S1 Ep. 21– Finding gold in employee text feedback

In this episode, you'll meet Andrew Marritt. CEO of Organization View based in Switzerland. We discuss the recent increase in companies capturing employee feedback powered by developments in technology. Importantly, the huge opportunities available to unearth gold hidden inside the vast amount of words shared in those surveys. Going beyond eNPS or 5-star ratings.

Listen to full episode :

Want to connect with Andrew?

https://www.linkedin.com/in/andrewmarritt/

Websites

https://www.organizationview.com

Experience Designers Ep 21

(I use a mix of automated transcript software and editing for readability )

Steve: [00:00:00] So Andrew, welcome to the experience designers one.

Andrew: [00:00:03] Thank you very much for the invitation.

Steve: [00:00:06] Yeah, welcome. Welcome. So as, as always with my guests, let's just, let's jump straight in. Let's a brief introduction to yourself, or a little bit about your background and your passions and yeah.

 Where you are in this planet and what you're doing. Okay.

Andrew: [00:00:19] So I'm Andrew Marriott. I am the founder and chief executive of organization view, which is one of. The oldest people, analytics practices. We were founded in 2010. My background was originally, I was a childhood geek. My, my father built a computer from a kit in the seventies, and I learned to program late seventies when I was about seven.

Geekness, wasn't very cool in the eighties. So instead of going off and reading computer science, which would report in useful now I went off and did maths and economics. Spent most of the nineties in consultancy. And then most of the following decade from 99 to 2009 in big corporates, mostly financial services sector sitting in the intersection between HR departments and the marketing and branding teams, doing things such as employment, marketing, brand research, those type of activities.

Then in 2009, sitting on the HR management team of a big European, industrial, and watching how badly everyone was doing data. In, in the HR department, I have to say it wasn't on to that last company. I decided to found a business to, to focus on it. So we've been going for over 10 years now. The, the key areas have started with around this bringing the sort of analytical techniques that we used in marketing, into interpret, looking at people.

And in the last five years we focused almost exclusively on. Text-based feedback and applying what you know, with a lot of your audience would associate with AI to very, very large amounts of feedback to give executives and understanding of what people are talking about within their organizations, in terms of passions.

I'm very, very fortunate to live up at Almost 2000 meters in summer X in Switzerland. One of the most beautiful places in the world we've been remote since 2009. As well as the, the normal things that people do up here. You know, I also keen mouth and swimming in the summer and in the winter King crest rider.

So go down there, go down an ice, cannot on my front five centimeters away from, from the ice.

Steve: [00:02:31] Amazing, good lifestyle. It good lifestyle. You know, I think what was what I'm super interested in exploring today you know, just hearing your background and I think some of the pieces there around that intersection between kind of HR and marketing, that's super interesting.

Still kind of still being. Kind of discussed and debated to a certain degree today as companies are still in some respects, figuring that out for some reason. But also kind of HR function. I think certainly as a, as a function itself, the analytics requirements has been steadily increasing year on year for, for some time now.

Probably driven by technology of course, but also the needs for listening and understanding and getting a much kind of Much clearer understanding about, you know, what the well, the most captive audience that you have as a business, which is short, which is your people. And then also the final one is I think for me, like text-based feedback, you know, it's, it's where the gold is.

And ultimately if you can mine that gold effectively, then there's this you know, untold possibilities and benefits for the organization. So I'm really keen, you know, and it's still definitely something which is Very very hot right now. And so keen to delve a little bit more into some we know service, some of the evolutions and the developments in this area with you as well.

Just just going back just to delve a little bit in, cause it was just something I picked up on your profile, actually, if I may which was some of your experience at UBS. So some of the work you did there just tying into some of that intersection piece, the, the new joiner experience you know, kind of.

Was that kind of say back then, I don't want to share your age here, but I mean, is it. Back then was that kind of seen as kind of a bit cutting edge, a bit kind of a kind of advanced back then, do you know thinking about new joiner experience, how people experiences as onboarding?

Andrew: [00:04:14] Yeah, I think so. At the time it didn't feel like advanced, so ups.

I mean, I've been doing marketing for, for several years before, so it wasn't, it wasn't the first area. And a lot of the type of experience work that we'd done. And, but ups was an, and is predominantly a wealth management. So private banking organizations is the world's largest and. A lot of the way that you sell management is around experience, right?

So they have a bit private banks or the wealth management businesses have big experienced design teams. They, they look at every single aspect, whether it's, you know, they have architects, they look at the smell going into the business the buildings, they, they focus on a really, really crazy amount.

So the organization was sort of set up and tuned up. For that type of work and, and, you know, the senior executives were used to it because it wasn't new for them. There was, you know, there was some data that I've seen from what, from actually one of the other banks where they'd done some research and found out that 70% of purchase decisions within banking were experience-based and we did some data, you know, I caught some data from.

I think it was a cop relationship comes one of those types of organizations who had done some research on what people wanted in an an employee value proposition. We found about 70% of them were experienced type related. So with that type of data on hand, it became quite obvious to do it. We all, I also, as I said, was sitting at least half the time in, in brand and research and, and, and all the folks I would be dealing with.

So, you know, Setting back. I was with some fantastic experience design people and, you know, it became obvious in some ways to try and apply that into, into HR. And we did some early work also in the early days of Of organization for you in 2010 on experience design, but it was mainly financial services.

It was maybe private banks and the rest of the rest of the organizations sort of nodded, but had no interest in. Inexperience in HR. I think that probably changed in about 2016. We started coming in, by which time, obviously I'd focused on other things, you know, thinking that HR was never going to get this experience and moved on.

So yeah. Yeah, it was, it was good, but it was very much coming from. What was then the sort of cutting? Yeah, it's really systematic approach to experience and experience design. And a lot of it was numbers based. A lot of it was research focused. It was very systematic. And you know, listening to people and looking at.

Well, for me, there's two parts from experience data. There's the listing part of why people are doing it. And often the more qualitative parts where you may be viewing people doing it, or using sort of more qualitative approach, but also combining with that quantitative actually sort of saying, well, actually it's.

Here's what they're telling us, what is actually what we're seeing in the data, right. And web and digital tools and they leave to track everything. So you can start building, you know, quite detailed pictures, you know, from an let's take a recruitment perspective. You can see where people drop off.

You can run. So that, that part then, and I remember at the time it was really hard to get the applicant tracking tools to provide us with drop-off data at different parts of the process. So we can optimize that. So we just ran usability studies and we recruited people into. You know, who is who our target markets, we fit them, watching it, using it and trying to get them started.

And that type of thing became really, really valuable. So it's always been, my approach to experience has always been very data centric, but I don't. Define data as just a numbers, right? It's this mixture of qualitative data and you have to balance the two together and that's where you really get the,

Steve: [00:08:15] yeah, I agree.

I agree. And from a, just from a listening strategy perspective, and specifically let's talk from an edX based CX how. What have you seen that evolution? How have you seen it evolve over perhaps the last 10 years or five years or whatever? The let's not put a time on it, but just how have you seen it evolve generally?

Or what are you seeing?

Andrew: [00:08:35] I think the big difference, certainly from what we can do now, compared to what we could do back then was the use of. Non-structured data or rather not conventionally structured data. So for example, UBS, we launched a. Certainly that went to all of the 16,000 people a year that we hired at the end of a hundred days.

And it was a traditional survey. It had 50 60 questions, and I'll also, those have parallels with engagement surveys so that we can track it across. Now, obviously that in hindsight, There's difficulties in there because everybody's still on the honeymoon period. So your level of engagement is 90 something percent and you're not, and there's no variance and therefore you're not learning much these days.

You know, the way that we work with our clients when we're doing experience type stuff is they're asking shorter survey questions. So they're shorter surveys and they're they're relying a lot more on the qualitative parts of the text. And that's, that's really technology. Yeah. Enable you to do that because you know, even at UBS with 16,000 people joining it.

Yeah. Looking at all that data and having a process to go through and identify trends is hard. And even if you were able to get a team to Revere it, they wouldn't spot the patterns. Right? So we, we tend to the way that we remember comments and things, we remember powerfully written ones. We remember ones that support our own worldview.

We do a lot of those type of things and bring our own human biases and. The algorithms can give, you can take a step back and just say, you know, this is an anomaly detection. We've detected this as a big issue in this particular area. And that much really unusual here's the responses here and you know, how do you interpret that?

And then that's, that's much more interesting. So I think we're seeing, seeing more of that type of type of work But, but you know, a lot of ICL lost HR departments doing making mistakes. Maybe that the customer experience people were doing 15 years ago. Right. So not relying on data. There's there seems to be too many people where they get a work working group of people.

Who've got sort of skin in the game, you know, heads of recruitment to design what the ideal. Process or, or what experience would look like without actually listening to, to the people who were there, you know, going through it. And you really just have to have to listen and you almost have to throw your own ideas out of the window and rely on the, on the, on that type of data, rather than trying to design something perfect yourself.

Steve: [00:11:14] Absolutely. It's kind of empathic data, isn't it? You kind of, you put, you have to put yourself to one side for a moment and just start to implement even on a first phase of listening, just to kind of keep it as, as. As non influenced on your own bias or asking the questions that you want to hear the answer for too.

You know, I know when, when when, you know, when we've done some candidate experience listening work it's, it's fascinating to see, and I'm not, I'm not a research expert. I always absolutely bring in research experts because I see it, it is an art form and it's a, it's a very specialist discipline in its own.

Right. And rightly so But it's just, it's interesting how the different stakeholders suddenly become experts in writing survey questions or you know, I know I've had some sentiment and there've been all pretty much closed questions and it's just like, yeah. So I think that's some of the individual that is that kind of stuff that you have to battle, but we're in a

Andrew: [00:12:06] tourist, we're in a tourist area and I think that's.

You know, so obviously hotels, some people are interesting and lots of my friends, so you know, running big hotels and that actually the, you know, some of them, some which is famous for five star hotels and the way that they deal with feedback is just immense. Right? So it becomes the thing that they live for and they have senior level meetings of all of their heads.

Departments every single day to go through every single bit of feedback and review it on a one-to-one basis, it's that their life. And but in terms of the way, if we think about candidate experience, which is probably the oldest part of the experience from, from HR perspective candidates. So to have this clear dilemma delineation between how you are as a business and how you are as a recruiter.

So we did a piece of re and also they're heavily skewed to people who like you as an organization, as you would imagine. Right. But we did a piece of work. I did a piece of work 16, 17 years ago with one of the big retail banks in the UK, where we looked at, we could find out who banked with bank. The.

Candidates banked with because we were paying expenses. So they had to give their account number and we can look at the source code and go, okay, well, they're banking with Nat West and their bank Barclays. And the bank concerned. If you looked at the overall population, you'd expect about 15 to 18% of all candidates to becoming as.

Customers given the time. And we were, we were doing a big campaign to, to hire a thousand people. So it was, it was big, it was over 50% of all candidates were customers. And when we've seen that sort of data we realized that, you know, it's the, if you're annoying them, you're annoying customers. If you're not treating them well, we did a piece of work with an insurer here in Switzerland.

Eight nine, 10 years ago, looking at candidate experience again, in that, in the recruitment process and the way the recruitment at the time working is that would, they would not reject anyone until somebody had got the offer, right. Sort of you've accepted offer. So the. The average time that a person who didn't get called for interview was waiting to hear anything was a hundred days.

Right? So if you're called for interview, they'd be hurt. They would appear after 15 days, half rich, right? Because the, the, the, we get called for interview or they get a cough. And but the people who didn't hear a thing would have to wait a hundred days that. You know, whenever we presented back, we had one on the marketing team.

And then I just said to them, you know, the dates that we've seen suggests that these people are thinking you are not as good recruitment pride, they're thinking what will happen when I get it, I need to have a claim. And I do that process. Is that, do they handle the claims process? Like they handle the recruitment process.

So I'm starting to bring in that data is. But that's showing how you can use a sort of sip data like times after the recruitment system and they calculations and that tells you what's happening, but it doesn't tell you why it's happening. And with the insurance, it was a really simple fix the recruiters when they decided to do first level, they, they rejected someone.

And if it didn't get to the end of the process, they could reach back out and sort of say, Hey, something's changed. We, we, you know, we've had a look at your profile again. We'd like you, are you still interested? And most people would be delighted to have that sort of call. And if, and if they're not interested, it's probably because they found another job and they would have found another job, even if you had gone back without a conversation.

So just, you know, putting in prep, part of that process is thinking from an experience perspective, but isn't, you kind of have to look at the data as well.

Steve: [00:15:58] So my, I mean, my. My it's not so much an assumption. So what I see and what I hear certainly in the industry, I think there's a lot more listening and research being undertaken generally, particularly from HR point of view from an employee experience, point of view you can see that by the various tools.

And I know this year, a lot of the listening organizations are doing very well as obviously organizations are increasing the. The policies or the, certainly just keeping a closer kind of finger on the pulse with how the

Andrew: [00:16:29] employees are

Steve: [00:16:29] thinking and feeling. But I just still get the sense that for particularly like situations like candidate experiences and example, there's not enough.

Appreciation maybe is the word or understanding of the opportunities that exist. If you just put some research budget aside and actually do some proper thorough research and the opportunities that exist in terms of themes or identifying moments that matter whether it's good or bad or indifferent to then turn that into a narrative and a story to influence stakeholders, to turn it into something meaningful.

That helps you deliver what you need to do as a stakeholder, but also make the necessary improvements in experience. That's something which I definitely, I really stand by and I think that's something which. I think there's not enough research happening. Yeah.

Andrew: [00:17:17] I would agree with you. And especially this year, lots has changed, right?

The business environment now looks completely different. The, you know, we've been doing lots of work on the experience of working from home and the data that we were getting in March, April is very different from the data we're getting now. Right. And it's when something changes you need to. Match more frequently.

Right? So if I'm heading for one of my long distance marathon swims, I might be tracking my weight and eradicate regular basis. So I'm trying to try and get, get fitter, right? I'm in desperate hope two months before, and I still need to bring myself down, but I never met my height. Right. I don't measure my Heights on a weekly basis, but I might measure my weight because I don't expect my height to change.

So therefore I don't need to measure it so often. When something changed, when you realize it changes, you need to capture more frequent data. Now, I think you've seen this year a polarization of what what's going on out there. There are a bunch of companies who are measuring frequently. Who've put things in place.

But I know a lot of companies who have stopped doing things, you know, sometimes the executives think that they'll hear bad news and therefore they won't ask. But they're a huge. Optimization and, you know, even just operational efficiencies that you can get from listening to people and working out what the problems are.

A lot of the stuff that we've. Tradition. Traditionally found in texts are really low hanging fruit right there. Often we put in a policy and there's this side effect. Or we've, you know, if you did something really small, that it would make a huge difference to people's lives there. And lots of it is just often it's just.

Somebody who's implemented something, which for 90% of the population makes sense, but for 10% of the population is an absolute pain in the neck. And usually they're really easy things to fix. Right. So. Putting those parts and doing those sort of small gains all the time can make a huge overall difference in terms of approach.

But sometimes you do find quite big things that people just aren't aware that's going on. You know, executives talk to then, you know, and minus two, and that's it. They never hear what Martha in the Berry branches really experienced. They're going to set it to give you the sort of, some of the more trivial examples is things like.

I know one of the big tech firms, their warehouses are so big that the people can't get to the coffee room and back, or the toilet in back in their break time. Because they can get 15 minutes break and then they've got this warehouse it's 20 something football field slug, and they just physically can't get to the wet.

So all of the thing is sped book, but some forwards to the, to the break room or, you know, we found places where the coffee machines and the time when it comes on exactly the start time when people go to work and people actually wants to go in 10 minutes early and grab a coffee before going out on the floor.

It's that sort of thing that, you know, Just put the timer on for 10 minutes, Saturday. Okay. That may improve.

Steve: [00:20:22] So yeah, there's a lot of those hygiene factor stuff. Isn't it? I mean, there's a, there's all of those kind of little basic nuances. When you walk into an office, is it the chairs break on a regular basis?

Is it in a call center for an example? Is that the headsets aren't particularly great? You know, all these kinds of things are, you know, they do add up as part of an overall experience. And I think rightly so. I mean, I think, as I mentioned in my last podcast episode was. I, you know, I always really, you know, advise is yes, you start listening, but also be ready, like, especially in that first set first or second set of data with kind of a pair of like white gloves on ready for these low-hanging fruits just to come out.

So you go, right. Let's do it. Let's, let's solve it quickly just to build that trust with

Andrew: [00:21:02] employees. So I think,

Steve: [00:21:04] you know,

Andrew: [00:21:06] that's really, really important. The And often these things that we're talking about are really simple effects. So, you know, if you've got A big surfer going on and you literally two to three days after closing, you can have a good overview of the towel off the text or the survey results.

These days. It doesn't take six weeks anymore to do this type of stuff. You know, put something on corporate com saying we heard a lot about this type of thing. Therefore it's fixed. Right. Two or three days after. And I remember actually pretext, we had a client a Swiss telco company, one of the big, big mobile phone phone companies, and that chief exec used to read.

Every single comment, right? Every single comment you took her home and what was more crazy and you'll know, no they're sticking in second stock him. He was Swedish. He used to read them in that, in the original language. He refused to have been translated than here, here in Switzerland. The survey was in four languages, right.

Because of the national boundaries. So he used to meet every single one. And I remember we used to get lots of comments about, you know, I know you're not going to read this. And he did the first town hall which was probably a good year when it goes. And you know, to the person who started that comments.

I know you're not going to read this book. I read every single one of these comments and the response rate and the, and the level of, you know, one of the things we often look at as the length of, of responses we see. The median response in employee surveys across most organizations about 25 words in a question, right?

So they, they provide reasonable ones. We see answers over a thousand words, right? So we see some really long ones. So we see a lot of people who say nothing or actually physically use the word, nothing, you know, what can we do to improve here?

Steve: [00:22:55] So the learning here is, is from a state senior stakeholder, whether it's a chief exac or MD or whatever title that it's, it's all about.

Them showing up as leaders to say I'm also either listening or I'm very much, I carve a proportion of it to actually really. Engage in this. Yeah, absolutely.

Andrew: [00:23:13] Absolutely. We, I mean, our clients are very large because this problem doesn't scale. Then if you've got 50, a hundred thousand people dealing with a tax from that is, you know, more than 50 or a hundred times bigger than if you have 1000 people at that company.

So you do need to start applying algorithms to start helping you read it. But. You know, we drink

Steve: [00:23:38] socially, not how can, okay, so sorry, because I put important question just came to mind. So how can organizations then, because I think this is around, it's kind of that, that kind of triangle, isn't it, you've got all the data and it ends up becoming, as you mentioned, this kind of one, one statistic that is seen at kind of C-suite level.

And, and, but the story or the. Something that's a little bit richer that could really hit home. I'm really influenced at that level. How can organizations kind of adapt something in that as well? I mean, I'll give you a very quick example. We say, I IDO for an example the world-renowned design consultancy.

As part of some of their research phase in the design process, they once they obviously synthesize and start to create personas, they actually then create proper life-size cutouts of the actual personas that give them a height and then actually bring it into the meetings and showcase them just to start to create.

You know, I mean, you know, it's, it's slightly different, but it just creates this visual, different visualization and a different story too, for people to connect with. How do you think companies could maybe do something like that with data insight data? Just to bring that out a little bit more outside of just the numbers.

Andrew: [00:24:51] Yeah. So then that reflects into one of my frustrations. 10 years ago. Right. So we would do some wonderful analysis, wonderful statistical analysis, and then we'd show a few quotes and the executives act on the quotes. Right. But I've also done this, you know, this statistical stuff took me ages. Hey, is it quiet?

So we, it, so, yeah. I mean, this is why we moved over towards, you know, we don't do no numbers but we think you need to mix the numbers and the data, the numbers tell you what's happening. The date, the text tells you, why is this happening in terms of humanizing it? So. I wrote something about this in September, I believe.

The first part of what you do when you're, you're doing statistical analysis on texts is you identify, what are the topics in that text? We traditionally have done a bottoms up approach. What is in the, in the research community called the inductive approach, where you look at. The text itself and sort of say, okay, what, what's the key themes within this text?

And the text for John newest partners share would be very different from the text for Google, because they would talk, the employees would talk about different things. In fact, they might use the same words, but mean different things. If you and I, if I think about what would a Google, a Google person might talk about wanting more healthy food choices.

And that would be about the canteen and the benefits and the free food as benefits. If somebody in Waitrose or John knows Punisher discussed that possibly Waitrose or property made products on the shelves. So even though you have to interpret the same words in a different, different context. So the first thing you do is you identify the themes and then you go through every single answer and code it and say, yes, this is talking about training.

This is talking about free food. This is talking about. My commute whole range of different things. But that doesn't give you much insight, right? So on the negative questions about what what's not going to work in for firm managers, almost always in the top three. Right, but it doesn't really tell you what what's this about.

And you can go into a bit more detailed and start having more granular, but, but just counting the instances gives you a sense of perspective, but it doesn't really build this story. So we then like using an analysis, which says, okay, well, managers are often. Linked with communication. People talk about those things in the same sentence, or then linked about micromanagement and micromanagement links to working from home.

And we built this, the sort of network this graph about, about how people are doing it. And then we can cluster that using statistical techniques and say, okay, actually, you've got. Six different populations who are answering this question in broadly similar things within that population, but very different.

So within a traditional. Surfing, they might talk about things to do a business and performance and the values, the organizations. And we know that, and we know that's one type of group, and then there are other, we'll talk about their benefits. Another we'll talk about operational different difficulties or that creative element.

So we, we create clusters and segments. Then we can take him the demographic data and that demographic data might be their age, that gender whereabouts in the organization workforce, but there might also be behavioral data. It might be how they've answered the concept of questions. And we can sort of say, okay, well the people talking about the values and the strategy of the firm, we know that two times more likely to be a high performer than you'd expect given the size of this population.

We know that they're 50% more likely to be a woman. That the university educated they've been here within the three years. And from there just by getting the, this type of analysis, we've got a good starter of a persona we've identified that needs we've mapped on a probabilistic basis. Who's more likely to be in this group as possible.

And then combining that with probably some quotes from that group gives a really human picture. I would love to have a full sized cutout of, of Jane in that group. Who's, you know, who's, who's who they're, but you know, we get, if you look at a typical persona sheet that many experienced folks use we're managing through algorithms to find that sort of stuff.

Really quite quickly. And we've got data on that. So and this, this is a bit of a difference between the sort of. I mean idea. Have you got the idea of cards, the 52 greeting? No, I

Steve: [00:29:29] haven't. No,

Andrew: [00:29:31] I've got very old copy of it here somewhere, but a lot of their things in terms of going out and watching people in their domestic environments or whatever, to do that type of way, that's super expensive.

Super time consuming research that you can only do with a relatively small number of people. Unless you have unit, if it's budget song. You know, that type of thing, even still, you can still only do it on that small ones. And certainly if we're doing things like a user usability study, you know, you will know that the general advice is you probably don't need to do many more than 10 and you find the results.

Right? So, so that deep qual stuff now, the work that we do is not such deep qual. It's not like coding interviews, which we have done. It's very broad, but. There's a few advantage Pantages over fret one, we can see a global population quickly. So if the experience in headquarters in London is different from the experience of a sales team in when it's areas we can see and show these two groups are different.

So we can look at this at the end. Recruiting for those sort of studies. And when we did usability, UBS, where you have to use to do it in the, in the London branch, because we could, we could recruit somebody who's recently jumped off a plane from Asia and we could get an Asians perspective of this rather than a, a bridge perspective.

But. Unless you start having those numbers, you can't balance out quite so well. And, and secondly, there's still is a worry for me about statistical. Felicity and you know, how, and it's all to do with half the 12 people or 10 people you recruited being free, him being fired. The last point is what we do is remarkably cheaper than the qual stuff.

Right? So UBS, I was, you know, I was at UBS before the financial crisis, but we're still pretty healthy. And we were able to spend well into six figures a year on research. Well into six figures, doing everything by the book, how you do it, if you you know, if you're the front office business, and we've had instances where we've have clients doing that on things like employee value proposition, and they've called in a big research consultants in they've spent six 50 euros doing this research, and they've got within a few percent difference of what we did.

In a day and a half, I'm looking at a big data set, right? So you can do approximation, you know, if I was doing usability now, I would probably get us using zoom, get the, get the design team in one room, get somebody using the site and they're guided through and do the usability study, but just doing the house right.

Do it today. And you wouldn't get, you wouldn't get such a wonderful view that you would get hiring somebody whose job it is all the time. But. You probably get 80% there. You probably get 90% there. And sometimes it's about that. It's about being pragmatic, especially in HR, when the budgets never. Well, that's

Steve: [00:32:36] what I was going to say.

And I think that's something we've going back to what you said earlier on about some of the changes is we think because technology has, is now and continue to allow us to the price is coming down. Definitely. We're able to listen to much, much more kind of on a granular level. You know, listening posts are, you can have them across the whole entire journey now, less site, the less siloed or moving away from just the annual engagement surveys.

So I think that the ability to capture. Has increased, which is great through technology. I think for me, I still think where we've got a little bit, way to go. And I think this is why for me, the, the text analysis piece, which is obviously a big passion and focus of for you and your organization, I think that's the bit, I still think there's this.

Just, you know, a little bit of kind of stick man, stick woman kind of, you know, explanation to go. This is kind of, this is how it works and this is kind of what actually you can get. These are the benefits you can get through this in a very, very rapid period. And, and allow you, your navel, you to do X, Y, Z.

Because I think sometimes when we get into some of the. You know, I, I, this is going back a few years, but I mean, I built a candidate experience, measurement tool. I've never done in my life that, you know, built a dashboard. We're starting to look for kind of qual reset quo, qual technology. And back then that wasn't that much on the market, but you can get deep into it very, very quickly, lose yourself as a non-expert or non-research mindset, which a lot of HR functions and on I'm not, you know, just that's the way they are.

So what would you. Thinking about this kind of from you know, a simple approach. Is there any kind of step-by-step approaches you would, if somebody is not doing any listening right now, is there any kind of rigid, basic guides or some low hanging fruit? Let's get back to that with people that could start it just to at least.

Cause, you know, you know, it's like, Andrea, you've got to influence internally to get the budget, but you've got to start somewhere in order to influence to get the budget sometimes or more often than not and, and do a business case. So is there anything, how you approach your recommend?

Andrew: [00:34:34] We've we've just got a client who I was speaking to earlier this week who had just done a trial, a trial on a free trial of one of the survey tools.

Right. So, and, and. There are things that getting an expensive thing can do that survey monkey can't, but that icing on the cake, right? You certainly can pilots and tests and, and a lot of what should we should, I believe what we should be doing. And as you know, Certainly an experience, but in HR a lot more is pilots and experiments and, you know, get something and just run it for, just get the permission to send everybody sort of thing there in terms of The design of the surveys, you know, keep it really short and simple.

I suspect that you can get a template off off a whole variety of sites of what would a great candidate experience survey look like, but you'd probably can keep it to 10 questions. I would, I would ask for a. Some form of experience EMPS as for that type of thing, I would probably look at a satisfaction type question or ease and friction type question.

But I would also make sure you keep to open text questions. And we always recommend if you have to do these in pairs. So you say what's the best thing about the recruitment process and what can we do to improve proof this? And the reason you asked two fold is very fast, similar to the way that when we're being coached, give performance feedback to our teams, we always.

You know, start with the positives, tell them what they did well, and then let's look at the challenges. You just get so much richer data. If you ask everybody those questions in the consumer marketplace, it was often they would ask EMPS question and then a Y. And they would assume that the people who were positive didn't have any improvements.

And they would assume that the people who are negative didn't think everything went well. And that's obviously not true. The other thing, and this is a, you know, Most of the technologists. Won't tell you this. When they're trying to sell you a tool is that sentiment analysis is remarkably inaccurate, right?

So there's been academic studies that show that even the best in class tools getting about 65% accuracy. In fact, it's a hot, if you look at the research, it's a hard problem for humans to agree on. So it's one of those things we all think we can do. But if you've got five people to review something and in the room, they won't get a full agreement.

So, so you obviously have four people who are wrong in that room. So it's a hard thing. And by us, by getting the question to ask what great things, what are the negative things to a lot, much greater degree, you enable you to understand whether somebody mentions it's a positive or negative. What the reality of.

Sentiment and text in feedback is that people don't write full sentences. Right? And if you get to a thousand comments from an employee survey, a human might not bear to understand a hundred, you know, what was the person BB trying to tell us here, they write a few words. You know, obviously when they start getting those meetings 25, but I remember one instance where we weren't getting the performance on one of our models.

So we just reviewed the top, the 5,000 shortest responses. And we could only work out about 700, what they actually meant. I just have no idea. It was just seven big. You see, you can't expect a computer to do it. Yeah, that, that counters, that problem, that the technology, it, I just don't the 10, I was just going to be there anytime soon.

So yeah, try something, try, try and quick. I mean, There's lots of information. I'm sure you find a template, but make sure, certainly at the earliest stage where you don't really know what the issue is, you ask open choice questions and this to counter the issue that I think you and I both had them certainly in the hotels you need, you stay somewhere and they send you a thing and say, You know, how, what was it, what was the food?

Like? What was the internet connectivity? Like? What was this? So they give you these 30 or four mics. And the thing that really frustrated you is you got into bed, you had to switch the lights off on the other side of the room. There was no way of doing it. They're all linked together. Right. And I just call, I just want to switch the lights off before I go to bed.

Right. This is, you know, and it becomes trying to understand, I have a PhD in hotel lighting systems. Yeah, you've laughed. You've been there. Right. And they don't ask that one thing, but you really care about, and by just asking, yeah. Open question and not go through each of the points, you, they can tell you everything.

Right. And they'll tell you things that you never thought about. They'll tell you that. One of the things that frustrates the executive team is that the car Huck's full of potholes, right?

Steve: [00:39:25] Yeah. Yeah. And I think that goes back to if a lot of the hotels, because you're asking about a specific part of the.

Customer journey that they want to be scored on. So how's the wifi. They want a score for the wifi. It's a, yeah, I think it's great advice. And it's probably the, the foundation is just to put all of that to one side and actually you're right. I, I think I'll share a couple of tools. It's like get feedback.com there's type form.

They're probably two of my favorites because it's UX and the UI is amazing. Yeah. I mean, you know, once you start to see some of those responses come through is really just a basic set, you know, EMPS or CNPS has been candidate You know, you're able to also then just slice and dice that by, by if somebody's giving it a one or a 10, whether they're an attract, you know, a promoter or a detractor, you're able to ingroup those detractors, pull it, pull it just to one side and then look at then the verbatim.

And then the comments just gives you a, it doesn't have to be perfect. But again, as you say, it's just playing experiments with it and getting a feel for what starting to come out.

Andrew: [00:40:29] Yeah. And certainly in early days when you're piloting and experimenting, Something like candida experience. So I said ups, when we were doing new joint experts, we had 16,000 tires.

We were right up there with the top 1% of organizations who have that type of that to the hiring issue. Right. So Whereas, you know, we do candidate experience for a bunch of clients. They are almost exclusively in tech sector and they are on massive growth. And therefore hiring is kind of what HR is about is the number one priority for the Ms.

Business. Most organizations will have a hundred, a few hundred maximum of people joining every year or going through these type of process. Just read the right. Don't put some good technology. Just read them, do stick the ideas on post-its and arrange them on the wall, you know, do really low tech stuff, bring those into, you know, bring those quote cutout.

Each quote puts out on a, on a post-it and then your workshop, just get, do, do a card sorting exercise, right. And just get people to do it in a really low tech way and to engage with the data in that, in that sort of stuff. So. Don't turn for complicated,

Steve: [00:41:39] especially, especially getting started. And it's, it's easily done.

It's definitely easily when you've proved

Andrew: [00:41:45] you've done it improved. What's working. What's not work. You can then go to an expert. And puts it in, are they extra 20% effectiveness you can put in processes, you can put in technology, but you know, the way that we build our stuff is we always just.

Prototype and pilots and, you know, and we, you know, we'll read an academic paper on something new technique in text and we'll try it out. And then we bend most of those ideas after, you know, after a week or so. Cause they just don't work or they're not giving us performance benefits or they're costing a squillion dollars on own Google compute servers to do the processing because it's, it's, it's just so heavy.

So, you know, Some of it, but you piloted as quickly as possible and you abandoned it as quickly as possible to having there's enough stuff on the internet that you can find out how to do a straw man for 10 and you know anything.

Steve: [00:42:40] Yeah. Yeah. Lots of good stuff. Well, look, Andrew that's the time is absolutely whipped by, so thank you so much for yeah, just sharing a little bit more into your world and.

The world of text analytics and the importance of this stuff. It's you know, the thing, the more, the more we at least start to listen and to understand what these words mean and in, in what context and what benefit, and then turn them into benefits, both for the organization and more importantly employees as well.

You know, it's vitally important right now and I think. The cornerstone of any good employee experience strategy is listening strategy. So it is the cornerstone probably in my mind, probably one of the foundations probably very close to in a target operating model. So yeah, no, really, really.

That's why I'm super interested in your work and thanks so much for sharing. And how can people get in contact with you? Have you got any, do, do you have any kind of I know you're a keen blog writer or I know you've done some. Some various posts over the, over the years of you. If you've got anything, perhaps we can share a link to on the, on the page or think of interest

Andrew: [00:43:39] that you think could, I'll pick out a few.

There's this, if you look on our website, there's over a hundred. Quite long form articles over the last that we've done them. And fast majority of them are there there's some, the most popular ones have been in segmentation, which has been a topic where explodes again early in the there's there's articles on.

Technical issues and there's a contact form and that's probably the easiest way to get in otherwise, you know, reach out on LinkedIn and try and get back. But yeah, and I'm happy to just get on a call and talk to HR teams who are doing this stuff for the first time and just give them a few, a few tips and habits.

We believe that, you know, our mission is to make analyzing text as easy as analyzing. Numbers in the firm, you're getting from a spreadsheet. And if we can do that, then you know, we're helping, helping somebody and really therefore helping the employers to get better experiences, which is a good, good way of looking at it.

Yeah. Yeah. I mean, we're definitely

Steve: [00:44:38] in the employee experience moment right now have been for a few years, but they said, well, yeah, Since he right in the thick of it and COVID of courses accelerated a lot of those conversations. Yeah. Yeah. In so many different ways. So now it's in a, in a funny way, I think there's some good to be had out of this awful situation as well.

So

Andrew: [00:44:55] yes. Oh, they just showing it's accelerated the trend that was already happening and it's, it's accelerated by about three or four years.

Steve: [00:45:03] Wow. Amazing. Well, that's a good, that's a good positive to end up speaking to your soul. Great, Andrew, thanks for your telling me. Really appreciate it. Okay.

Andrew: [00:45:12] Okay.

Steve: [00:45:15] And there you have it. Another episode of the experience designers podcast. Thank you so much to my guest, Andrew, absolute pleasure to have you on the show. And of course, if you wish to connect are all of the links and also a couple of other helpful blog posts or helpful documents that we talked about, the end of the show.

I'll of course put links in there for you as well. Thanks, ever so much for listening and until the next time. Bye for now.