WB40 – Episode 89 – Data Curiosity
On this weeks show we chat with Ben Schein, VP of Data Curiosity and Innovation at Domo. If you want to hear more from Ben, there’s a slightly extended version of the interview available here.
We also see the welcome return of Yin and Yang.
Here’s the Yin:
https://www.newscientist.com/article/2186512-exclusive-uk-police-wants-ai-to-stop-violent-crime-before-it-happens/
And here’s the Yang:
https://www.1843magazine.com/features/a-rockers-guide-to-management
Don’t forget you can join the conversation over on our WB40 WhatsApp group. Just send us a message on Twitter for the magic link.
The following transcript was created by https://Otter.ai – it’s included to improve searching capability on the website and it might even be useful to read!
Matt Ballantine 0:20
Hello and welcome to Episode 89 of WB40 the weekly podcast with me Matt Ballantine and Chris Weston
so we reached the month of December sleigh bells are ringing the Christmas tree is up the turkey is starting to defrost How are you
Chris Weston 0:55
well I’m defrosting nicely and as you say that up I’ve got it’s a little bit distracting actually because they’re on the outside of the house and they’re flashing away
and little white icicle type of things. And they
and then I forgot about them and I’ll be working on my computer and then they’ll see them on the corner of my eye. So that’s slightly distracting. But yes, it’s a it’s December.
It is a time of year when people start thinking about packing them for Christmas. So the amount of work that gets done is going to take away dramatically in between now and Christmas holes. But still plenty going on. You’re you’re busy. You’ve got new new customers I gather.
Matt Ballantine 1:38
Yes, I have.
I did a fantastic session last week with a retail client,
very high end retail client and what they let me in there for but no really big Lego airplane building exercise to be able to illustrate some of the concepts of agile and it was fabulous. Really, really good fun.
And I was also with university last week as well in sunny Ipswich when I say something I mean really, really rainy.
And then there’s a couple more that look like they’re coming out with would work in the moment as well, which is some Yeah, it’s good. It’s good. I was trying to look at I think I’ve had six or seven different client things either work or, or meetings about potential work in the last five working days.
Chris Weston 2:28
It’s quite high hit rate. That’s the way to do it. You know, you’ve got to be having those conversations, things to come off. We’ve got I’ve got a piece starting next week, hopefully with the colleague Dave who can do a piece for a manufacturing company around their their it and another project which is thundering towards some sort of inception. So yeah, it’s busy and interesting time.
Matt Ballantine 2:55
Having said that, I was talking to another potential client today who said that their businesses basically falling off a cliff. And the consistent reason has been given by their clients or potential clients is the B word and I won’t leave it to anything off of your imagination to where they were more that way. I don’t, I don’t want to everybody. So anyway,
good. We have an exciting show lined up this evening, we have an interview with the Vice President of data curiosity and innovation at a software company called Domo talking about curiosity. And what it means and my organization struggle with it. And we got yin and yang as well, because
Chris Weston 3:38
that’s what we do every other week. That’s right. It’s it’s very exciting, we should be should crash on that. So everybody will be absolutely delighted to learn that we do have a uni and this week, it’s something that people look forward to, and so much to highlight of some people’s fortnite. When we do this. Now, you realize that
certainly highlights of mine
Matt Ballantine 3:59
now that I can find.
Chris Weston 4:01
So let’s start off with the staff with the misery with your your, your own dark place on the internet. Yeah, this is something that I can’t remember, I picked up on
Matt Ballantine 4:11
this. But it’s a story that was in the New Scientist. And it is about a system called the National Data Analytics solution, which is using a combination of AI and statistics to try and assess the risk or somebody’s committing or becoming a victim of gun or knife crime, as well as the likelihood to someone falling victim to modern slavery. Now, you know, becoming a victim is one thing and that, okay, I don’t have as much of a problem with but I have a really, really, really big set of misgivings about this being used to proactively try to be able to stop people committing crime
Unknown 4:53
because it’s
Matt Ballantine 4:55
using data and statistics, which means all it will do is extrapolate all of the biases exist, conscious and unconscious that have existed in all crime recorded up until this point, and it will end up if you look at the first that there is, and quite rightly, the first and there is around these to stop and search this has got nothing on the fast that there is around stop and search. Because the problem with this is, what it will do is it will target people from poor income groups from non white ethnicity groups. And it will reinforce a bunch of stereotypes about the police wanting to be able to persecute those groups for one reason or another, which will not make crime any better.
Chris Weston 5:40
Well, okay, if it’s done badly, it will do those things right now. And as you can imagine, I’ve got I’ve got an opinion on this. I mean, I’ve got up on everything. But I’ve got slightly bit more informed opinion on this given I’ve done quite a bit of work in the criminal justice over the years,
there are there are lots of models out there for predicting criminal behavior, there are various assessments that people do around the prison system and probation system to say, How likely is this person to refund and it’s, it’s a, these are all models based on
they’re all based on research of one kind or another, right, like, other than they are actually based on academic studies. They’re not just made up
as much as you can ever say that about anything academic. And,
and it’s important to do that, you know, so if you can do it at all, you’ve got to say, are we having an effect on this person. So we got somebody who’s criminal behavior is down to some sort of anger management problem, or send out some sort of side psychotic issue down to the valley, they’ve got a job, or they, but they write down their relationship of relationship or family issues. And you can, and you can then predict if you can figure some of those things, if you can get them a job of getting somewhere to live, you can reduce that propensity to refine. And that’s, that’s a fairly well understood thing.
So prediction is, is a useful tool.
And whether it’s, whether it’s well done or not, is another issue. Okay, and you can bring AI into it. And then you’ve got a whole nother set of problems, okay, in terms of
a lot of the production and tools at the moment to a bit like the kind of, you know, the one that the heart one so that they’ll say, Okay, do you smoke, or how heavy Are you What’s your blood pressure, whatever. And then they’ll give you a score and say, you know, whether you like to live past 50 or not, or they have to go and start it into something.
And it’s all pretty, you know, there’s a lot of clever stuff got into a lot of hard work, a lot of research, but the algorithms and everything that they’re pretty, they’re pretty straightforward when you enter the AI world, because then you like you say, you are in danger of building a whole bunch of boxes, you don’t really understand how it’s how it’s working. Because because of the way the AI works in this in a very encapsulated way. And there’s also some things about, you know, this is, are you talking about people who are on probation, there’s a sentence passed on, and then they’re being supervised? Or are you just talking about people who are,
let’s face it, these are probably going to be people who have been through the criminal justice system, right? So
and you do your time, and you pay your price, and all that kind of thing. And that’s fine. And then you’re a normal citizen, but they are going to be people who are known to criminal justice system, and it should be about as much about helping them as doing anything else. So it shouldn’t be about it shouldn’t be about
just coming with a whole bunch of things and saying, Oh, no, you’re a black person. So we’re going to come and see you. It should be about the the actual indicators based around criminal genetic issues. Not you happen to be from that street. Okay. But if you
Matt Ballantine 8:55
and this is the thing that would work for me, massive alarm bells go off about this story.
Absolutely, if you can find ways to be able to be able to reduce the chances of people re entering the criminal justice system repeatedly. Because if you want to know one good indicator, somebody likely to commit a criminal act as if they’ve been involved with the criminal justice system. Because recidivism is incredibly high. Because the system is at the moment doesn’t work. It’s this weird thing between being a preventative measure through punishment, a purely punishment thing, and there’s very little going on these days, as far as I can tell, about actually trying to be able to prevent people from committing crime again in the first place.
But if you read in read the story place, funding has been cut significantly over recent years. So forces need a system that can look at all individuals already known to offices with the aim of prioritizing those who need interventions. My surgeon he says in Donnie, the police lead on the project as for exactly what will happen when such individuals are identified. That is still a matter of discussion, says Donnelly, he says the intention isn’t to preemptively arrest anyone, but rather to provide support from local helpful social workers. Now, we know already that local health and social workers and mental health is in absolute catastrophe state at the moment, because it’s been so on funded. So basically, they’re working on the bit that can identify some potential people that you can then point and go, they might point and those people might commit crime, but it’s doing nothing to actually talk about what you might do about and that’s totally I’m sorry, that is totally ass about face.
Chris Weston 10:27
Well, it’s also very narrow in that sense. So and you can you can point to, and I don’t know the numbers, and I can’t remember the details. But there are several instances where you can go to a town, whether it’s somewhere like Nottingham Bradford or, or Milton Keynes, or wherever you can go, you’ll find that a massive proportion of police and social services time is taken up by a very small number of people. And you know that that’s that’s where your problem is coming from. I don’t
Matt Ballantine 10:59
need it. I
Chris Weston 11:02
think so. So I suspect and this is maybe cynical of me, but I suspect this is a very narrow piece of work
based on some things that that they know the answer already to see if they can get the AI to predict it, and see if they can build it on it a bit more, I suspect is quite small beer if I’m honest. But because it sounds a bit. ly Minority Report New Scientist and making a slightly big thing out of it. So I do, I do understand your misgivings. I suspect it’s probably not as
there’s this either. I don’t think anything’s going to come of it. Not not to die. Anyway.
Matt Ballantine 11:40
Anyway. Okay. Yeah, I’ll have my rant. So anyway, bring us bring us bring us up again, Chris, what have you got, what is the, what is the, the gang of the internet for you this week.
Chris Weston 11:49
So the younger the internet this week is, it’s all about management and leadership and the dynamics around that, but it’s really good article,
I can’t remember where it is what
Matt Ballantine 12:03
it was on the economists and,
Chris Weston 12:07
and it was all it’s about leadership and management and using rock bands
as the template. So the author then talk takes us through different styles of management dynamic, so are you friends and, and, and, and it talks about the Beatles, I think, is the friends
model where they were no interest to be, you know, they lived in each other’s pocket for many, many years, I will very good friends and this kind of work very well for them. But they all do, they did fall out, you know, in pretty spectacular ways at times as well. And it talks about some some examples from the business world off people working like that, and, and almost making wrong decisions, because they didn’t want to not fall out with their friends. And there’s also examples of democracy. So it talks about REM, being very much a democracy where they all had the same important and the fact that that can also work. But it also can be a bit of a struggle. And it also talks about more,
Unknown 13:12
I guess,
Chris Weston 13:13
less less democratic system to talk about Bruce Springsteen, for example, in the fact that they recommend his band and E Street Band, it works because he is the boss, I guess that is the is the terrible kind of conclusion you gotta draw from that. So, but he takes takes decisions, he also takes more of the more of the money and things like that. But yeah, that’s the way it works. And it’s and people say that only it’s only work for that long because because of that, the that because that was the case you saying, if you’re taking responsibility, and you’re, you’re taking decisions Well, that’s, that’s part of being the boss. So he
the article is really good. I just think it makes you think about a different dynamics where we have in business because we do we go through these things that we we have people that we get on with very well, and you may be work with those people not slightly differently. It’s not favoritism. It really it’s more of a
you’re aware of this kind of there’s more dynamic than just you know, oh, here’s personnel who are happy to work with and there’s people you don’t get on with so well and there’s people who just won’t engage in the have that complete the are not here to make any friends you know, we’re just going to you know, Dutch people and
and there’ll be you know, absolutely ads I’ve worked very friendly but inside work is very transactional and we are all different and we react to people differently depending on the way of the relationship is me so I thought is really good piece of that is very thought provoking.
Matt Ballantine 14:37
Okay, good. I love the metaphor that sounds like very good metaphor. I just trying to think of the banter I think probably a lot of autocratic
musical geniuses with surrounding cast of people who change regularly said people like Matt Johnson of the who’d never really sustained a band for more than an album because
Chris Weston 14:58
it was actually really only rather than for example is a Marquis Smith was that
Unknown 15:03
oh yeah
Matt Ballantine 15:04
yeah
and yeah I see it
in smaller businesses I have to say like my own where I am basically a solo artist and that’s probably for good reason if I think about it but
Chris Weston 15:18
I’ve heard you described in similar
the way it’s the story kicks off actually easy nice that all the stories is really well known but it’s about Mick Jagger
coming back to the hotel late at night or in the early hours and and drunkenly trying to get hold of Charlie Watts say where’s my drummer and Charlie Watts terms of a bit later and punches mix Mick Jagger in the face and says never caught me your drummer and it’s and the whole point is that these overstep this thing is not it might it might be the leader of the band The child wasn’t withdraw from Charlie was one of you that’s kind of all right we we might be hit with that that might be the case that your lead singer but I am not your drummer I am a virtual me and when people can overstep these on said boundaries people do for that and it hasn’t been established in that sense early on
some people can see as a tech is taking liberties and we can all i knows exposure on coming
Matt Ballantine 16:27
this is very true, very true. Okay, good stuff. So that article from the economist and also a link to the New Scientist article will be up on the website at wp 14 podcast.com
Chris Weston 16:41
So Matt you talked recently to Ben shine of Domo P of the extremely long job title.
And it was it was interesting towards it, because you are really asking me about the subject of curiosity and how it applies to, to the $10 works with
Matt Ballantine 17:00
Yes, it’s a subject that fascinates me. And I was really interested to have somebody who actually has the term curiosity in their job title. So here’s what we talked about.
Unknown 17:13
My name has been shined and the Vice President of data curiosity and innovation at Domo. We’re a data analytics platform that helps people run their business from their phone and organize all the information they need.
Unknown 17:26
What does vice president data curiosity?
Unknown 17:29
Yeah, so I was I was lucky enough around like, I created my own crazy wonky job title, really the core of of trading the role what we’re trying to do, I’m sort of came out of my journey at Target, so the private Domo, I worked at Target large us $70 billion retailer. And what I saw was that really the way we started to change how that it was used within the company that I really I thought about, it was like, ways to create that curiosity. And that it was about more than just any set of tools or any report or anything there that but the way we structure the way we thought about things, the way we empower people had as much even more to do with it, then a product. And so my pitch, ironically, maybe the demos is the product was that I really wanted to explore how do we what are those other things that you need to change the culture around that and create data, curiosity with an organization independent of product, right? Because I really believe that you could have a very expensive products that can be implemented and create no data curiosity as well. And you could have a scrappy team using Excel and finds ways to create the data curiosity. And so really wanted a chance to think about that to go and test with our customers both and also test these things outside of retail outside of the United States, different markets, different cultures, all those things. And then eventually based upon the thinking and based upon some of the field testing, come back to our product team and say, Yes, we are selling a product, how do we continue to create a space that encourages these best practices and encourages curiosity? And so how do we change culture outside the product more on the process, and people in those pieces
Unknown 19:23
talk is, is I guess one of the legends of the use of data, it is a story that was a target with the with my friends,
Unknown 19:31
I story amazingly, I could not have been recruiting a target for data analytics, you cannot be someone without bringing that up. And then I think that is some ways frames it up in an interesting manner. Because like, one is the use of data to operationalize a process, right. So like, how do I use data to send effective direct mail, right, or email. So not that was all stories of the father saw it was offended that they was for pregnant person, the daughter of the later found out that we were right, that she’d been buying things that indicate that she was pregnant. And so I think, and then maybe it is a microcosm of word that has been like, that was very effective use of data. But whether the people running those campaigns were using that every day of understand what they were doing, or we’re engaging the data directly versus sort of this pipeline of like, I’ve require an algorithm, send it over the fence, have to come back and operationalize it. So I think that operational analytics is important. But it’s not necessarily changing the culture of how that is day to day, whether organization
Unknown 20:36
is the thing I think, because that the operational model for a mass massive retail like targets, Ashley many ways is the antithesis of curiosity, because, yes, it’s so process entry.
Unknown 20:48
Yeah, and I think it’s, um, in some ways, like those core prophecies benefit a little bit less than the Curiosity partly because they don’t need as much and partly because they are the core, right, because the they have been prioritize where I think like really creating curiosity and more flexibility, it’s more in the long tail of priorities, right? So we’ve always been able to say, how do we use data in these top five processes. But where I was always concerned was, what about number six? Does that mean, okay, if I’m really bad organization, that doesn’t mean just my top five priorities that I have, that scientists and engineers working on used at every day, I needed to also be,
Unknown 21:29
you know, number six, and number seven, and number 20. And so things that target like theft reporting, right, it’s important, right, if people are stealing from us, and how to understand it, but it never would have gotten, you know, like a full fledged data scientist, that engineering team, but how do we frame ecosystems and platforms that encourage that team to use that and have that curiosity culture on equal to the teams that are the number one thing like supply chain or a certain planning or things like that.
Unknown 21:55
So there’s some patent that the term of it is, what’s the essence of curiosity?
Unknown 22:03
I mean, I think a lot of ways the essence might just be asking questions, but then pairing the questions with actual it’s like being able to ask questions, but also be able to get the answers, right. So I think for a long time, what’s happened is, if you had someone who is curious, it asked the question, then you could said, No. So many times, you can’t get the answer, then you just shut it down. To me. It it’s asking the questions, the sense of wanting to engage with the data and with the data providers, whether it’s an analyst or that engineer, however, the some setup or even just a tool in order to to make it better and more relevant and go back and forth, right. And so I think the one of the things that would frustrate me like the sense of just wanting to be fed data, just wanted to be fed information like, this is the processes how to do it, right. Even, even if my team I thought, got the process exactly right. I just know that things change too much for that to be the way that we operate. And then when we have a more flexible platform, if you it’s, it’s the flexibility doesn’t matter at all, if you don’t have engagement back and forth, right? So if someone just wants to be given data in a specific format, and, you know, dump it into their system, or do whatever, it’s, I need them to ask questions, I need them to engage and need them to challenge even right, even if it’s challenging to my team, and I’d rather have that dynamic and go back and forth. And sometimes a challenge might still waiting to Wow, that’s going to be a little hard. But if, if 90% of the time because we do yeah, let’s try that. Let’s do it together. Let me show you how you can do that on your own. I think it sort of goes there. And again, like, I think it’s a mistake to think about it as just like this innate skill Jude’s let to be, it needs to be nurtured. It needs to be encouraged and guided. And so I often would tell my team, it’s like, there’s a fine line between self service analytics and telling someone to go take a long walk off a short Pier, right? It’s like, it’s not just like, here’s the tool, go do it. And then the analytics team security grumbles and says, No one uses it will know like, they need encouragement, they need to be shown how things can change and how to do it. So it’s not about just saying, hey, curious, it’s it’s how do I help someone be curious and help get them there? How do we go about trying to be able to, you know, allow people to make steps to change, you know, so I think it’s a process where we can expect it to change overnight, I do think there’s a certain level of empathy that reporting teams to have. So I think it’s easy to say, Well, no, you don’t really want that. And then that might be that might be 100% true statement to say, you don’t really need this, you need something else. But if either they think they need it, or maybe they do need it for some operational thing. So how can I honor that need and that and empathize with what they do need, right? So that doesn’t come across as like, this guy’s just trying to tell me something different, and he doesn’t care about my day to day, how can I honor some of that whether it’s okay, well, here’s the here’s the ugly version. Now let’s start unpacking what you really need on or sitting down with them and saying, Okay, what we’re going to start with this, let’s leave this to the side like that’s what if we looked at it this way, what show me what you do with the next and those kinds of things. So I think
Unknown 25:15
it’s dangerous not to come in. And one of the concepts that I really love that was brought out in the VR piece of curiosity is intellectual, if you will, right. I think that you can have with engineers, data scientists with analyst you, some of them have this intellectual hubris that can be dangerous, and it just comes across as, like, I know better, and you might know better answering things. But that’s not how you get people to change behavior, right? It’s not coming in as this great genius. It’s empathize with what they’re doing help and help them take baby steps from there. So maybe it’s okay, well, now you have your report that you always had. But then, oh, let’s do a summarized version because it actually saving time. Okay, once we do that, what’s the next thing? Oh, well, really, just trying to find out why what set up in the words that sends you when there’s an outlier, and you don’t have to wait to the report runs once a day or two, you click on it, it’ll tell you, right. So it’s like it’s taking people along those steps that then free them up to do other things and search showing a new way. Because if and it might end up in the same place. But if you just came into that I know best you should do an alerting system. You don’t need this report that feels very much like you don’t understand what I’m going through. And then there’s no empathy there. It’s you need to help them psychologically,
Unknown 26:26
it feels actually there’s public parallels through this journey, that organization to taken from moving from waterfall approaches to project delivery. To agile. Yeah, for sure. And the barriers there aren’t about the method. Yeah,
Unknown 26:39
and I think that was really the premise of my role. And what I’m trying to do is like, you do need you like, you need to think about your people, you need to think about agile engagement models and all these things, if you really want to change your culture for us to implement a system and have a report pop out. But But I don’t think that’s really there. And I think,
Unknown 27:01
you know, there’s not a lot of value just changing system, it’s just because you want a slightly better report.
Unknown 27:07
So I mean, maybe part of that, then it’s about being able to get people to understand why this is important. And the known certain world, the known certain the business environment, the political environment, or whatever,
Unknown 27:22
at many levels of analysis. Now, you can pretty much say that that’s got
Unknown 27:28
this concept of uka volatility, uncertainty, complexity and ambiguity that we have a world whether it’s, you know, what was going on in politics, and whether it’s what’s going on in in markets are the emergence of technology and the players? Yeah, that that feels to me to be the compelling organizational reason for this. Yes.
Unknown 27:47
And I often would say, like, Look, I can I have personal motivation for wanting our tools to work and to create this kind of dialogue. But it’s, it’s almost like, Okay, if I can’t even get someone to engage questions and curiosity around their data, or their reports, that’s a little scary, if I’ve asked her asking that person to engage, and how do I, you know, compete with Amazon and compete with Walmart, and think about these different things, and think about multi channel and, you know, omni channel retail, and all these other things. And so as much as like, I care personally about the data side, ultimately, it’s also like, well, how are we going to compete as a business and the overall objective, which is satisfying guests, leads and giving them, you know, good product and a quick way at all those things. It’s just a little scary to think like, if that person is not in this isn’t this mindset of just execute, execute, execute? Bonus questions, there’s no way you can say you’re going to compete with an Amazon
Unknown 28:42
Yeah, and I guess maybe within that as well, there’s also a shift from information systems being seen as profits as things that can predict the future for Yeah, as opposed to information systems being things that can be used to test assumptions and to be able to do low cost, yes, you know, theorizing and testing without having to take it to market and then you can take it to market on a small scale. And, and, and things emerged as well.
Unknown 29:13
Yeah, it is, it is very much that piece. And I think if you think about all the swirl around AI, and job replacement, all those things is it is like, you read these, these one attitude of like, Oh, my gosh, I’m going to be replaced by a robot. The other attitude can be, hey, I have a lot of value in context to provide as an input into an AI algorithm. And so how do you, you set it up as, Oh, okay, this, this, this is how I test things. This is how I do pieces. And certainly the the context and a lot of ways like this journey of curiosity is like, how do I sort of harness and the line and leverage all the knowledge in context that sits out there in the business, right, I was at talking to a potential customer last week, where they were frustrated about, you know, not being a to get the GL and, and all these other things. And it was sort of like, Well, that’s because they have these engineers just trying to figure out the GL, if they had a way to expose it out, and let the account manager who knows everything about everything in their GL, and all the different things, then they they don’t need to try to teach accounting to the engineering team, you empower the context of the knowledge already have an organization to help be an input into framing how that is done, right? So I think it just has to be there’s a lot of knowledge other even in an AI world that humans have. So how do you get them engaging with the data with the in a way that adds value as opposed to really was competing, this computer is competing with me for my job.
Unknown 30:42
The reason why XL has become so dominant is because you can use it relatively speaking to do just about anything, there are all sorts of sins the day to commit to that, you know, the amount of data that is embedded in Excel sell, for example, but nonetheless, because it has agency and it has flexibility.
Unknown 31:01
Yeah, well, it’s interesting, I think one of the things, you know, as we look at the data platform, and something that we hear a lot is like, people are scared of like, well, I don’t want to give people too much power, like, why I don’t want them to be edited. And it’s a little bit like, you are aware, this is happening in Excel with no visibility and know consistency, and all these other things. Like, why don’t you empower them to do it in a platform where, yes, they have the agency, but I also have some governance and some visibility, and I could see, hey, someone to share something with the CEO, and he’s looking at it, maybe I should just go double check. And the outcome might be, hey, that’s a great idea, we should integrate them into our core KPI job, or it might be that is calculated, you should be sharing that with the CEO. But that’s still a lot better than if they just did an Excel and email.
Unknown 31:47
One of the concepts I talked a lot about, as I’ve been traveling and going to conferences and whatnot, is sort of this dynamic, I think we can unpack and more of like thinking about like, old waterfall end to end thinking. And then really, I do Excel as we sort of end only brands, like, I just have this thing. And as I think the middle ground for me is like, end to almost end, right? So how do you create data platforms that are leaving a little bit of the agency within user, but still, you know, doing some of the foundational work to get you there. And I think that’s part of what we tried to do a target was thinking about, like, how do you empower lots of analysts, lots of business users to do that last mile of insight delivery on their own, because, again, like Excel, that end only thinking was awesome, because I can change the color, I can reorder the columns, I can add something new, I can do a quick graph, those are all really horrible things. It’s why Excel with been so popular well, and I By the way, I think the empathy goes both ways, right? I think they’re at the same time that there’s people that are on on aesthetic to the business case, the business and like, like, if you come into it, and say, Hey, you know what, you need to let everyone do every thing like, that’s also not being true to their legitimate desires to make sure things are secured to comply with GDPR to do all these other things. So like, it has to be a two way street with that, but the to say, Okay, well, you know, what are we doing? Or how can we, you have governance concerns? Okay, let’s make sure we have the certification process with make sure you can monitor usage and different things that like, lets you encourage some of that freedom but honor the fact that like, yes, you need to have some structure and governance even just legally and also and then there’s nuances between that right HR data about who reports to who and how much money people are making has to be more secure than sales data, a target right like that to come in and say, Oh, this free curious Well yes, just curiosity curiosity does not extend to knowing how much your mate makes versus yours our right let’s hop part of what we’re doing and so you need to meet them to empathy needs to go both ways you can’t just come in and be one sided on either either one of those
Unknown 33:59
one the criticisms that has been made of current management style and if you look at people like Tom Peters immediate, the old guard now with management theories with people at Tom Peters or Charles handy, they took quite often about how business school churns out a certain sort of, quite short term, it’s no leader business leaders that probably isn’t fit for purpose.
Unknown 34:29
No, I mean, I think agility, which we mentioned is the core part of it, right, just a target, we, we tried to focus our agility routes, three core values of small teams that have trusted accountability, transparency, learning quickly, right. And so like we just said, Look, we, you might try different practices you might be set up, you might be introspective, your team might be more con bond more a scrum. So I think like that kind of a curriculum that is really teaching the essence of agile, not a specific practice. But like, how do I think little bit differently
Unknown 35:02
about those things, I think is a big a big piece of it, you know, I think the other one is in MBA programs, and I have an MBA to is I can think back, but a few years now
Unknown 35:15
is, is really around like understanding the nuances of data. And that financial statements are different than an analytically directional analysis. And then some basic understanding of
Unknown 35:30
you know, of about some of the algorithms and the concepts and other they have to code but they should understand what’s there. And I think I do worry now an educational business ever sort of bifurcating a little bit of like, there’s like business education, that is all, especially the stage and all these masters of analytics and data science coming up. And I worry that they get really focused on just the checklist of skills. And then and then I’m, it’s on to that if you don’t have the soft skills, right, of understanding, empathy, and how do you work with people? How do you think about teams and those kinds of things. And so, you know, we just released a survey, some the fourth industrial revolution, or whatever was part of the big data, London conference this week. And one of the things was very concerning is that in the survey of battle leaders in the UK, a lot of people were de emphasizing those skills, liberal arts, agility, all those things, and I think that’s a very risky thing to do. Because if you just like, Okay, I got 10 Python people, I have 10 our people. And then if those people just have these technical skills without doing it, it’s just as bad as having the short term just business PowerPoint skills, or whatever, without knowing the data, or even Actually,
Unknown 36:40
it’s just that the have quantitative data skills as opposed to a more rounded set. Yeah, there’s a guy Christian match Burke, who’s done quite a lot of work around it’s no coffee and talking about how without the qualitative, yeah, the constitutive can be utterly mislead.
Unknown 37:01
Yes,
Unknown 37:02
I think there is a challenge there, because actually, the
Unknown 37:08
still to this day, if you want to be able to do any sort of analysis about qualitative data in a system, you end up turning into quantitative data. And that’s not the same as truly analyzing qualitative data, you know, things like, you know, techniques of being able to do
Unknown 37:29
sentiment analysis and those kinds of things. They’re not qualitative analysis, they’re constantly analysis of qualitative data.
Unknown 37:37
Yeah, well, and I think, you know, a lot of term, I think a lot of it is how do you use and certainly,
Unknown 37:45
you know, various, you know, approaches for categorization and classification algorithms, I think that becomes important, because then that’s part of what I think classification some of those approaches within the sentiment analytic space does is it how do you get a subset of qualitative data that is meaningful enough to give it to a person who can use that quality that piece, right, so, so target if we had on, you know, lots of data points, you have the search on our website, you have Google and Facebook, and all the social media and Pinterest, you have people coming into the stores and asking a team member, you know, question and they type it into the handheld device, it’s a lot, we would take all that data, we would use data science to help classify it, right? So, so that,
Unknown 38:31
you know, we can say this, this set of qualitative data is most relevant to this department. But then the ultimate value of it is not, not like how many positive negative are coming up, it’s okay, now I can deliver this qualitative information that’s relevant, because certainly plays like Target, we’re selling really everything, if you just have, here’s the qualitative data, that’s too much. And to your point, if I just said, here’s a indexing and positive and negative sentiment, it’s sort of interesting, but it’s still just very general. So one of my favorite examples was, I think there was a case where all of a sudden, they saw nitric oxide, some some chemical was popping, and they didn’t understand why a lot of asked in search of what it turned out to have these bath bombs for kids, my kids love them, right? to like, drove to the bathroom. That’s like, visibly, people were trying to make homemade versions. And so we saw that you can Google it. So whether that means you should put the Bad Moms that I all or you should think about important in the stores that sometimes people want, these things are, we should order more because more people are doing it. But that only comes like that would just be noise in this qualitative mess. If you didn’t have classification algorithms to help say, this is relevant to the home cleaning supplies department, because we know it’s an ingredient that maybe is there often, or whatever it is. And so finding the ways to it goes back to like, how do I how do I help get the humans inject the value but help them to sift through all that, that that and before they get to it,
Unknown 39:54
and there’s still that element of
Unknown 39:57
needing to take risk. So if you look the, you know, the anti example 40 years old now, the new coke Yeah. And that the data said the new coke was going to be more successful the Pepsi because in blind taste test, it was tastier than Pepsi, and it failed terrifically much more recently, was really just this morning, actually, about actually, all of the analytics about Amazon Prime said it would be a disaster. Yeah, and, you know, there’s also blessing is laughing all the way to the watch. So, so that was
Unknown 40:27
no, yeah, I mean, again, it’s like the data it needs to frame it. I think the other thing that we looked at a lot of target was there’s also a difference between measurement and strategic intent, right? So certainly like to retail, you’re always looking at sales, but what was your strategic intent with this initiative, right? Because you can hit that sales number and the like, it looks right. And like, maybe you say the test successful, but when you try to attract new customers, we’re trying to grow existing customers, you’re trying to move people off of a brand, all of those things start coming in. And that’s not a matter of the data. That’s a matter of clarity of strategic content, and then tracking that strategic content through metrics, right? So if you don’t, if you don’t get to that muscle, like articulating your strategic intense in terms of data, yes, but it’s, it’s different than to say that you can have all the metrics, you could say, Hey, we’re hitting your sales number, it doesn’t mean it’s a bad thing. It’s awesome. Did your cell number, but maybe you still need to go back to the drawing board to understand how to attract new customers, because you hit it by growing existing basket, not by,
Unknown 41:30
you know, going in, and actually growing your customer base and blistering
Unknown 41:34
10. Again, that’s how adaptability so that’s the ability to be able to use data to be able to ask further questions, rather than just try to get it to predict the future. Yes, it’s a full time. Yeah, if an organization wants to be able to become more curious, what are the sorts of foundations it needs to be able to put in place?
Unknown 41:53
I think, one is to, to really focus on this idea of and to almost end up How do I get people platform that allows them to do some of the work like, and it’s this idea that I can have lots of small teams empowered to, to go to work on it. And he goes back to that he of agile, and a lot of it was how can I say, have a small team with trust and accountability if they have constant dependency and someone else? So how do I create tools that a few people with maybe a limited skill sets or engineering skill set can do a lot of the almost end to end of their little project with most of the accountability being try like a little mini data product is How often would think about it, right? So I want to create platforms that let little mini product owners of data products exist.
Unknown 42:45
So I think that’s a piece of it. I think the other thing is, is creating, you know, creating the avenues for how you help people I mentioned this, this sort of myth of self service analytics. And so how do I help people along on the journey, one of the one of the analogies I use a lot at Target was that often because we have all this advancement of data science and all these fancy tools, you certainly into the situation where some of the smaller day to day, things get pushed to the side. The other interesting thing that we, we started experimenting, and it’s probably one of the things I still sort of ping my colleagues to target to see how it’s going was not just that you have this urgent care, but then we started said, You know what, and sort of drift on the medical analogy, we’re going to have people from all over the analytics will come and do two week residency with that team. So I am data scientist, yes, I’m curing cancer. I’m doing the complicated things over here. But I’m going to spend two weeks and I’m going to teach some of the other people maybe what I know. But I’m also going to sit down and help someone add a column to a report, I’m going to hear a question about sales to, oh, yeah, someone told me, there’s a status on sales, I can build you a card on that. And I really believe that that makes you a better data scientist. Not that, like, it’s the best use of your time, 52 weeks a year, but for two weeks of the year to sit down and sort of deal with the quote, unquote, regular people, I think, is a powerful lesson to keep you tied into how that is use day to day in the business. So I think some of those, there’s some of those very lightweight mechanisms, I think, start doing this. And again, it can’t just be like a one, one presentation and you’re like, hey, everyone’s gonna be curious to be the support mechanism to help people along that journey
Chris Weston 44:28
was really interesting conversation, Martin, it sounds like Ben’s got a really cool job,
took a look, look at some of the things that they’re doing a demo,
it’s another one of those companies that will produce hard one to dashboards and things like that. But it sounds like
where they’re coming from, right? It is about making sure it’s relevant and useful. Because God knows, God knows, I’ve seen some, I’ve seen some arcane and completely useless dashboards over the time.
But it seems like he was really pushing on the,
on the, on the needs of the user and getting to think about what they wanted and why they wanted it.
Matt Ballantine 45:11
Yeah, they’re a company we didn’t bits of work for. And the interesting bit is actually they’ve been selling quite successfully into the CMO.
And I think if you can sell a product into outside of it is the old Salesforce model of sell to the sales director, and then get permission from the IT people after
Unknown 45:35
it’s
Matt Ballantine 45:37
about I think it organizations that do that, by necessity, have got to have a closer attention to the people they’re selling to, because they can’t sell by proxy, which is where the traditional IT cell is going to spice things up systems still a thing,
but that concepts that just that very concept of curiosity, and the extent to which I think organizations do struggle with it, they struggle with it, I was talking to a guy run to small business today. And he he sells consulting services into big corporates. And he’s he’s very worried, actually, because he thinks that there’s a recession coming, putting aside whatever happens with the Brexit developer. And what he is hearing is people saying, they don’t have time for that kind of thing. Now, they just have to double down on the stuff that is hard revenue generating. And that’s a really risky strategy. Because that’s great to get you through just about, but when you come out the other side, you ain’t got nothing. And actually, I think organizations that are able to be able to work out how, in harder economic times more uncertain economic times, they are able to be able to ask questions, be curious,
that I think is, you know, that that’s the science of the companies that will come out of
depression, how many times potentially stronger?
Chris Weston 47:02
Wow, I mean, absolutely, I I can only think back to my time in an engineering company where we were very exposed to the automotive industry as we went into the recession in 2007 and we didn’t know what was coming we had no you know, not nobody knew what was coming. But in the couple of years that I’ve been there before that I’d spent a lot of my time Building Information Systems dashboards
not not a kind of sort of speed speed oka into dashboards that were very, very popular at the time still early, but more kind of control panel dashboard to say in doing ended with your businesses, what is going on in your business? Or how much of yourself how much how much work here you got coming through the door? What’s your, you know, what, really just the most important indicators to say, What are you doing, how much it is valuable, how much of it is going to coming up next next one Thomas, you’re going to sell basic stuff,
but available very, very easily, very simply and understandable by everybody. So when we got it to that point, when we suddenly hit that time where we had a massive drop off in our in our sales because the the world when when to ratchet quite quickly. We knew exactly what was hitting us we understood we could see what the problem was. And we could then make decisions sometimes hard decisions because it was closing down plan here or you know, or laying off so people they’re not decisions you want to make. But at least you know what you’ve got to do you understand what your choices are, if all you’re doing is continuing to do what you’ve ever done at hope that when the when the wave comes that your head will be above rather than buy it now I don’t think that’s a that’s a sensible choice and when when when and if something is done has happened again, it’s a one isn’t it because recession will happen at some point
I would say for any company in that in that space they need to go and talk to people who did it they went through it last time and figured out how they how they they get their business in shape to get through it is not hope is not an option I don’t think
Matt Ballantine 49:21
so. Anyway thank you to Ben for making the time to catch up with a little experiment on this one this week if you want to hear a slightly extended version of that interview goes to the website web 40 podcast.com
and there will be a link there to be able to hear a bit more conversation that Ben and I had
so that’s it time to go out and the Christmas shopping or whatever it is you’ve got to do for the rest of this week what does it oh yeah
Chris Weston 49:50
yeah I’ve got my copy this week actually and I’ve got some
ongoing at got a Christmas party with with my good friends g mark the end of the week
going at six other people as well this week is going to be it’s going to be really busy. And since I was out with some customers
last week was which was really good fun, but also quite late. I think it’s it’s my portal body counts. I can take this you know, I’ll be glad when it’s time for Christmas break about yourself.
Matt Ballantine 50:21
I am heading out to Cambridge on Wednesday through a session with a client out there at the Mahler Center at Churchill college, which is, I think, one of the best
kind of academic provided places to have conferences and things that I know it’s very, very good indeed. So I’m quite looking forward to that very early starts where to get there
and I am going to go into somebody’s tomorrow actually for showing the new year and we’ve also think got coming up in the neck or maybe it would have recorded one more show before that. But we’ve got an next session with Paul into better record the next version of the next episode of the flexible move. And so looking forward we’ve been listening to some of the interviews that Paul is done so far with a few people and there’s some interesting stuff there already so it’s interesting to hear where her mind at the moment in terms of of how that whole thing shaping up so lots on and it will probably slow down in about a week and a half, I reckon.
Chris Weston 51:22
There we go. Good. Turns out we got a plan
Matt Ballantine 51:24
I think so
Chris Weston 51:29
thank you for listening. Thank you to Ben. Thank you to Matt for being such a splendid co host we are going to be here again next week. No doubt you can catch us at Wb 40 podcast on Twitter and you can get some our website wb40podcast.com on Stitcher and iTunes and all good podcasting platforms. And if you get time please do leave us a review.
