Episode 9: Revolution or evolution

AI Transcription:

[00:00:00] Ben Culora: Hi. This podcast is brought to you by Practicus. Practicus is a recruitment, consulting and advisory business, specializing in change and transformation. We hope you enjoyed the podcast.

[00:00:13] Barry Panayi: Hello. Congratulations. You’ve made it to the final episode of the dated files. In this episode, we’re gonna look back at what we actually started this whole affair with was five years ago, and some predictions for the CDO role and data will then take a canter through the wonderful guests we’ve had on this show and pick out what we think some of the really important areas where we agreed, where we disagreed and some of the things we learned.

And then in the end, maybe, if we fancy putting our necks above the line again, we can think about what the future of this looks like in another five years.

[00:00:50] Tony Cassin-Scott: So thank you very much for listening to our series of the data files, and we’d like to thank Rachel Purchase Lydia Collett, Steve Green, Brian Price, Helen Crooks, Pete Williams and Rosh Awatar. And finally, also, I’d like to thank our producers, Ben Culora,, Lawrence Hill, and our fantastic soundman, Harry Radcliffe.

[00:01:14] Barry Panayi: Yeah, thank you very much guys. Behind the curtain of the data dungeon, you’ve done an excellent job.

[00:01:19] Tony Cassin-Scott: Five years ago, Barry, we came up with our thoughts about what five years in the future would look like and, and here we are five years in the future.

One of the things that we said is, does the CDO role need glamming up?

[00:01:32] Barry Panayi: I do remember saying that, and I do regret writing it down so that we could dig it up again after five years. I think there was, uh, very much at the time, the CDO roles in its infancy. It was, banks, insurance, other FS dealing with regulation, talking about governance, talking about data management.

And it was a turnoff to a lot of the exec who saw it as a necessary evil. And I, I trying to justify why I said that, I think that’s why I said it. What I do think perhaps we’ve done as an industry and I, I put my hands up as part of the, uh, fanning of the flames here. Is we did hype it up quite a lot, and we’ve got to a place where everyone wants a piece of machine learning or AI. No one wants to invest in the enabling capability. I’m summarizing of course, but perhaps we’ve shown a bit too much value too quickly in tactical ways to make it sparkly and nice and perhaps the harder, slower route, would’ve got us to the panacea of a data driven organization quicker.

Now, I don’t know, Tony, what notes you took down in your burgeoning book of notes I can see in front of me, but certainly many people were talking about complications of getting business cases for data management type initiatives, whether it be metadata or quality or cataloging, and people are being asked by their bosses, What’s the ROI and so on.

I can’t help thinking that my urge to hype things up has perhaps been repeated elsewhere and people have come to the conclusion that maybe that stuff you just get away with as little as you can. But actually we are hearing from our guests that, it needs a bit more investment, a bit more thought and should be thought of as an enabling capability and not something that returns value.

I dunno what you think. I think we’ve been very successful with blaming it up perhaps too much.

[00:03:36] Tony Cassin-Scott: I think there’s a pay to play element here. I think the sort of go slow to go fast approach you mentioned earlier is logically correct, but I think it’s probably unappetizing. If I’m on the board and I’m going to invest in this, I want to see some quick returns.

And telling me it’s gonna take time is probably not the right answer at this point in time. Yes, I will understand that in time things will get better, but if you’re gonna take a year or two just to develop a data platform that may or may not provide some value later on, isn’t what I’m looking for. So I, I do think the need to hype is really important.

I think what’s missing is the fact that if you wanna make it sustainable, , this will require investments and I think that conversation hasn’t necessarily happened at the level, which I think it should have done.

[00:04:26] Barry Panayi: And, and that’s manifesting itself now, of course, in people talking about ML ops, pipelines and products, which are all about hardening, to use a, a IT term or industrializing stuff. And I think that is maybe our second bite of the cherry now as an industry to, to get this done. Talking about product, talking about ML ops. So, Okay. That was a mixed, uh, prediction. I would say victims of our own success, if we wanna put, er, well, a slightly false lens on it that makes us look better than we are.

There is something we said five years ago, which I think during the course of this series of podcasts, has manifested itself to be true, which is the importance of community and engagement, not just within data teams, but across the business. And we’ve heard from a number of people that have talked about, we have cross functional teams from the business.

We have product owners who maybe don’t do data. If we go all the way back to Rachel in the first episode, we also heard from people talking about hackathons, comms specialists. And just getting the message out there in a community way, not a push from the center. And I, I think that that was a fairly solid prediction we made that success will come in communities of data curious people as opposed from just the data teams. And indeed, we might get onto this at the end, what is the data team of the future?

[00:05:50] Tony Cassin-Scott: That’s a very good point. As you were speaking, I was, it’s a term which I now hate to use, but I used it very liberally five years ago, is about data democratization.

[00:05:58] Barry Panayi: Is it the data or is it actually the capability to use the data that’s democratized?

And perhaps we were a little bit, lazy. The industry was lazy with saying, we’re gonna democratize data cuz that’s not actually the point, is it?

[00:06:09] Tony Cassin-Scott: No, it’s not at all. It’s about, uh, getting more value outta the data that you’ve got and the tools and the methods to do that. So that is what I suppose as a data function or as a data entity, we are striving to hence all these datathons and what have you.

So I think it’s about embedding and training and educating people in the use of how to get more value out of existing information assets. Which is very different if you look at it to say something like IT or finance, where that isn’t, I’ll use the term democratized to the same level cuz you still have those fun core functions that provide enabling functions and reporting functions on those specific topics.

I suppose the question around the data function is, is it an entity in its own right or is it a core enabler for the rest of the organization to benefit.

[00:07:02] Barry Panayi: That leads me to another thought. I had the amount of inconsistency in this series that we’ve had on. You talk about job titles. Everyone’s had a pretty, a different job title, but if we forget about that and talk about the number one person in charge of data for an organization, they’re called different things. Let’s just call them number one for now. And then under that number one, there’s no consistency as to what is in there and what is out of there. We’ve seen, functions that have engineering, platform, data science, bi and vis. We’ve seen examples where the engineering’s not there. We’ve seen examples where the data science isn’t there. We’ve seen examples that are all about product and cross functional. I’m wondering is that good or not good? Does it depend on the organization or does it need to be codified? You know, a, a data function should have as a bare bear minimum. These things in and if you step even further back, have we got our stuff together as an industry to help people understand what is in a data function? Because it seems five years on, we haven’t really defined what a data function is in the same way. Sorry, IT and finance and HR have, and we are a support function in data. Right?

[00:08:21] Tony Cassin-Scott: I, I think that’s right. We we’re still on a journey. We haven’t reached the destination and we’re not quite sure what the destination looks like.

If you were, you mentioned finance and IT they are far more mature in defining what their destination is, what their purpose is. And although it has probably gone through more recent changes in the last 10 years than finance, and that’s probably cause finance is more mature, data is way behind the curve in terms of those level maturity. And therefore the question is, is it a function of the lack of maturity of which it is in, or is it, in fact it’s an implicit future capability that doesn’t need to stand alone at all, but is implicit in the way that people operate.

[00:09:03] Barry Panayi: I think that’s a good challenge, uh, to me because is it that we are not mature or are we mature and it is everywhere? Not to say it’s nebulous, but a goal would be that products become more important. For example, and we did hear both from Lydia and from Rachel about the importance of squad. That deliver data products and those squads are spun up based on the domains you’re operating in. And I do wonder if that is a trend that we’ll see catch on more the chief product officer pulling stuff together.

At the moment, the chief data officer maybe has a head of data product that is fulfilling that role, but I wonder if it’s all just product in the end. Uh, and I did think that through a number of our conversations actually, we, we may build up the foundations. Pete talked about it in about the three Es that you’ve gotta build that enabling capability all the way through to exploiting it. And I wonder once one gets to the exploit end of things. If that then means that that exploitation, again, another word that probably we should think of a a better way of saying it, but you all know what I mean, that those product teams are the ones that are drawing on that and the businesses and trying to push things out from the center.

We talked five years ago about hub and spoke. Maybe that was just too crude to term. Actually it’s building the enablers and letting people decide how they arrange themselves around them.

[00:10:27] Tony Cassin-Scott: I think that that’s right. Going back to what we said five years ago, we were talking about. The data function being a center of excellence if like a factory type function.

But actually I’ve got a feeling now talking to all the people we’ve spoken to and also just outside of these podcasts, it’s an enabler that allows things to happen rather than a function in its own right. And what I mean by that is there’s a level of education that needs to happen, but once that is in place, surely the data function, if that indeed what it’s still called, does data, does data become a, a defunct name now, but that enabler and the use of the exploitation, I know you don’t like the words, but in terms of getting more value out of the information that we hold, is that what the enabling function does? Does it allow people to get greater insight and greater value from the information they have to propel their business along? Therefore, the whole point about data governance, data management are still implicitly there, but they’re built and by design. You don’t have a special function that looks after it.

As in, for example, you don’t have a special function today that looks after the water services in your building or the facilities management. They use third parties to provide water. It’s a commodity. Are we going to a world where these are commodity activities and the expertise in house is actually how to get the best outta these commodities?

[00:11:53] Barry Panayi: You might be onto something. I, I go back to the chat we had with Steve Green in the leadership episode. He talked about one of the main things of his role is distilling information into stories and how his background as a historian has helped him do that. And I wonder is there a future where some, what we now call data functions are actually the storytelling functions, But some may be the data risk management functions, if that’s what’s important and the exploitation as you put it of that data sits elsewhere, but the pulling together of the value and the inspiration and the what’s possible and the underlying infrastructure of data still sits in the center, but the use cases sit outside. I got the impression from Steve that he’s very much the, he lights the blue touch paper and he then he tells his stories to, to make sure that there’s that cross pollination.

[00:12:49] Tony Cassin-Scott: Well, along those, I mean, Lydia mentioned about the explore exploratory mindset. That she wanted to encourage within organizations. So which would, would tie in what you just said there. So I, I think it would providing those people with that platform to do it.

So for example, if I use an old example now, sort of search engines, Netscape, Google now. Before those existed, there were specialist people who worked in the library function that looked up information for people. They were specialists, they were trained, they got certification, what have you. Well, that doesn’t exist anymore. Everyone’s their own self-service information. finder now. Everyone does it from the mobile phone to anything.

So what does that space look like in the data space in the future? I think that’s potentially where we’re going. Someone’s done all the hard graft about automating search engines, for, for, for text. What does it look like in the data space? Or am I dreaming there?

[00:13:45] Barry Panayi: That would answer some of the tasks, wouldn’t it? I mean, I, I go back to something that Helen mentioned, which was the tenure of CDOs being quite small, and I wonder, is that because it’s not well defined, what the success factor is of that CDO in that organization. Is it tagging up all the data to make sure it’s discoverable and people can do the bi, which is what I read your point to be. Is it making an absolutely hot team of data scientists who can do the best models ever in the middle and push them out? Is it managing the risk? And I wonder if there’s just such a huge range of options that the exec or boards of companies think that they need all of it, but maybe they only need a bit of it and then they get someone in who focuses on the bit that they don’t think is so important.

[00:14:46] Tony Cassin-Scott: I think. I think that’s right. Um, it’s interesting what you just said there, cause it, it sounds like a very fragmented landscape we’ve got here. If I look at IT, which I would argue was equally fragmented a few years ago, people and they were spending lots of money repeating the same thing and getting it wrong and what have you. So there are lots of methodologies that were created to manage that cost as it was seen at cost rather than benefit. Are we at the stage in data where we’re seen as a cost cuz it’s so fragmented? Or do we have a more joined up message and offering to deliver to the business.

[00:15:26] Barry Panayi: This joined up piece is something that, there’s something in that, uh, we look at what Helen said, where she semi glibly said that the D CDO stands for diplomat.

Um, I know out there, there are other data podcasts. Um, They talk about talking to people and tea cake and that, that being the most important thing. Stakeholder management is a management term for it, isn’t it? Rachel talked about it in hers. Pete talked about it. I wonder if there is that tying together and making sure everyone understands what the rules of the game are. So what are the absolute bare minimum. Data, things you need to do before you can go to the next level.

Now, we may have gone a bit too early with delivering value, but we had to do that as an industry, otherwise you wouldn’t even have got the foot in the door. Now there’s an element of, as the data leader one has gone all around the organization, you go back to the board or the exec or however you’re structured and say, there are some common themes here. No one is gonna pay for all of this, but we need it. They’ve all agreed it’s important. But it needs to be funded and don’t ask me to give you an ROI on it. That is a conversation that I’m not sure I heard in this series has happened in a mature way within exec and a board cuz the exec and the boards want bits of it and don’t understand perhaps, or it hasn’t been articulated properly, or maybe they do and it’s not that important ,what you need to do to enable that.

[00:17:09] Tony Cassin-Scott: I think that’s right. And what I, Just to, to frame that a bit more. I don’t think there’s been any business architecture design around that. The glue, or I’ll call this the glue about pulling all those things together doesn’t exist, in a single place. It is dissipated, it is distributed and is very disparate in how it’s managed as well. And I think the role of the, the, the data lead if you like, they’re often filling gaps, pulling the different threads together cuz no one else is. I would say before that data lead the on it function was probably the closest to pulling those together. But again, they have a different slant. They have a less commercial slant, whereas I think data side have a more business led commercial focus and therefore probably more appealing to the business community. But still, they’re the ones that are trying to pull together these very disparate threads.

[00:18:02] Barry Panayi: I, I go back to, uh, the ethics and governance episode that we had with Roshan, where he talked about, I think quite smartly his tactic of, drawing a line from the AI and ML that people want, the shiny use cases, back to governance, uh, and explaining that in a story. And I think that’s pretty, pretty strong and we could all learn from that. It’s worth pulling out.

[00:18:27] Tony Cassin-Scott: And I think that that also illustrates me another point, which, which hasn’t always been clear as that line of sight. Between activity and value. So often I’ve come, I’ve worked with several organizations where governance and data management is seen as a necessary evil, something I have to do because the value hasn’t been articulated. Say, if I don’t do that, this is gonna be the outcome. It’s not gonna be very good.

[00:18:51] Barry Panayi: Of course in FS, if you don’t do it, go to prison. Exactly. And that, that’s a pretty good, the Smcr regime does, does help with that. But going back to that, Rosh’s chat on that, I remember him talking, saying, well, if you just take data away from data ethics, it’s ethics. Same thing. If you take data away from data governance, it’s governance. The same thing. Would we even, I’m not gonna put words in Rosh’s mouth cause I’m not sure he would say this, but data analytics. Data science. Why is data in there? Is it, is it just analytics? If you’re in the business, just some analytics that you can do is have, by having the word data there, have we somehow put dibs on, Well, that’s us, not you. And is it time to talk about analytics and science and governance of which data plays bit bit of that. We don’t talk about, I’m gonna do an IT PowerPoint for you. That’s true. Er other, other slideware programs are available.

[00:19:48] Tony Cassin-Scott: I think that’s absolutely right. It’s a bit like the I in IT isn’t it? It’s kind of all technology based. Uh, it may or may not involve the movement of information from one element to another. So it’s technology rather than IT.

[00:20:01] Barry Panayi: Are we talking ourselves out of any sort of job in data here? Um, are we saying that it’s everyone’s business or not? Reflecting on what we’ve pulled out of all of these things or how it should be everyone leaning in and everyone understanding it, and analytics isn’t just for the data team, and ethics isn’t data ethics, it’s just ethics. What does that leave like what, what are the threads over these episodes that really point to the need for a function? I’m being a bit provocative here. Do we need this whole bloody thing?

[00:20:34] Tony Cassin-Scott: I think the short answer is yes at this moment in time.

[00:20:37] Barry Panayi: Phew, that’s good.

[00:20:38] Tony Cassin-Scott: Yeah. Okay. You’ve got a job.

[00:20:39] Barry Panayi: Thank you.

[00:20:40] Tony Cassin-Scott: Uh, but. Going forward, I don’t think you do. Oh, sorry.

[00:20:45] Barry Panayi: Um, what a way to find out.

[00:20:47] Tony Cassin-Scott: I think, I think we’re at a moment in time, I think we’re filling a gap that exists, a real gap that exists. I don’t think the, both the capabilities or indeed the technology is at a mature enough level to benefit the business, and therefore you need a guiding hand.

I think the data function at this moment in time is the guiding hand to provide that level of support and insight, and I think we’re a few years away from that guiding hand being removed. So for the foreseeable future, so until my retirement anyway, I think it will be required. But 10 years time, less so is my prediction.

[00:21:28] Barry Panayi: Less so as in it goes away or it becomes smaller. The leaders that we’ve had on talking specifically about leadership, talked about a quite important, but generic set of skills that they see as the most important. They talked about influencing and diplomacy and tying together and funding and business cases. Is that actually the future and are people doing that already And we’ve got nothing to worry about cuz the real data leaders have clocked on. If so, I’d done some rudimentary back of a post-it analysis. 70% of our guests think that the CDO or equivalent has to have, been a data expert and 30% do not. There was only one, uh, bit of both answer from Rachel, which I thought was an excellent answer in that you can’t do all of data. So why would anyone say they have. Is it the case that the CDO role on or similar will mature to a more generic leadership role? But at the moment practitioners are still needed cuz it’s still quite early on?

[00:22:40] Tony Cassin-Scott: I think that’s right. As I say, it’s a point in time on the maturity curve. I think in time it will, the role of the data leader becomes more of a commercial job, because the technology in place will take care of the governance for you because it’ll be built in by design. So I think there are a lot of, there will be a lot of automation going ahead in the future, and the role will be to make the most of that technology rather than design the technology that will be a commodity, is my prediction going forward. In fact, there’s already some, some movement out there already, so I, I think that the function become smaller and more of an enabler than a end point delivery function, which it’s much of the time it is today.

[00:23:26] Barry Panayi: Looking back over these episodes, we didn’t talk much about technology.

[00:23:31] Tony Cassin-Scott: No.

[00:23:32] Barry Panayi: Why? I mean, I don’t really know much about technologies. That’s probably why I didn’t mention it. But, uh, Are we seeing it as a commodity? Do we still believe that tech sits outside of data or data sits outside of tech and it’s not the same? Cuz I was on my high horse five years ago. Data is a thing. It should never report to the cio. It’s nothing to do with technology, was my vibe back then. I still think it’s completely separate and should be. We didn’t really talk about technology at all. Is it complete red herring? No one, no one mentioned it.

[00:24:03] Tony Cassin-Scott: Well, it’s because it’s there. So if I go back 25 years, you couldn’t do what you’re doing today.

[00:24:10] Barry Panayi: I was far too young.

[00:24:14] Tony Cassin-Scott: Far too young. No. The, uh, the, the technology hadn’t caught up with the theory. It didn’t exist. You, you couldn’t, Everyone, the theory has been around for decades longer. The technology is relatively recent, so go back 20, 25 years, it wasn’t there, so you couldn’t do what you’re doing today. So techno Well, the reason we haven’t spoken about technology is cuz it’s been, I would argue it’s taken for a bit of a granted, uh, sorry, it’s been taken for granted that they’re already there and therefore you don’t need to worry about it.

25 years ago as a statistician, you’d probably be working on the technology in order to enable you to perform the statistical analysis that you wanted to do. Mm. Hence people like, uh, the guys at Rothamsted Experimental Station, the Genta and Li, they had to write their own applications just to be able to do the analysis that they want to do, but that it was like they had to do all the hard graft just to do the stuff which they were originally employed to.

I think where we are now is a lot of that technology is in place and therefore we haven’t mentioned it because it’s there.

[00:25:21] Barry Panayi: Yeah, I think that’s good. I was just racking my brain for the why, what haven’t we talked about on this series of podcasts that we thought we would of and technology. I meet some data leaders who are obsessed with talking about this platform and that platform and this software and that software, but what was quite refreshing is, I think we may have got that five years ago. What data quality tool are you using? Um, Even now we didn’t get a, how are we doing ML ops? Are we using tool layout, tool B? And I think that shows some maturity in that we, we see the technology as somewhat, ubiquitous. Some tools are better than others. We’re talking more about in this series of podcasts, how you make things stick and how you structure yourselves.

[00:26:04] Tony Cassin-Scott: Yeah. I, I think that’s right. Five years ago would’ve been about which the best tool for this job. And, and there were a fewer tools around, they were less mature. So the reason it’s not a topic today is because they’re all, they’re all much of a muchness. I mean, some people will argue with that, but I, I would say they’re probably good enough to do what most people want to do.

[00:26:23] Barry Panayi: Talking about. The skills of the data leader, the CDO whatever we wanna call them. There was an interesting point in the Brian Price episode I thought, where Brian made the argument that yes, the CDO or equivalent should be pushing this message out and educating and so on, but where’s the pull from the board?

When we’re talking, not the exec actually, uh, uh, Brian meant the board. So within the non-executives and the executives who’s asking those questions of the exec, where the exec then think, Oh, I think we need a data expert for this. Uh, I, I think that is a very good point because we, we can’t expect a function to be pushing, pushing all the time. There does need to be those pull and quite pokey questions, the right pokey questions coming from the top. We didn’t hear much of that in the podcast apart from in Brian’s one, but I can’t help thinking as the role matures, as the industry matures, we will need to see some more of that.

[00:27:27] Tony Cassin-Scott: I think that’s right. I’m just wondering if that was the case, 30 years ago, if I look at it or there was nothing or whether it was seen as a subfunction of finance as it was. We had the IT direct report to the cfo, et cetera, and I’m not sure those questions were on the board about the value of IT. So fast forward to now, I think it’s the same issue. Mm-hmm. , I don’t think many of the board members ask those questions maybe because it may not be in their sphere of knowledge. Um, it all depends on the company, of course, but Brian’s right, it’s, there is a lot of push and very little pull. I suppose the question around that is, why and how do you remedy it?

[00:28:12] Barry Panayi: Is it a function of time? And are there enough data leaders who have been in situ and made enough impact and have enough broad leadership capability to demonstrate their value to a board or, is it the boards don’t recognize it as important.

[00:28:29] Tony Cassin-Scott: I think that’s probably a function of where we are today. As in, I think we have a lack of maturity.

I think we’re on a journey of maturity and that probably permeates across from top to toe of an organization. So the reason those questions aren’t being asked is because they’re probably not in line of sight of many people, so they don’t understand.

[00:28:51] Barry Panayi: That’s one to watch for the next five years. Well, I’ve thoroughly enjoyed the podcast themselves and I think we’ve put our guests on a pretty, uh, intense spotlight. We’ve asked them some pretty hard questions.

I wonder if it’s time for us to put our necks on the line a little bit and make some predictions for the next five years and maybe Practicus will ask us back to see what idiotic statements we’ve made.

[00:29:17] Tony Cassin-Scott: Yeah, let’s, let’s make idiots of ourselves.

[00:29:19] Barry Panayi: I don’t have to try too hard.

So what, what word do we think the future is? If we summarize, we think data is a thing, it will continue to be a thing, but maybe there’s a retrenching on the enabling capability and maybe the education isn’t on, this is what AI and ML can give you, but more on these are the foundations you need to deliver the AI and ML. I think that the AI and ML algorithms will become quite commoditized and actually the advent of low code, no code and so on will mean it’ll be more accessible. Compute will become cheaper, and it will be those enabling capabilities and the ML ops pipelines and the engineering, they’ll become important.

And I’d say my second bet for the next five years, is the rise of data product. We heard about it with Rachel and Lydia especially. I think these product skills will become very important, treating these data pipelines and data products as actual products, whether they be internal or external, it doesn’t matter. So I’d say data, product and retrenching on the foundations. That doesn’t mean rebuilding data lakes and SCVs. It means the, the data management foundations. That those would be my two bets that we’ve kind of come, will come full circle, I think, in the next five years. And there’ll be an awakening on, on the enabling capabilities needed.

[00:30:41] Tony Cassin-Scott: I think I agree with you on that. And as a consequence, I think the data function as it stands today will be very, very different. As in it’ll be smaller, be more of an enabler, uh, a selector, if you like, of what those, commodities are for the specific organization. So I think the role will change from a build to a buy type function and consequently will be far more commercially aligned than it is today.

[00:31:10] Barry Panayi: Hmm.

[00:31:11] Tony Cassin-Scott: Out of the people that you would’ve liked to have, that you would’ve liked to interview but haven’t, they can be dead or alive would be obviously hard to interview them when they’re dead. But anyway, metaphorically, who would you like to interview and why?

[00:31:25] Barry Panayi: Well, I dunno if I’d, uh, be smart enough to interview them, but I’d certainly like, like to ask Warren Buffet if he thinks this data thing’s a flash in the pan or not. Cause certainly he’s made the right technology investments in the past. He knows what companies, what good companies look like. So I think it would be a, a ringing endorsement or a death nell to the, the data function if it, if we could understand in his mind, the way he looks at the world, is it a differentiator to have nailed this or not? So I’d say Warren Buffet just cause I think he’s fascinating human being, but also he’s the oracle, isn’t he?

So he is. Yeah. I think he will be able to tell us some stuff. How about you?

[00:32:08] Tony Cassin-Scott: I think again, another contemporary figure would be Edwina Dunhumby. She started this gig off many, many years ago. Has it gone the way that she’d expected and what would her future view be of where the space goes?

[00:32:23] Barry Panayi: Yeah. She’s certainly one of the big pioneers in this space. Good. Good answer. Thanks, Tony.

Well, I think that’s it, or a wrap as they say in these sorts of, uh, situations.

Thank you all for listening. Tony, thank you for sitting opposite me,

[00:32:37] Tony Cassin-Scott: And thank you for sitting opposite me, Barry.

[00:32:40] Barry Panayi: Till next time. Bye.