, and I would tell them what the problem is.” It sounds like you cannot meet until you understand what the problem is in the first place.
Dana: Well, you would know what the problem is because you would have bad returns in some way. Where you are implementing marketing plans and they are just not working the way you want them to, the way you expect them to. You’re probably missing targets, for example. So you already have this data because you’re doing bad job.
Maybe it’s not you doing a bad job, maybe you have fantastic SEO and you’re ranking really well for all the stuff you’re supposed to be ranking for. Then people sign up and then they are horrified, and leave. Or they’re leaving you bad reviews. At some point, there is a disconnect, and so that’s where the data is going tell you where that disconnect is. Then you can go up and say, “Alright, so we need to do a major overhaul to fix these particular issues.”
Louis: Where do you spot the disconnect? How do you spot it?
Dana: This is where you need to look at all the different systems, not just your little area. For example, a lot of people spend a ton of money on Facebook posts, just boosting Facebook posts. One of the things that we see is “Is this actually good value for your money?”
We can actually take a look at a metric, which we actually just introduced at MozCon called “content consumption,” which can measure if someone has spent long enough on a page to read the content on that page. In addition, if they scrolled down to the bottom of the content. If both these things are true, it measures it in Google Tech Manager, then it fires off a custom metric that we call content consumption.
What we look at for those boosted posts, are they being consumed or are people leaving right away? Because time on site is a really bad metric. Time on site only works if someone goes to a second page of the site. If you’re boosting posts to the same audience, they already know who you are, they don’t need to go to a second page.
Are they actually reading that content? If they’re not, okay, we have a disconnect. The content isn’t very good. Where are they bailing? Then you can start to do a little bit more digging and see where the problems are.
So this is where you probably have, and honestly, when I talk to in-house marketers they know stuff’s bad. They know things are bad and they know that there are these situations going on. They just don’t know how to get from everything is terrible to we have a plan. It may not necessarily be your job to come up with we have a plan, right?
It may be your job just to say everything is terrible and just start raising up flags and flares from there. And say, “Please pay attention to what’s going on here, because we’re spending all this money on our ad-buy. We’re spending all this money on offline advertising. We aren’t seeing the ticket sales, or the software sign-ups,” or the whatever it is that you’re doing, “That we should be. What is the disconnect?”
You can really start to look at your data and see that user behavior patterns bearing out in your analytics. Then from there you can say, obviously, we’re either targeting the wrong people. Which if you’ve done your research, you don’t think you would be or there’s some other disconnect at some other part of the process that is causing people to turn away.
Louis: This is an amazing example that you just mentioned the content consumption metric, which is an interesting way to look at what the user is doing from an IT’s point of view, and not only by just those methods that don’t mean anything. So that’s very interesting. Have you published anything online about this exact-
Dana: Yup. There is a blog post on our website that explains how it works. We have the actual Google Tech Manager code that you can go ahead and implement. We built it for WordPress, and we’ve included some examples if you’re on another CMS. We’ve built some implementations for clients who are on custom or even ASPX/CMSs as well.
It can work on anything because it runs through Google Type Manager, so you don’t need to do anything in your CMS itself. You just have to have Google Type Manager on the site, and you just need to know what you are going to measure and how. Once you have those variables, anyone can do it.
Louis: We’ll add that to the show notes, and make sure that listeners can check it out. It’s definitely in the episode page if you are checking the episode today.
That’s one example that is really practical and that shows the difference between tracking user-based metrics, and actually tracking metrics related to what people are actually doing, and to identify the problems that are staring at your face.
Can you share any other examples from your experience, anything that usually screams at you that says you’re not using user-centered design? You’re not talking to users? From your experience, what tends to happen the most?
Dana: Yeah, I think it’s people come in with really high expectations on where sales are going to go, and then there’s no actual reality in terms of what happens. Sometimes we’ll ask people, “Last year, did you hit your sales goal?”
“How much did you miss them by?”
“Oh, we missed them by $300,000.”
“Okay, well, that’s a third of your estimates. Why did that happen?”
And if they don’t have a good answer, now we’re talking about a disconnect in that way. That would be where, what are those sales numbers based on? Do people actually want to buy what you have to sell?
“No? You haven’t done that research?”
It’s interesting when you work with some organizations who, for example, if you’re a concert hall, and you are putting on shows, have you actually done the research to make sure people are actually going to want to go see the show that you’re putting on?
If you think about it from a concert promoter perspective. They’re not dumb. They know where artists are going. They pay attention to things like streaming services. Is an artist really popular streaming in a particular area. If they are, great, we’re going to make them go up there on tour.
Even though it may not look like it makes sense. Here in Edmonton, we don’t necessarily get a lot of big acts all the time because touring in Canada is very expensive. We are pretty far north. So people don’t necessarily want to come up here. But we had AC/DC come by. Why? Because they had streamed a ton here in Edmonton, and classic rock’s really popular here. The concert was really well attended even though it was freezing. I’m pretty sure it snowed.
That is where that data leads you. If you look at the music industry, they’ve obviously recovered from streaming quite well but concerts have always been profitable for people.
Look at those industries that have gone through this cataclysmic change and come out the other side. What are they doing? They take that data. They do that research. It’s really easy to tell which industries do sit down and do their research on who is interested in what. You can see that in a lot of different campaigns
Louis: Let’s say we know where the issue is, and we know we don’t have enough sales. We know that we have no fucking clue why people should buy from us, we don’t use their voice, and we don’t use the words they would use. Our experience is shitty, and all of that. So we know that.
That is step one to me because step two is getting buy-in or you try to get buy-in. I had a few listeners contact me about, for example, we have this email list of 10,000 people, and our open rate is abysmal like 10%. I know why, because people are completely disengaged, most of them. Most of the emails bounce back. Nobody gives a shit about what we do, except those 10%, but my boss doesn’t want to clean the list. Right? It’s a small example, but-
Dana: No, it’s a perfect example. Yup.
Louis: Who gets the boss to agree. At the end, I replied to the email and I said, “Have you tried that?” She said, “Yes.” At the end, I said, “I think you need to move. To move on.” It’s time to move on. From your experience, how do you convince people to change leadership? If you can’t, what do you expect people to do?
We have a bit of freedom, not being that employer, where we can come in and say, “are you actually happy with this information? Are you happy with how this is?”
Because if they say, “oh, well, no. I’m not happy that there is only a 10% open rate on this mailing list.”
Say, “Great. These are the changes I want to make.”
“Well, I don’t want any of those changes.”
“Okay, well, you have a choice. Either you are happy with the 10%, or you’re willing to make these changes. If neither are true, then you are irrational. You are making irrational decisions.”
At that point, go look for a new job. Sometimes after talks, people will come up to me with their problems. One person, their boss had entire call center Googling some keywords every ten minutes and recording their rankings.
I said, “You need to find a new job.”
That is horrible. She talked to this person, and he was just not convinced- that rank tracking software is useless. “The only way we can do this is through the call center.” She explained incognito in personalization. He was having none of it.
I was like, “You know what, you tried, you should just move on.”
I honestly think that that’s the thing. These bad companies with bad owners who aren’t willing to take a step outside themselves are going to end up with bad marketers. You know what, that’s okay. We can’t fix everybody. I was just tweeting about this recently because someone was saying about these mastermind classes where they sell a bunch of crap to people- learn these secrets, this one weird trick.
If people are going to pay for that and they are going to get fooled by that–if they choose to see the error of their ways and decide to do things properly later, that’s great. But if they refuse to accept that maybe that isn’t the right way, you can’t change people’s minds. You can’t make people feel a certain way. So they have to be open to change in the first place. In those kinds of situations, this is what you ask: “Have you brought this up to senior leadership?”
“No, because I don’t know the right words to say.”
Okay, so then we have a situation where “Great. We can help you find those right words to say, and the right ways to convince them.”
But if you say, “Yes. I’ve tried this and this and we can’t think of anything else,” Then it’s not going to work out. It doesn’t matter how awesome your team is, because if you have that kind of croft up at the top that isn’t going to let go, then your team is also isn’t going to stay there for very long either.
That sort of thing is terrible for morale. That is one of the big differences between successful companies and not successful companies. Is the leadership team actually willing to accept that they don’t know everything and is willing to take actual advice from someone? Not just someone flashy, but someone who actually knows their stuff.
Louis: Right. So let’s assume that leadership is actually quite knowledgeable or at least open to change. They know they need to change some stuff to make stuff happen. Now you’re in there, and you know the problems that need to be solved.
You started to hint at something that I think we can touch on, which is the research side of things. You naturally speak about it and you naturally take examples. But it sounds simple like this, but making and doing research, which might be step three. I’m not too sure what you’re going to say.
Dana: It is going to be step 3.
Louis: How do you approach the research as a whole? This is a big fucking deal, right? It’s a big thing.
Dana: I think it is an enormous thing because you have all sorts of different divisions and clients and everything else. Maybe just start with one little thing. I’m really a big fan of trying taking 10% and making a change and seeing how it turns out.
You could say, “Alright, out of all these different things that we do, one client of ours is an apartment rentals. One building is having a hard time converting. There’s a building that converts really well. They never have a problem with vacancies there.”
“Okay, great. We don’t have to worry about that building. You know that building is fine, so let’s not worry about that one. Let’s worry about this one specific thing that seems to be our problem child. So, what is it? Is it location? Is it bad photos?”
Interview people who have gone to check it out and just decided not to rent it. Talk to leads who decided not to close with you and ask them for feedback. Not from your salesperson.
Typically, if you have, say enterprise software, and there are sales people and they walk people through the demo process and they follow up a thousand times and eventually they get a no. What was actually the feedback?
Sometimes people can be very forthcoming if they know they are talking to someone who isn’t a salesperson at the organization. Someone who is a marketer, and they want real, honest feedback.
We did this once for a Chamber of Commerce. We called people who left the Chamber and said, “Why?” Boy, we got some stories. But they all had a lot in common, which was that the Chamber wasn’t actually taking their needs into consideration, and wasn’t listening to what they had to say. They were just representing the views of one small group of users.
After these phone calls, and talking to people, you get to know some of this stuff. This is one of the things we really like doing is actually picking up the phone and calling people and talking to them. Then the other half of it is this cohort analysis, where you can take all the email addresses of everyone who’s ever bought from you, or signed up, or downloaded white paper, but didn’t end up converting, and run it through a tool like Full Contact.
Look at their social information and find out what sorts of things they enjoy and they like. What are the ways in which you can speak to them, using the kinds of phrases that they use? Who are the influencers in their market? Who are they paying attention to?
You can have the best marketing in the world but if everyone thinks that this person over here, everything they say is gold. Then maybe you have to get that person on your side, instead of trying to convert these 100 other people over here, who only listen to the word of this one person.
Those are the kind of research has to do.
Louis: You said so many things here.
Dana: I know. Sorry.
Louis: No. No. That’s my job. Let’s break it down. The first step is this. I like this quality feedback, where you actually talk to people–which is a grandiose idea, isn’t it? It’s like a 2018 idea. It’s insane that we still talk about this kind of thing.
Anyway. Yes, talking to people actually makes sense. You said something interesting. You said talk to people who actually ended up not doing what you wanted them to do, in a sense, right? So talk to people who ended up not converting.
I think it’s also interesting to talk to people who actually ended up converting in the same way. But from my small experience, I would say that the former is more powerful. Why didn’t people do what you wanted them to do?
If you had to pick one or two questions that you always like to ask during these interviews, what are they? What do you like best?
Dana: It actually does depend on where we find the disconnect is. Let’s say we are trying to figure out if we should invest more in social media for someone. Then we would ask things like, what makes you want to share something on social media with your friends?
Sometimes, if we are trying to tell a level of sophistication of someone we’ll say do you know how many Twitter followers you have? If they don’t know they probably aren’t a Twitter power user, because people who spend a lot of time on Twitter always know how many followers they have, or approximately. Right.
Louis: How many Twitter followers do you have?
Dana: About 8,000. Laughs. That’s the kind of information that people pay attention to, right? Other things could be, what kind of things do you pay attention to now? So we’ll ask people, what kind of events do you go to?
If it’s a concert hall client we’re working with, what kind of music festivals do you go to in town? What are the radio stations you listen to? Also, why? What makes you want to go and see something?
Often people will be pretty forthcoming with their answers. I would say when you start doing this process–if it’s your first time doing it–the first couple of interviews are probably going to go awkwardly, and that’s okay.
Because you need to comfortable with talking on the phone. Not many people are comfortable talking on the phone. You or them. So that’s fine. But definitely don’t have a set of questions, just really let the conversation flow. Then take notes after each call, and figure out how you can make the next conversation go better based on what you learned in the previous conversation.
We will usually have a set of questions and then a decision tree. Okay, so if they said this, then we might want to ask about this. Often when we are doing the calls we’ll really just try to let the conversation flow naturally. Definitely try to keep it to 15 minutes or less. You really want to respect the person’s time, and of course, ask permission to record it, if it’s a place where you have to ask permission to record it.
It’s always really important. Even if you don’t use the recording, sometimes it’s nice when you’re thinking, “Oh, somebody said something and now I can’t remember what it was.” It’s always good to have those recordings to go back to it later on. It is also good for you to get used to the sound of your own voice because that’s something people are really uncomfortable with. At some point, you are going to want to hear yourself.
I think when you do these conversations, definitely make sure that you’re asking the right kind of questions that people… See how they respond. If you can do a face-to-face, that’s even better because then you can see their body language. You can see how they feel when they’re answering the question. On the phone is a little bit trickier unless you are using something like FaceTime.
Louis: That is an amazing answer. I would add to that. Instead of rushing to take notes at the end of the call, definitely record it, and get it transcribed. That’s one thing that we do is – we’ll have two people on calls at all times. Just in case I ask a question and my colleague will think of something else, which they could ask for more, so none of us worry about taking notes. We have the full transcript. Then it’s just, okay, you free up your mind to, “Okay, I only listen.”
And your other thing-
Dana: I would say one thing to that is try not to be on speakerphone. I find that people really hate speaker phone and conversations will go a lot shorter than they should–just because they may seem fine to you but probably can’t hear you at all. If you are doing the two-person method, try to have mics or something so it sounds a little bit better to them.
Louis: Yeah. What we used to do is using Skype or Zoom or anything that have three people in.
Louis: It’s way easier. Yes. I think like a journalist instead of a salesperson. One thing I found super easy, super good to do, from the start of an interview is to say, “Listen, I’m not trying to sell you anything whatsoever. I’m only here to understand you, what you do, like your problems.”
And people love to talk about that stuff. So it’s not as if they be like “No, I’m not sure.” No. They love this. “As long as you permit me to listen, I really want to know more about you.” It’s very rare that people will say no.
Thanks for sharing that. That was the first part of your answer. The second part is even more interesting because we are getting into the nitty-gritty, in specific tactics that you use, that I’m not sure many people use.
Rand Fishkin–who was on the podcast last year–mentioned something quite similar about influencer and understanding who influences who. In fact, he starting about a start-up like SparkToro about this exact problem. How to identify influencers, or rather people who influence you and people who you give a shit about.
How do you go about this? You started to mention going to FullContact. You actually export list of contacts, emails that you have, and try to understand who they are. Who is the person behind the email.
Dana: Yup. Absolutely. FullContact gives you that information. Actually, it’s really helpful that SparkToro may be a tool that we can use to do this because it is a manual process now. It kind of sucks. It takes a lot of time, and sometimes if we get an email list, and it’s got 20,000 people on it, we run it through FullContact.
Now, we’ve got 10,000 people with full social profiles. Nobody’s going to want to click on all those people. There isn’t really a tool where we can upload a whole bunch legitimately through actual proper channels, upload a bunch. We don’t do that stuff.
Yes, people listening, I’m sure you might have a tool that will let you do that. You should not be using it.
So that’s why I’m hoping a tool like SparkToro will help to close that gap. I have actually emailed Rand a little bit about, “Here’s what we do now. It’s terrible. Please help.”
I don’t know if this tool will fix it or not. That’d be great if it did, because it is a manual process, and some of it is where we go through the list, and say, “We know that we have these 100 users on Twitter. They are pretty prolific. What are they talking about? Who are they sharing? Is there someone in common?”
You can take people individually if you have some individual users that you decide represent your standard persona. Then you can put them in a tool such as BuzzSumo, and then you can see what sorts of sources they share regularly. Who they generally interact with.
Followerwonk is another good tool for this kind of information. That can be a good place to start as well. Really what you’re trying to get at this point is the kind of information that people share on social media that they won’t tell you on the call. Because I think a lot of people share stuff on social media and they don’t think about it.
For example: If you run my profile through BuzzSumo, I’m pretty sure most of my links shared are from The Verge, but I don’t talk about The Verge. It’s just something that I share their information from because I think it’s a good publication, right?
I’m always looking at because I’m coming up with ideas for my technology call that I do on the radio every Monday. It’s something I would never mention in a phone interview. If you said, “Where do you go for news?” I would forget about The Verge. I wouldn’t mention it. I know that.
But with BuzzSumo, if you look at my profile, The Verge is something like 30% of the links that I share. Okay. So clearly, humans are fallible. I think this is where the data really needs to back up those conversations as well.
Louis: And humans like to make themselves look good, right? They wouldn’t admit that they spend half of their day on Reddit. They like to be smart and say, “I look at all the marketing blogs, and this is all I read.” Yes. When you see what people actually do, instead of what they say they do, it’s usually the difference is quite messy.
An interesting point to make here, because every interview that I recalled with marketers and people like you is that, I try to think, okay, let’s say BuzzSumo and SparkToro, and all of that. They don’t exist in five years. Is this principle, is the thing we’re talking about here, still relevant in 5, 10 years, 20 years, 50 years? The answer is abso-fucking-lutely, right?
It’s not because social media is on right now, and everyone is on it and use it. In five years, ten years, you will still need to understand who you’re selling to, who influenced them, what they care about, what words they use, regardless of the channel or the format or whatever.
Louis: So I’m glad you mentioned a few tools, but I also want to make the point. It doesn’t really matter. The principle behind it is what matters.
Louis: Once you have those most profitable customers and you understand who influenced them, like really those people who matter the most to you, and those keys personas as you mentioned. What do you do with this information? How do you act on this?
Dana: We typically put together a list of interesting things you should know about your audience. That is: did you know they like these particular pages? This is something that people mentioned a lot. Just, sort of, facts that we didn’t know before we started doing this research and sharing it with the client.
Sometimes, at this point, they’ll say, “Oh, yeah. That makes a lot of sense.” Or “Oh, wow, I didn’t see that coming.” That’s really good information for us to know too, because then it’s like, “Well if you knew this already, why weren’t you doing that?” You can ask that kind of information.
At that point, that is the basis of building our personas. The personas, of course, people are like, “Oh, I don’t need personas.” Well, personas often are based on fake people that you just made up. Our personas are based on real, actual research that you do.
This is the basis of your persona. It’s based on real research. Not only is it based on real research, but that’s measurable, so we know that this persona is going to do these things. We can set up a segment in Google Analytics, or some other tracking method, that will be able to segment these personas into different buckets in your analytics.
You can understand if the behaviors that you think these personas are going to take are actually bearing out in reality of these personas are taking. If it isn’t, again, you’ve got a disconnect there. Is it because your research was faulty? Or because the flow on your website is currently faulty? Then that can help you identify disconnects.
We do have a blog post on our website as well that walks through our persona process. It’s called “Why I Use Personas.” I’m sure you’ll include the link in the show notes as well. There is also a video I did at Learn Inbound in Dublin in 2017 that walks through the process as well.
I recommend people watch that if they want to learn how to do personas. Teaching people how to do personas is another podcast all on it’s own, but it’s an important part of this process because those are the users in the user-centered design.
Then when you are thinking about the decisions that you’re going to make, you’re writing and making decisions for that one person. I heard once that BuzzFeed has two specific personas.
They have a woman whose recently out of university, has her first professional job, is one or two years out of school, is feeling a little disconnected from the social networks that she had when she was in university, and stays on BuzzFeed to keep up with what’s going on, and that second person is that woman’s mom.
That makes a lot of sense. So if you look at BuzzFeed, and in that light, the decisions that they make will make a ton of sense, right? I think that’s something where they have these really strong personas and they stick to them.
This can also help you with decisions when somebody comes in. This is my favorite. The VP goes off to a marketing conference and comes back just like, “Oh, this is really cool. We should totally do this.”
Well, do your personas want that? Does your brand voice bear out? Is this a good fit for your already existing strategy? Or are you just slapping something on because it seemed new and cool because somebody persuasive talked about it at a conference?
This can really give you some ammunition if you do have that kind of leadership who gets all fired up after they go to a conference and they want to try something new. That can help you be like, “All right, simmer down. This is why we can’t do that. Or this is why this isn’t a good fit for us.” Yeah.
Louis: I found that to be extremely effective. It’s funny, because as soon as you do persona research that is buyer persona research based on behavior not only demographics, which I like because that’s what you mentioned as well.
That’s what we mentioned in the podcast before with Adele Revella from the Buyer Persona Institute, which is exactly what they do. This concept of persona, not just demographic data, put together into this fake name and fake photo it’s what people do as well. What make them successful.
We’ve talked about that extensively in a few episodes so I’m glad we can move on to the next step. I think, I’m not sure we’ll have time for more than the next step. Let’s say we now have those valuable persona identified, and proceed with those two personas. What is the next step to weld this to user-centered design? What do you usually tend to do.
Dana: Now you have to turn to your analytic set-up and you have to figure out how you’re going to measure those personas. So are there certain pages that one persona visits versus another persona if the search terms that they’ll be doing?
Figure out how you can configure your analytic software. I’m assuming you’re using some probably some combination of analytics tag manager, maybe using Adobe. There’s a lot of different behavior targeting you can do, particularly when it comes to things like custom dimensions and custom metrics, that will help you capture those additional pieces.
If you have a really strong path, that makes you know that this person is this persona, and this person is this other persona, then put that in analytics as a custom dimension. Then you can segment by those different personas and see if they’re doing what you expect them to do.
Because at this point you’re not necessarily making changes unless there is some sort of horrible red flag that’s come up, and you know immediately if you do this one thing, things will get better. Easy win that way. Go ahead and do that.
But generally, now you’re implementing. You’re making sure that you can track, and you’re establishing a baseline for yourself because you need that baseline in order to know if the changes you’re making are actually going to help you.
Then that is the next step: Figuring out that baseline, knowing that this is where “I think they should be doing, but this is what they’re actually doing, and I know I can accurately measure this. Now when I make my changes, I can say, did this help, or did this hurt?” Without that baseline, you really don’t know if you’re effective in the changes that you’re making.
Louis: Outside of pages, I would call it use-case page. Let’s say your agency website. Let’s say you have two pages, one to says for marketers, and the other one for business or whatever. What you would do, is you assume that people visiting one of the two page, like the marketer page, is a marketer.
Someone related to that and vice versa for the business owner. What other type of behavior do you usually see to be used in understanding in who they are when they visit your website? What type of other stuff do you tend to use?
Dana: Something you could probably find out from your research is how they like to read content. Do they prefer video? Do they prefer downloading a pdf? Do they prefer reading on the web? That could be something you could measure. Are there more pdf downloads for this persona versus another?
We found this out, actually for a client of ours who is a convention center. One of their personas are wedding planners. We found that wedding planners actually liked to submit multiple forms at once, for all the different weddings that they’re planning.
They were screwing up our analytics because you can only have one goal per session, and they would be filling out the form six times. So great. We had to have a different method. And we only saw this when we looked at the unique events versus the number of goal completions and the number was off. I’m like okay, something’s going on here.
Then going back we actually did recordings of the site using FullStory, so you can actually watch what people are doing. HotJar is another tool that can do this. Then we could see them going back and filling out the form again and again. Okay, we need to have a different process if you are a wedding planner.
Then that tells us now we have this baseline of what they’re doing now. How can we improve that? We knew they were already interested. They would download a lot of PDFs, so one of the things that we weren’t tracking, is we were tracking when people were downloading PDFs, but we weren’t necessarily considering that to be a goal.
Let’s change that and make this pdf download for this specific set persona segment only a goal. You can do that through Google Tech Manager.
Another client of our, we realized that people were taking forever to buy tickets–this is for a concert hall–because they would check with their friends. “Hey, do you want to go see so-and-so? Okay. Great. Which date do you want to go? Okay, where do you want to sit? Okay. How many of us are going? Okay, there’s six of us. Okay, everybody send me the money, I’m going to buy the tickets.”
That’s four separate sessions that have all happened more than half an hour apart, so now you’ve got four sessions and one purchase. So it looks you have a 25% conversion rate. Sessions don’t convert, users do. Actually, it’s a 100% conversion rate for that user. We changed over their metrics from focusing on session based conversions to user based conversions.
Obviously, it’s not completely foolproof because if you are on a different device Google doesn’t necessarily recognize users between devices. That’s fine. They know the limitations but it’s really helped them figure out, “Okay, so when we look at a Facebook ad, this conversion rate looks really bad. Well, it’s because people are coming 18 times before they actually buy from Facebook.”
So what is it with Facebook? One issue that we diagnosed was a technical issue on their website, where people would log in with Facebook and would delete their cart. That’s why people from Facebook were having a much harder time converting than people who came from any other channel. It wasn’t something flawed with Facebook itself. It was actually something that was broken on their website.
Again, we only saw that once we start to dig into Facebook people, specifically, and say what is the problem here, look at that specific persona, and say why are they struggling?
This is where you really need to look at that data. You don’t really want to look at too much data at once. I would say pick a segment that you feel you can have the biggest benefit, and just start working on the specific segment. Then once you’re done with that, once you’ve got them to a point with them, then you can move on to the next one.
Especially if you’re a solo marketer; you really can’t tackle everything at once. It’s best too for your own headspace to be able to get deeply into one person at time before you move on to the next one, so you’re not trying to balance, “Oh, is this this persona or this persona?” You’ve just got the one to worry about.
Louis: I think it’s a strong case of FOMO most of the time. Marketers want to try to do everything at once, and they’ll boast, but then what about this, what about that? I 100% agree. Focus on your most important buyer persona. Focus on your most biggest, most important biggest problem.
Let’s try to solve this problem for them, and if you make this happen, if you make changes that actually solve it, then you’re going to have a much bigger impact than trying to wander around. I think we do have more time for the next step. I want to quickly speak more on this.
But I want to say this is quite amazing the ability that you have. The knowledge you have to translate real user behavior, like what people actually do into analytics tracking it in the right way. I think its quite rare to talk to someone who knows both. One to understand the user and one to understand the analytics. That’s why you’re speaking everywhere at conferences.
Dana: Laughs. Well, I’ve also been doing this a long time. I honestly think it’s because I’ve been doing this for 18 years. I think that makes a huge difference. I started using analytics back when it was urchin, so I’ve been doing this a long time.
I think a lot of people do ignore … I got to say, in all the analytics profiles I’ve ever gotten from every company I’ve ever worked with in my life, only one of them had custom metrics set up properly when I looked at it. No one else did.
If you’re sitting there struggling, think about well this data isn’t doing it for me. Well, is it set up properly? There is definitely more data that you can get. Don’t just limit yourself to the basics of what analytics gives you. Start to do stuff, like learning Google Tech Manager.
Oh, if you are in the Dublin area, or planning to come to Learn Inbound, I am doing a Google Tech Manager workshop the day before the conference. If you would like to learn all the dirty things that I can teach you about Google Tech Manager.
Louis: I actually live in Dublin as you might know. I will go Learn Inbound. Yes. Everyone listening– whether you want to go to Europe or not–I would definitely recommend people to go to Learn Inbound and check it out. It’s really good.
Let’s say we set up the right analytics, the right metrics, the right custom metrics. We are able to track at least one of our key persona, what they do. What is usually the next step? You started to talk about making changes. What do you usually see happening after that?
Dana: Hopefully, you see a positive change. You need to decide on a few key metrics that are really going to be your canary in a coal mine for this particular movement. Then again this is actually something I just talked about at MozCon is that we need to cut reports down to the bare basics of what we need.
Because I find that we often fire hose people with data, particularly senior leadership, in the hopes that they’ll find something valuable in there. When really, you could probably just boil it down to three key metrics. Decide what those three key metrics are going to be and really focus on that–and nothing else.
The other thing too that I really want people to start thinking about is the difference between reporting and monitoring. Reporting is what has happened and what insights can you bring from this. Monitoring is as it happens, actual tracking of what’s going on.
A good example is–every agency person is going to say, oh yes, this happened to me–is you get to the end of the month. You look at the report, and you realize that they lost 20% organic traffic that month.
There’s no reason for it. It’s not seasonal. You don’t know what’s going on, so you have to give this client this crappy report that says, “Hey, your organic traffic is down 20%. We don’t know why. We’ll get back to you.” Right? No one wants to do that. That’s a bad, bad time.
Whereas, if you had monitoring set up and you were actually pushing notifications to Slack or something that would say that organic traffic is going down. One report that I really like is in DeepCrawl. They will tell you the number of pages that have had organic visits.
If that starts to go down, that’s usually a good sign that something’s up. Then you can actually be proactive about it. So monitoring really lets you do stuff faster than the regular reporting cycle.
In general, a month is really a weird, artificial time period to choose for reporting. It’s completely arbitrary. The months aren’t even the same number of days. If you compare August to February, that’s bad obviously. Why do we report by months?
I prefer in Google Days Studio, by default they do 28 days. That’s nice. That’s a good time period. You can do six weeks, for example, but really do start to try to do little changes to show people better reporting metrics.
I think that helps a lot too. You say, these are the only three things we ever talk about when I give these other reports, so I am just going to cut it down to these three things. If they ask for other stuff, then you can talk about why they feel they need that information, because maybe you’re not presenting it in the best way and they feel insecure that you’re actually telling them what they need to know. What is it that you can do about that report?
A common theme in this conversation we’ve been having is a lot about egos and insecurity. I find that that’s a lot of what we end up dealing with. No one is having a good time. No one is making any money. Things seem bad. They really got their own personal egos wrapped up in it.
Sometimes you really just need to go in there and say, “You know what, just separate yourself from it.” Try to be zen about it. Get rid of your own ego, and think about what is going to help the user. I think that that is something whereas a marketer, you do really need to be able to take yourself out of this equation and think about what’s going to be best for other people.
Maybe that means you’re not the best solution for this. That’s okay. Maybe that means that you didn’t have the best ideas for this because you didn’t truly understand the audience. That’s okay too. Accepting that you have these flaws is an important part to making things better. Otherwise, you’re just going to try to keep hitting things with a hammer, because everything looks like a hammer.
Louis: I think that’s a great way to end this step-by-step process, Dana. Thanks so much for going through this in-depth explanations through each step. I think listeners got a lot of value out of it. A lot of people will tell me and say that this podcast is the only podcast where they actually had to take a notepad and take notes.
Thanks for doing that. I have a few more questions for you before you go. What do you think marketers should learn today that will help them in the ten years, 20 years, 50 years?
I would program out of a book I took out of the library called “How To Make Video Games Out Of Basic” Then I learned how to get around DOS, and I learned visual basics. I was doing a lot of Excel stuff, in Excel 1.0.
My first software job was at a Microsoft Access shop that made Microsoft Access software that did environmental reporting. My second job was for a Lotus Notes company that made Lotus Notes Customer Relationship Management software. Then I got into websites.
I’ve done a lot of different code in my life but I am not a great coder. That’s fine, but I can read and understand other people’s code, and know where it’s going. As marketer, sometimes you don’t have the programming resources for someone to explain it to you, or again if a programmer is feeling insecure about why you’re questioning decisions, they may not want to explain it to you.
You need to be able to understand that code. I think that is a really important skill … when I started coding was a long time ago. It was over 30 years ago. Code has changed a lot in that time, but I still know how to read code. I think that is a really valuable skill to have, even if you start with Basic. At least understand how programs do what they do, and why they do what they do, and then you will be able to read other people’s code, no matter what language it’s in.
Based on that answer, what are the top three resources you would recommend to our listeners?
That’s a very useful thing to have because I think a lot of people are scared of Regex. It’s awful. Whoever programmed, whoever made it originally was not thinking like a human. They were thinking like a robot. Regex is terrible. If you’re going to come up with a replacement, please do. I would love to see it.
In the meantime, we are stuck with Regex. Go in there and start doing data manipulations, and you really can’t mess up your original data when you’re creating custom fields. You can just go hot wild and make whatever you want, and try out some different stuff there. I really that as well. There isn’t really a super great Google Data Studio resource site yet. There probably should be but the forms are quite good, and I try to spend time there as well as answering questions.
There are lots of different places you can learn code. Code Academy is where I usually send people, to be honest. There’s a few different places out there. Pick your favorite.
Louis: So Code Academy, Google Data Studio. Is there any book or conference or anything else?
Oh, boy, well. Learn Inbound. Learn Inbound is an awesome conference. If you’re in Europe–and if you’re in North America, the plane’s not that expensive. Dublin is a very nice city. I do recommend going, even though it is a long flight. I love the conference. I think Marcus put together a fantastic line-up of speakers. He always does. It’s just a really well-run conference.
Of course, in the US, MozCon. Again, they really work hard on making a great line-up of speakers. If you are looking for something where you want multiple tracks, and you want to talk about a whole bunch of different things, SMX advanced specifically, in Seattle in June is another good one.
If you’re in more of the paid side of things, Britney works here in Dublin, and does a lot of our pay work, went to Hero Conf this past year, which was in Austin. I believe it’s going to be in Philadelphia this upcoming year. She, two thumbs up, loved it. I recommend checking that out as well, if you are on the paid side of things.
Louis: Dana, thanks very much for your time today, and for answering all my questions, and for being understanding of my constant interruption. Where can listeners connect with you and learn more from you?
DanaL Probably Twitter is the best place to go. So it’s @danaditomaso, which is super easy to spell so check the show notes. Of course, you can always Google me. There’s only two Dana DiTomaso’s so Google usually figures it out no matter how badly you spell it. The company is Kick Point, and definitely check us out.
We have a weekly newsletter that I do really love, and I think you should probably sign up for. Every Friday, we send out free digital marketing design articles that cover interests this week, as well as music pick, as well as something cute and adorable we found on the internet. It’s really short, and it’s a nice way to keep up if you can’t keep up on Twitter all the time.
Louis: That sounds like the perfect newsletter to me. Thanks, Dana, once again.
Dana: Thank you for having me.
I’m a no-fluff marketer living in Dublin, Ireland (but yeah, I’m French).
I believe you can treat people the way you’d like to be treated and still generate results without using sleazy, aggressive, hack-y marketing. This is why I’ve started Everyone Hates Marketers – a no-fluff, actionable marketing podcast – as a side project in April 2017.
I’m also the Content Lead at Hotjar – a powerful way to analyse people’s behaviour on your website or app and understand how you can improve their experience.