Why is it so easy to lose the customer in the heat of product development? We tend to forget for whom we do the project? In DevOps, we say that we need to focus on user needs by amplifying and shortening the feedback loops. But how is it done? What should be done to entice customers to be part of the development process?

Maria (00:05):

Organizations themselves have already tons of data that unfortunately is often unused. It's so easy to start from the organization data, then start gathering from outside. And it's good to have both qualitative and quantitative data. Methodology isn't the one that dictates how good a persona you do.

Be sure that when you are doing a persona, don't just take any template or tool that is available. Try to first understand what you want to know and then add up.

Lauri (00:44):

Hello and welcome to DevOps Sauna. Co-founder of Unbound, a Canadian software company once said: the customer isn't always right, but if you don't listen to them, your product won't be either. When talking about DevOps, people iterate that we need to focus on user needs by amplifying and shortening the feedback loops.

But why it is so easy to lose the customer in the heat of product development. That's what we are talking today with Maria Wan, Jarmo Parkkinen, and Marko Lindgren. The participants ask what should be done to entice customers to be part of the development process. Join me in the conversation.

We are here in the DevOps Sauna podcast. Thanks very much, Maria, Marko, and Jarmo more for taking your time. 

Maria (01:39):

Yeah, thanks. Thanks for inviting. 

Marko (01:40):

Thank you. Great to be here.

Jarmo (01:43):

Thank you. Nice to be here. 

Lauri (01:44):

And we have a very cryptic topic for today, which is lost customer. Why we don't know our customer? And as I was preparing myself for this, I knew that there was a quote that relates to this and I had to go back and look it up.

Sam Walton was a founder of Walmart and Sam's Club, somewhere early 20th century. He has been quoted as follows. There is only one boss, the customer. And he can fire everybody in the company from the chairman on down simply by spending his money somewhere else. And that quote has lived on now for a century.

And we all know it. Like if we do a lousy job, it's the customers who eventually gets to judge the quality and the output and the value of our work. Yet, I hear from you that we lose our sight of our customers. And we know that we have design team, we have UX team, and we have a ways of constantly reminding our teams and customers about enticing customers or their customers to take them into the development work.

But the question is why is it so that even though there's a century-old expression that everybody knows, and everybody agrees. The customer is still lost. And more importantly, what can we do about it? Who wants to start? 

Maria (03:04):

I do. It's my statement. And it is, yes, it is a very bold statement. Unfortunately, so it is at least somewhat true in so many organizations, but after all, it boils down to the fact that how the customer is defined. When we talk about customer, we actually mean market, segment, and various kind of personas. But if it's just customer as is, it’s actually just talking to whomever, it's just some random person or living being instead of better defined. That's my taking into this lost customer bit.

Marko (03:46):

And also for me, coming from user interface and user experience design in digital services, customers often thought that they can be anybody even when the product or service being sold has quite specific solution to offer. And it's not always easy to recognize who is that customer who will benefit mostly from the solution.

And then again, I may be starting to talk only about users who are kind of already taking a customership and not looking at the new customers. 

Jarmo (04:21):

Yeah. I agree with both of you. And then I think, especially nowadays, since we have been talking about the software, and as service business, then we have customers who are not paying customers. They are like using this free part of it. And then we have customers who are paying something for it. So are those customers different? 

That's also, I think an interesting viewpoint. And do we know what makes the customer pay for the product that they using anyway? 

Lauri (04:48):

Yeah. So B2B market there, I have to put it out there, to begin with. I have always been extremely distinct between user and the customer. In the business-to-consumer segment, it's like when I'm going on buying a new stereo or a new watch or new car, I am the customer and I am the user. It is me to blame if I make a wrong decision. And I will improve myself to make a better decision next time.

But in the B2B, there is a difference between those two. And I always think them very distinctly. Are we talking about B2C here? Are we talking about B2B here or is it all the same? 

Marko (05:30):

Both. I would say that my background is mainly in B2B side. And there's a transformation happening, some blame millennials but they get blamed for everything. So I don't think it's their fault. Instead, it’s happening that people are using these well-done B2C products and they are accepting the same or even better service in B2C side. 

And also decisions are off per test get in a way personal in many organizations. So, there again comes the little bit hazy line between our customer and user. And yes, I also always, I want to ask are we talking now of the customer or the user, but they are getting more and more mixed. 

Jarmo (06:23):

Yeah. And then when we talking about knowing the customer, I don't think it's so essential to separate, this is a corporate customer or individual consumer customer because knowing the customer actually defines that then.

Maria (06:36):

And I would even bring the customer in a wider perspective. And this is for, let's say a software development team thinking about the customer. It might be the influencer, which is the most important customer for that person. It might be somebody inside the organization and for those who are in product management, they talk about markets, product, market fit.

And therefore for me, the pseudonym for customer, in that case, is actually market. Because if we just talk about customer, you're afraid that you might be doing that product development thing only for us, one B2B customer. So I would like to totally dismantle the word customer to define so many different things.

And as long as you don't know these differences, how to recognize and make differences, what the customer good to me for you as a person and for an organization, I say you're lost. You don't know who you're talking to. You might be just developing for yourself and that's the biggest scare. 

Marko (07:43):

Yeah, we fight if I can add that to that a bit, so that's like you can find out, you can make user research. You can find out what the customer wants, but you can't possibly really find out exactly what market wants. It's more like vague and more like a fluffy thing. 

Jarmo (07:59):

Sometimes hear being presented with product-market fit. And I know Maria that you didn't use it in that specific intention, was that when somebody says product-market fit, there is this underlying assumption that there's a market somewhere out there. And we can take a pile of it and all we have to do is fit with it. 

But then there's this alternative point of view, which is there are customer needs and could be, there is no market for that yet. And as soon as we discover what those needs are, we can then do almost like a product need fit. And then that market will come to exist, which is more like a category design view that many marketers have advocated in.

Like, don't try to fit yourself into the market. And again, Maria, I know you didn't mean it. What I take away and took away from you, Maria was demolishing something which is well established. It's always a bit scary. Because we have some established conventions, how do we research the market? How do we construe the definition of customers?

How do we define the market size and all of that? So I'd like you to expand on your suggestion that like, let's, you didn't use those words, but like let's redefine. Let's start over this market definition. So talk us through on that idea. 

Maria (09:21):

Yeah, thanks. Well, here again, what is a market? Let us not go into that. But exactly for me, it is so important for organizations to realize that it's just really like whom you talking to, whom you're trying to influence, whose needs are you following? But it's also like you said, that could be something new. Usually, it's so common that the user or customers, we don't yet know what we want.

But we want it, but we don't know yet what we want. So a good organization, which is a bit ahead of time, can also look into that. So for me, I greatly enjoy when I hear companies discussing what the customer is. It's so easy to ask first. The company's like, so who is your customer? And they start thinking about it among themselves.

And it takes them quite a few hours to realize who their customers are. Although in the beginning, they said that we know our customers. I'm not sure if this answered your questions, but I got just driven away. 

Marko (10:31):

I really liked that part of having set methods to do research because, to me, the only way to understand the customer and customership is to go where the customers or users, or both exist and observe what they are doing. 

But for example, I don't know of your case when you are doing shopping, but quite often you don't buy a B2C thing only for yourself. If it's a TV, it’s used by everybody in the household. And even if you like a TV set that has some pictures, there may be another family member who at the end of the day makes the decision, which TV suits to that interior, et cetera.

So the difference of customer and a user yet again varied. And if we don't observe what people actually do when they are making a decision and when they are using the product, even when they are discontinuing the use. Then we really don't know what customers or users are. 

Jarmo (11:45): Yeah, I agree. And especially if it's not what I start working, if you're talking about the B2B business, the business center working environment has changed. So that actually are those who used to be users who had not seen what was bought or what what they were using, nowadays can bring your own device, systems that they can buy whatever computer they want. 

And then they just, like log into cloud systems, and actually, they are now customers and user at the same time, even though we are talking about the B2B context. 

Maria (12:12):

I like the fact that you brought in kind of influencers in buying or those who are like, the stakeholders when buying a TV for example. But it's the same thing in company decisions, especially, if you go abroad, in Asia, it might be an Indonesian grandmother actually doing the final decisions. But in a company, you may not know it because you're dealing with the executives in that company. But somehow it’s the grandma who makes the decisions. And if you don't know that your game is lost. 

Marko (12:44):

And in the digital market talent services business, it's much more harder to say that when and by whom does the decision really happens. With the TV set that you can point that, alright. That dude bought that TV. But for example, when they are trial periods in big companies or different kinds of analyses and tests and tests trying things out are made, the actual end-user may actually be the customer who says that I don't know how to do my work with this thing, but I know that this other thing works because I saw my colleague using, all of them were using it in my previous working place. So well, collect data. 

Jarmo (13:30):

Yeah. I think not knowing the user and the user research, even though it's tedious and takes a lot of resources and effort, so that's the best way to understand what they're really doing and wanting. 

Marko (13:42):

I'd like to break down this inside the gathering or the entire process of actually getting to the bottom of what those customers’ needs are. And how to get them into account when we are delivering and developing services. So should we try to put a little more substance behind, like what should be done in order for organizations to have this customer on board and to entice the customer to be part of the development.

Maria (I14:14):

Sure. Well, why don't I start with a very concrete example? I had a team that I was consulting and a development team, and we decided to do personas for them. And one of the personas that they chose, let's say that his name was Bob. They actually made a cardboard person out of it into their team wall room so that they were able to see Bob all the time. And it has also, like the description of what kind of a person Bob was or what his needs and desires and obstacles, etc. were.

So it was a good thing to bring the persona, Bob, to the room. However, it was there for such a long time that eventually it became sort of a wallpaper because nothing happened to it. Also, many companies have their personas on the wall in papers saying, but if they're constantly there on the wall, do something different for them?

Get their hair done. I mean, change the shirt or try to get the person a new hobby as long as it fits the persona, just so that they are somehow alive. But of course, everything comes from recognizing, gaining insight, and then putting it into practice, and make it a friend, use it. 

Marko (15:41):

Yeah. Bring it through your coffee pause and have a chat with the persona there. Yeah, I think it's also that not to forget it on the wall, is also to take that persona and the persona changes as essential part of the whole development project. So that they say, it's checked up and said, hi, how are you doing? The person on frequent intervals that we're not forgetting it. 

Jarmo (16:02):

To me, one really important thing with a personas or any other means to come to understanding of users and customers is that unfortunately, what Marko said is cruel, it is sometimes a little bit costly to have research. But I would like to say that make your personas based on real data and real observations rather than creating imaginary things.

So it's much more better to have one persona that reflects real understanding than 10 personas that were invented in the said coffee room. Maybe this question be then left to the coffee room and make the one guy who woke up will present the knowledge and understanding of user or customer or both into the right wallroom. 

Marko (16:54):

Yeah. If I may add to the stereotyping thing is also now, especially if I'm working a lot of with digital accessibility, and though it's very important to take into account that also this comes from a different impairments or disabilities that people have so that we have personas who will represent the whole, like with and multi diversity of the clientele or customers base.

And so remember that there's a lot of people who have this kind of a temporary situation or a permanent disability, and need to be able to use the service or software anyway. 

Maria (17:29):

We're saying that we have, there are so many personas in a company, and I guarantee you, once you start digging the data that you have, make it into insights that are. And then create the personas from different perspectives. And not all personas are valid all the time. You don't serve the 25 personas that you may have, hope you don't have that many, but choose the ones which are closest to you and which are important at that point. But I like the fact that the sort of the customer is sort of dismantled into smaller personas. 

Marko (18:06):

I personally hate working with personas when they are not based on real understanding and real data. So, what many times I have heard the question that how many persona should we then have? And I always say that one is definitely better than zero, but then the right number depends on the types of customers and users you are able to concentrate on the given moment. 

So as Maria said, update them, use them, let them be in the resting position for a while, and then take them back again when it's time to concentrate on that side of business again. 

Lauri (18:50):

Maybe some of that answer lies in the maturity of the company.

Marko (18:55):

Definitely, yes. 

Lauri (18:56):

And also the complexity of the buyer journey. So if it's a very nascent new company and they're just trying to figure out what's best for them and what's best for their customers, then maybe fewer is better. And then when the company gets mature and can address more distinct customer segments, they can afford to do more research. Then could be that, and for the fact that they can address more segments and more types of users and buyers, then maybe that's also one way of looking at the right number.

Lauri (19:35):

It's Lauri, again. When producing user insights and making them actionable, some methods are useful when you're planning a new product while others help improve a product with an existing userbase. Our folks at Eficode wrote a guide about creating product user insights. The guide introduces various techniques such as user personas, wireframes and prototypes, and user story mapping, just to name a few. 

You can find a link to the guide in the show notes. Now let's jump back to the conversation.

Lauri (20:10):

I was actually thinking about this from a statistics perspective. So in statistics, often time we use K-means cluster, which is basically… here is the sample. 

Can you split this to me to five different cohorts? And you tell them how many cohorts you want. And then k-means clustering algorithm basically tries to find the sets, which have the maximum consistency within the set.

And it's maximally distinct from others sets. And if you think persona development and let's start from this assumption that you are doing it the right way, which is, of course, there's many ways to do it, but one right way to do it is go get to the bottom of the real people who know about these things, who have opinions about them, have deep interviews, have like quality data gathering.

And when you discover some insights, then apply quantitative research to that, to really get the magnitude of that observation. So, when you do something like that, you will inevitably get to a situation where you have a relatively broad quantitative dataset that represents your personas overall. And then you do the k-means clustering, either statistically or just intuitively.

I sort of presented you one way of developing personas, and I'd like to hear your thoughts on that, and also this sort of statistics to approach for that. 

Marko (21:38):

I like combining the hard and soft sciences or having both qualitative and quantitative data. And I'd like to say that what you described this kind of what everybody would love to do. But then there are problems, like people are quite complex and getting really to understand the whole customer base qualitative studies takes lots of time. And during that time, the customer base and the user base already changed because people are also nasty on that.

That's how they learn new things. They forget unnecessary things. So having the right amount of each type of data is the big question. And I think it costs to the point of maturity. So when you were starting this work, start from small. And then roll on that knowledge you have caned instead of trying to kind of turn to the end right from the starting point.

Lauri (22:43):

Yeah. And I like the statistical approach, but then the challenge there is I guess, that how to make that persona produced by statistics and like relatable and an actual person. And that's what Jarmo said about these changes that, yeah. So, people have life changes. So personas also should be valid through these changes in life like getting a job or getting children or getting rid of children or what not. 

And that way the customers go through all these faces and now persona also be still valid when they go through those phases. 

Maria (23:15): Yeah. I add to whatever was said before is that it's so important to have different kinds of research methods to get that data. Organizations themselves have already tons of data that unfortunately is often unused. So it's so easy to start from the organization data, then start gathering from outside and it's good to have both qualitative and quantitative data. So that you get as much as possible. So methodology isn't the one that dictates how good a persona you do. 

It's more likely that be sure that when you are doing a persona, don't just take any template or tool that is available. Try to first understand what you want to know and then add up. 

Jarmo (24:06):

Yeah. I subscribe to start from small at-up approach. 

Lauri (21:11):

Yeah. From behavioral economics, it is well-established that there are a lot of these, what are called biases. Effectively, we are not aware of the ways, how we make decisions. We are not aware of our needs. And so, these can be abstract sort of definitions of what biases are. But I think a very concrete example of what bias could be is when we respond to survey, we read a question and we don't understand it. So we conjure a question in our mind that we do understand that is close enough the question which is presented to us.

And then we replace that question, which was presented to us with a question we came up with ourselves. And then we answered that, that question we came up ourselves with the answers or options given to us. Then we ask, okay, so we do this a thousand times, and then we come back to the results. And I think if we were to be careful researchers, we should ask ourselves, have our responded answer to the question we ask or have they answered to completely some other question. And what impact does this bias have, you know, a result. 

Maria (25:19):

I love that you brought that up. I totally love it. As researchers, it's our scene, but it's also the one who's answering. It's also his or her scene, is how we understand what is wanted from us or what is asked. But it's also what we want to, what kind of an image we want to portray of ourselves?

Is it just kind of wishful thinking? Am I answering something that I would like to be, or am I actually answering honestly? So, and this is where the methodology or the fact that you have so many different kinds of methods is so important. This ensures that the biases are lost when you have enough of data.

Marko (26:05):

Yeah. I'm like, your optimism when you say people try to answer a question, which is close to the question they were asked. Not always people lie. They totally misunderstand the question. They totally have different vocabulary than the person who is asking the question. So again, what needs to be iterative and I like it when I can go and observe what people actually do. And ask them to tell me why they did that and how will this and that thing affect to the next thing they are going to do or will it affect? So, observing what people do.

Maria (26:50): 

Are you as a researcher asking the right kind of a question, is it honest or are you already biased? 

Marko (26:56):

I don't make mistakes that I admit. 

Jarmo (27:02):

Yeah, I think. Yeah, I agree with that. Then you can ask a question and the person answers the question, and then if you ask more like, a little bit more about it. Then the person might change that question. Like, so for example, when I did research and asked if people would throw out the use of mobile apps or website, and then there's the apps that the people don't use mobile app. But then after a small discussion, they said that actually they would rather use for important things, they would rather use a computer and a website. So then what is important to them? They use a big, like the casual things, they use mobile for mobile app rather. So that again is actually the answer changed totally when you ask a second question. 

Lauri (27:40):

Great example is this everything I have failed in, I have answered truthfully. And you usually get that question at the end. Now my question to you is when has it happened to you that you have gone back to the beginning of survey and reviewed everything again, and basically filled in the survey one more time because you wanted to answer to that question truthfully. There's a result from a research that when you present that question in the beginning of survey, then people actually respond more truthfully than when you present it at the end of the survey. 

Because what other options do they have? They have already responded to the survey. The only thing other than ditching the whole survey starting over is to say yes. So first ask them for commitment that they are going to answer truthfully, and then present the questions and you are getting more truthful answers. So there's this meta-level of questions as well. 

Jarmo (28:34):

Yeah. If I may add to that one, I have also one user research case, was like that. And I think the best answer was that the user said that, well, let's say that this means that I answered truthfully to the best of my knowledge. So then I can put a tap on it. So exactly the user answered the different question that was asked. 

Maria (28:53):

Yeah. And I sometimes, depending on questionnaire, I sometimes ask at the end instead of did you answer these truthfully, I asked, like, how are you feeling when you answered? Because then if you've had a bad day, the answers may be a bit negative, but when you've had a good day feeling, then the answers tend to be positive.

So that's how I've been trying to evaluate the answers. And by the way, is it technically possible to go to previous pages on the survey? 

Lauri (29:23):


Marko (29:24):

Usually not.

Lauri (29:27): 

The other, these are of course tidbits, but it's okay. We can, we can throw out some tidbits to our listeners as well. The other interesting tidbit is the background variables.

Often it happens that we asked the background variables first, and then we go to the questionnaire because we feel that if the responses are incomplete. Then at least we can analyze incomplete responses with respect to the background variables. But thinking again, when you ask background variables, you are reminding your respondents of the kind of categories that they represent.

And then they are assuming those sort of category identities, and they are going to respond to those questions or their category identities that they assume like social economic status or their age or their gender or their income, or any of those. They are going to influence the responses that they gave. Somebody could suggest that if you want more truthful answers, ask the background variables last. 

Maria (30:26):

That's true. But to add to that, I would say that those demographic questions are last season. Now at the time of empathy, we don't anymore want to know if you are a man or a woman or something in between. Where are you from?

And what's your age? It's more likely the lifestyle of living and your brands and your identity that we are more interested in. But of course, depending what you want to find out and not the tidbits to go again. I agree, but.

Lauri (31:00):

Yeah. Yeah. So I was thinking if you respond that you are a huge fan of Gucci and BMW, will they influence your responses even though they are not socioeconomic statuses. 

Jarmo (31:12):

Always a measuring act will have an effect on the thing that is being measured. So I, like what Maria said that I would phrase it a little bit differently. Like, ask only questions that are relevant to the subject you want to understand. So if it is really important to understand that is the person giving the answers in higher income bracket, then yeah, go ahead. And ask it. But if it is not relevant, then don't ask it. 

Maria (31:51):

Yeah, it's not a consolation, but I mean, compiling the questions and what to ask in a survey. So it's a difficult task and it takes time. So it should be given that time. 

Lauri (32:02):

Yeah. Maybe we can start rounding up. I think we have one important question to the listeners. It is what can they take away from all of this? 

Marko (32:15):

Maybe I’ll go first with my repeated message because I love qualitative and quantitative data. Everybody else should too. So gather data, but whatever you do, get some data and then form an idea or expectation or even maybe persona if you are so prayed. So observe and analyze the information, what method is most easily available for you.

And only after that, start trying to understand what you can do next. So build your understanding, build your organization's understanding based on data and wisdom you have gained so far. 

Jarmo (33:03):

Yeah. And then use the data to really understand the customer and not make like assumptions or ideas that what they are, but really, like make real people, real users, real customers they are.

And remember that there are permutations. There are customers go through different life cycles and they have their different disabilities and different diversity in the customer base. And then support the design process so that you really have people there. And they are known to the development and the people who are actually doing the product so that they understand and know the personas, or however you are creating this character.

Maria (33:43):

Okay. Maybe my final guidance for you, dear listener is how to go about is, number one, start by creating a cross-organizational workshop where you define your market for roughly you know it, but anyway, your market, segments, and personas. Then number two, gather data, a lot of it and some you already have. And third, see what you are missing in the data and fill it in.

And then number four, collect from the data, the insights that are valuable, which you can work on forward. 

Lauri (34:26):

So stop assuming, start small, gather data, and refine them to insights. And then just do it again and maybe break things in the meanwhile. This has been a great conversation. Thanks a lot, Maria, Jarmo, and Marko for joining. 

Marko (34:41):

Thank you. 

Jarmo (34:42):

Thanks. It's been great. 

Maria (34:43):

Thank you.

Lauri (34:44):

Thank you for listening. To continue the conversation with Maria, Jarmo and Marko, head out to the show notes to find their social media profiles. Alongside the profiles, you will also find the links to the content referred in the episode.

If you haven't already, please subscribe to our podcast and give us a rating on our platform. It means the world to us. Also, check out our other episodes for interesting and exciting talks. Finally, before we sign off, let's give the participants floor to introduce themselves. I say, now take care of yourselves and remember to apply both qualitative and quantitative information to your customer insights.

Jarmo (35:21):

I'm Jarmo Parkkinen. I have in working with usability in user testing or usability testing and with other methods to understand the usability of different products. Lately, I have been working at a UX and interaction designer, mainly in B2B setting, some B2C things here there. And I usually love it when I kind of work with complex systems where it's really difficult and hard to understand what is happening. That's when I'm at my best. 

Marko (35:59):

I am Marko Lindgren. I'm a user experience and accessibility specialist. I've been working with digital services and software for like over 20 years, always with the user perspective and trying to remember that the user is the most important thing that we are working for.

Maria (36:18):

Hello, you. My name is Maria Wan. I'm a service design consultant. My talent and passion is to help companies to bring the customer, or however defined, to the center site or wherever the organization strategy defines. I love making a difference in customer experience and customer-centricity within organizations.