Listen to Gerry McNicol on #TheBigLift Podcast

We are thrilled to have our own Gerry McNicol in conversation on the #TheBigLift podcast with Webtrends Optimize.

If you want to learn how to approach data, make sure you’re generating valid insights and understand how to make your data work for you – Season 5 Episode 2 is the episode for you!

The Top Points

  • 04:48 “Something as simple as a sale could have multiple definitions”
  • 12:23 “Customer data is even more fragmented than ever”
  • 14:11 “Data is becoming a revenue operations service, sitting behind all teams in a business”
  • 20:36 “Data is like a river. It should flow through your business. It doesn’t just flow to dashboards, it should flow back out to operational systems”
  • 22:40 “What to do with your marketing strategy is absolutely now!”
  • 29:06 “Who should have the responsibility for the quality of the data?”

Listen below or find the episode on Apple Podcasts or Spotify.

Transcript in Full

0:06
You’re listening to #TheBigLift, the podcast of Webtrends Optimize – the CRM solution that enables marketers and developers to maximise the ROI on their digital properties. Webtrends optimised is a powerful, feature rich and easy to use solution all delivered within a fixed price contract with no additional cost for increased functionality ever.

Gerry and I will try to unravel how to make the best use of your data. Avoid analysis, paralysis, and all without the need. To employ an army of data scientists.

So Gerry, who exactly are Distil AI? You see an agency is a software solution, or what is it?

1:24
Were a mix of a little bit mix of both really. So we built a platform called Distil, which is a customer data platform with analytics capability on top of it if you like.

So what our platform allows you to do is combine datasets from multiple different sources very simply to create a single customer view and then have analytics on top of that. But we also recognise as a business that with data there is a deal of strategy and as deal of systems required as well because data can be complicated. And also understanding how to operationalize it and get the best out of it can be complicated as well. So we offer service layer on top of our top bar software. So really, you can think of us as a data team in a box that you can plug into your business and take your business from struggling with data to being the cutting edge within a matter of weeks, like to say,

2:17
Does it have a self-service solution or does it all come with the added value service on top?

2:24
Yeah, we do have a self-service solution, especially within the ecommerce sphere, because the data sources are pretty well known. But typically outside of that businesses have a you know, a huge variety of data so we help make sure that data is stitched together to provide that single customer view.

2:46
In preliminary discussions, we discussed about avoiding analysis paralysis, and it’s a quite a well used phrase, especially over the last 10/15 years. Could you give me your view of what analysis paralysis is?

3:02
Yes. So one of the things that you find when you start looking at data is the temptation is to plug dashboards into your existing existing datasets. If you do that, you end up with, you know, similar data dashboards and data set spawning around all over the place. And sometimes that means you can’t really you know, you can’t see the wood for the trees or you’re not really looking at the same thing, versus another thing. And so, it’s actually really important to make sure your data is telling you useful information. But also, data isn’t just about dashboards or numbers. It’s also about definitions and about making sure that the thinking of the business is aligned and I think one of the things we talked about beforehand was the question, what is a train?

3:55
I’m intrigued.

3:57
So, we did some work with with Network Rail, and, you know, on the surface, and this applies to any business with anything really on the surface, you kind of think you know, what something is, you know, what’s the sale or, in this case, what’s the train we were doing work with? So Network Rail, and actually quite a lot of the meetings came back to well hang on a second, what’s the train again, and you think you know what a train is, but actually, it’s more complicated than that. Is a train a series of rolling stock, is it a single car is it something that’s, you know, being applied to a schedule? And what is it when it’s not those things? So actually, as a business, getting your definitions together, around your data is super important. We work with businesses and especially ones that have grown through acquisition and have multiple different facets, something as simple as what is a sale. It could actually also have multiple definitions because a sale could be when it’s been delivered or when the payments been made, or when it’s partly delivered. So actually, part of data and making sure you’re not overburdened by manual thinking is get your definitions as a business correct right from the start. So what is the train? Well, we can define what a train is when a sale is we can very clearly write that down. And then everybody knows and if and if you do that at the start of any analysis, you then tend to find yourself be much clearer and much more concise with whenever you can do and that’s really what you’re looking for with them.

5:31
It’s interesting use the definition of what is a sale because you might early days is being a salesman, I was always told a sale is not a sale until the money’s in the bank. But that’s obviously changed with ecommerce where people tend to pay things well up in front before they actually get the goods delivered which is quite strange.

5:48
Or we can equally be incredibly you know, some people say well, isn’t isn’t a sale is not selling to the products being delivered, you know, and the 14 days has passed and it doesn’t really matter as long as you have a very consistent approach throughout everything. Then you’re comparing apples and apples, which is important. That might seem like quite a complex, elongated experience to kind of define all the things do you have a kind of a template of the things that you need to define or is it completely different for each individual customer you deal with.

6:16
So there’s an approach rather than a template. And typically, that approach is workshop where you sit down with me, it’s not actually data people and it’s not engineers. It’s normally the finance team or the heads of the business and say okay, how do you think about how do you think about the business? What a lot of companies do is they when they do analytics is they basically analytics on the structure of the data that they have? And rather than the structure of the data as it should? Be within the ideal world of a business. And actually doing the latter is important because actually, you have data within a business which is lying around in all manner of different formats and meant for all different things, you know, often your CRM system or your marketing system or your finance system, they’ll look at data in a certain way and that that way will be to feed that operational system, but actually have collectively how the business thinks of, of data is actually how you want to be analysing it comes back down to what is the sale and the one system might think of a sale is one way another might think of it another way may make sense in the context of those systems. But actually, the approach is to go through it and say, Okay, let’s just talk about how the business works. Let’s make some definitions. And then let’s take the data and transform it into those into those definitions so that the data is always accurate against what the business is trying to achieve.

7:46
So one of the things that I would kind of come up with that is, what do you define as good data? If I use an analogy of you know, when you’re cleaning out your garage or whatever, you want to keep stuff in, you want to get rid of stuff and you want to be able to think so that might be useful, okay. And as I get older, I’m finding more and more of the stuff that could be useful, isn’t there? is no longer being useful. So how does that work out? Can you can you define what good data is in a particular business? I would say?

8:15
Yes, you can. And, but to answer your question about what you should also keep and what you should throw away is complicated as well. It’s about how you presented really, data is it can be it’s an interesting beast, it can morph, and it can change underneath you give you different answers. So I’ll give you an example. If you have a you have a list of a list of customers and one of those customer fields you have, you know, Is this person a student or not a student? Yes or no? Very simple. This is Bob, here’s a student. Yes. And then you have say, for example, all the sales records attached to Bob. You can say Okay, give me everybody, you know, how much stuff did we sell to students? You know, and we started this month. Or last month or whatever. And it’s very simple. You find all the people who are students, and yes, in that column, and then you sum all of the things that they bought, that gives you an answer. 500 grand or whatever. It’s a great result from underground of stuff to students last month, but if you’re being good bookkeeping and keeping the records, and Bob will eventually not become a student, and you’ll change Bob’s record from being a student, to not being a student, you could ask exactly the same question A year later, how much stuff did we sell in that month a year ago to students because Bob’s record has changed from not being a student, the number will be different. So Bob was clearly prolific students spend the money spend 100 grand on something, we now spend 400 grand, we’ve now sold 400 grands worth of value to students. And that’s wrong. It’s incorrect information. But yes, so actually, the way data is structured and the way data is held is really, really important. Now in that example, there I can show you very simply that you can have total your answer is going to be total nonsense. What you think is really simple question is actually not a simple question at all because it was a student is not a student and you’re literally basing your decisions on nothing. So what just throw away you is really important.

10:13
But one of the challenges I think, you know, I’ve been in marketing for half of my career and one of the challenges is, data is always out of date. That’s a big statement to make, but generally speaking, because as soon as you put some data in there, it’s historical. It may not be accurate 10 minutes down the line. Because as you say, this particular guy was a student, he’s no longer a student. Somebody couldn’t be living at this particular address, but they’ve just bought a new house. This particular person could be, for instance, in a particular socio economic class. And move to a different socio economic class. There’s lots and lots of things that are constantly inside of data. And it’s always as a B2B salesperson that I was in in marketing. It’s all about making sure that your data is as accurate as you possibly could be. So you can take action based upon that data. And the examples that you’ve put about this guy being a student, then not a student, that could throw his communication differently because you’re still talking to him as if he was student. How do how do companies try and grasp that or does your solution offer a solution to this?

11:25
You can only you can only stitch data together that you have, you know, you can bring data in and you can do as much as you can with it. The importance is not to have silos information and it comes down to the single customer view. And that means if you’ve got a record of a customer, one system and a record of customer and other which invariably, you know, most businesses do have, that you actually pull all that information together in a single place. And then that forms if you’d like the, the backbone or the business’s understanding of the business and the operations and all the customers so you’re right, someone will move and they won’t let you know. But they might if they have let you know or if there is a signal in one system that information should be replicated across all your systems.

12:09
The single customer view is I would describe it as a holy grail for many companies. It’s been part of the agenda for, I suppose in marketing for maybe five, even maybe 10 years to be able to have this single view. But there are you know, this customer data is even more fragmented now than it ever was before. So how do you bring all that data together in one place, and know that you’re actually got I’ll use the phrase good data in that set?

12:38
Well, technically, it’s not easy, but you know, that problems been solved and we’ve solved it inside our platform. You know, it’s simply bringing them all the data together from all the different systems into into one sort of pot and making sure that the system – if you think about a business, you will have all your various different departments. So you’ve got sales department, the marketing department, you’ve got the finance department. They will all have operational systems, and that could be HubSpot or Salesforce in the sales team. It could be a series of email platforms Dotdigital, Klaviyo or Braze or whatever in marketing in Xero, or Sage, whatever in finance. It’s important to take all of that data together and put it into a central repository. And it’s like the veins running through a business that data in one thing happening in one will feed the single customer view, which will feed all the other will feed on the other data silos. Have to keep them all up to date. When we started to steal actually, this concept of that was was was how we thought about it, and we were a bit we were like this actually is you know what, what pigeonholed we put ourselves into where we are marketing platform or we’re finance platform. And we know actually data is important across all of these different platforms. You know, data is technical, but it shouldn’t be really should come for finance, but it’s not a finance system. It’s not really a marketing system either, but it should be involved in marketing. Since then. A kind of a thought has sprung up, I guess, in the industry around this concept of data being a revenue operations service. So businesses are now starting to think about rev ops, where actually you have a data team or some data software which sits behind any of the revenue generating operation of the business. So sitting on the finance teams and the marketing team sitting on the sales team, and making sure that data is exchanged and utilised across all of those teams. When you think about data in a business, we need to think about the way we look at it is actually it’s a bit of everything you’re going to understand customers LTV, you’re gonna understand who your high value customers are, you’re going to market to them differently. But also you want to understand that from a finance point of view, and you want to understand it from a sales point of view as well. So it pollinate across all of it. And that’s that’s that’s where we sit and sits as a fundamental pillar. Of a business and revenue operations in the business and how to support the other functions of that.

15:03
So how can so many companies still don’t have that single customer view?

15:11
Well, quite a few companies have multiple single customer views.

15:14
That’s a Get Out of Jail Free card that one

15:18
Well, it’s not really is it because it’s not maximising. You know, in our platform, for example, you can’t send an email from our platform when you can’t run payroll reports and things. So we’ve seen marketing platforms adopt single customer views. But actually, all you’re doing there is you’re giving them more capability to an email platform. There’s no way you want to run your finance and your board reporting solely off your email platform. But most of the businesses won’t want to do that because when they actually need to improve our reporting, and now we need to ditch our email system for it or we need to improve our email system and we have to redo all our reporting. So they don’t have it because it’s, it can be complicated and you got to tie all those departments together. That’s what we do. And that’s the becoming accepted new way of how to use data inside a business to support all of those, all of those teams.

16:21
So in recent years, I think lots of companies have been employing data scientists to try and stitch all this stuff together and not only stitch it together, but actually interpret the data that is available through dashboards, etc. To think companies really need a data scientist, or is it all within the bounds of possibility to do this with your software?

16:43

Well, it’s not a data scientist, per se. You mean so you really need to do data properly. You need three, I guess individuals or roles. You need a data engineer. That’s the person who stitches all that stuff together. And there’s an expression is done. It’s done with the process has done is extract transform load “ETL” and there’s the same ETL is hard. And it really is hard. Getting data out and seeing it is a really difficult thing to do. There’s so many variations. There’s so many nuances and so many decisions that need to be made. It’s thinking data is really difficult. So firstly, you need a data engineer data scientists, they won’t necessarily have the right skills to do that. It’s a decent data engineering job. But on top of that, you want to be what you want to be in business analytics. And that is having an individual who can understand the business and how it works. And that comes back to definitions and structuring data and then producing analytics and dashboards which, which which represent that you also then going back to the data engineer, you need people who can then take that data and put it back into lots of different systems. And then then it comes to the role of the data scientist, which is actually how do we do something? How do we do science with that data? How do we make the company break away? How do we operationalize some really key insights, bringing in machine learning and AI and all of those kinds of things, and that’s where the data scientist comes. So actually, if you want to do data properly, you need a whole team of people and then you need someone to manage them and that’s very, for as a business is growing up for any business. It’s quite a it’s quite an expense, and it’s quite a quite a pond. If you figure this question to get that right. You know, it’s there’s a lot. There’s a lot to get right in getting that right. So, again, distil that’s what we thought we can shortcut a lot of that to our platform does all the data engineering does all the unification for you. We also with the service element, we come to support to some extent both the data science aspect and also the analytics aspect, working closely with the business as well. De-risks that whole process.

18:49
Just moving on a little bit. We’ve now kind of said – we can hopefully get a single customer view. We’re utilising your software solution and your people to be able to give us some view of what the data and how to interpret the data. But actually using that data, which you would say would be in the roles you just described – a data scientist. It’s trying to be able to look at things which are behaviour, trends, seasonality, churn – does that work within your system to be able to give people what I would describe “a view of the future” potentially immediate future nonetheless, but not just looking at historic but looking at forward trends and seasonality, etc, etc.

19:33
Whether it’s Distil or not Distil really, any data solution that you put in place in your business, we say should do two things; it should tell you about what’s going on, and it should help you operationalize things. I’ll give you an example. You know, if you’ve got if you’ve got a metric which you’ve used calculated, which is like your particular churn metric e.g we think if a customer engages with this on this level, and they’re really good customer and then they start to indicate signs that they’re going to leave they know start looking at things they started logging in all of that, and they started indicating a churn risk. It’s great to have that in a dashboard. You know, we’ve got 20% of our customers who are indicating return risk. That’s not great. But what we can do is we can, you know, we can look at it and we can go okay, there’s 20% we can try and do something about it. If you’ve identified a customer who has a term risk, you know, you want that information to be used somewhere. You don’t want to just look at it and sit back and go, well, there we go. You want to take that person’s record and then you want to put it into an email system when you want to send them something or you want to give them a discount or you want to get to salesperson ask them to call up and do something.

So data is like a river it flows through your business. It doesn’t just flow to dashboards should flow back out to operational systems, that you can automate the processes which will make your business more efficient. And that then also comes to things like forecasting often, you know, we look at seasonality and seasonality of individual products and product lines. And then of course they differ, you know, if you’ve got something which sells well in summer, that’s great, but it’s not summer, around the world all over the place and also, some things were selling summer some things were selling winter, having that information flowing through distribution to manufacturers and to sales and marketing is also really important as well. The data science comes in where you’re working with that stuff out but then the operationalization of it. Right What it’ll do, you will do that is really important as well because then you can do great things operate greater efficiency at scale without having hundreds of people doing stuff manually making mistakes and costing a fortune. So today, when we talk about analytics AI, it’s not just looking at data, it’s making that data work for you and and flow and then the data scientists and the data people within your business can sit back and observe and make sure that is working well. And that of course gives you opportunities to bring in machine learning artificial intelligence to really smarten up those processes to make those decisions for you as well. And then the business becomes observing data and how it’s being put to use rather than necessarily having to put spreadsheets together and calculate things on a daily basis. Which is of course a great efficiency.

22:15
Because that’s where I kind of see this, this moving much further towards which is more artificial intelligence, being able to interpret the data on your behalf by giving it some rules and being able to is that still something that you would describe? As more in the future? Or is that actually happening now other companies that are using AI to be able to forecast what to do when was their marketing strategy for instance?

22:39
Oh, yeah. What to do when with your marketing strategy? Yeah, that’s absolutely now people. People are doing that all the time. And it’s not really an artificial intelligence. It’s using patterns and trends. I think AI within software is being able to help the business understand what’s going on. Clearly, you know, again, it takes a while maybe of internal people analysing data and allowing you to AI as a tool to help you explain and work through things is important data but of course, with inside locution such as marketing execution and pricing. AI allows you to run multiple iterations and get really tactical solutions to things such as what type of copy to run you know where to place your ads as new bidding all of that, all that kind of thing. And again, data from the business side of us observing the effectiveness of those, so the AI and machine learning in the way I see it, is improving tactical execution of various things around the business. To do that, those things need to be fed with, you know, raw data and more raw data, which of course, makes that single customer view and that flow of data even more important. It’s a very fast changing landscape. It’s very interesting and it’s a war you know – it’s like a cold war as a who’s using the technology the quickest. But it’s to me is, AI is around the edges because if something does something, well, you use it but it will be replaced. So you want to, again, getting your data flow your single customer view, right is really important because that means you can plug and change tools really easily rather than having them be major engineering transformation projects every single time.

24:22
Because because I was thinking because there isn’t just one flavour of AI, say for instance, the chatbots or whatever or ChatGPT, etc. And presumably they all believe that they’re the best and obviously promote themselves as such. But there must be nuances with regard to what one does and what another does. And the other bit that really concerns me in a way is it once you using AI with your data, does your data then become part of a dataset of other people’s AI to be able to make sure that they’re doing the same kind of stuff? I would have thought that you’re opening a door to lay out some of the company’s secrets, are you not?

25:03
Probably I mean, depends who they are. Yeah, probably. I mean, all machine learning or AI needs to be trained. It needs datasets that are fed into it so they can detect the patterns and know how to respond to that. So that’s how they work they do it’s called machine learning for reason – the machine learns it’s organised itself and uses the information. Whether they share that information with other businesses or depend on the technology and how it’s done. Again, what it’s doing, because obviously the Internet is a vast amount of information and they will learn from interactions, but yeah, yeah, exactly.

25:39
Because because it’s interesting because two students both have the same question to ask for their thesis. And both of them used a AI tool to be able to make this happen. And whilst there was some consensus of thought between the two, they weren’t by any means exactly the same. They were kind of nuanced in certain ways. And some of the facts and figures were in one and they weren’t in the other, and vice versa. So it must be quite a challenge to be able to hang your hat on a piece of technology to interpret data on your behalf. But presumably, you’re saying that if you teach it what you want to do, will it only do what you want to do? Or will it use the great unwashed data that’s sitting out there as well to help formulate insight for what you should do?

26:33
Well, exactly. If you’ve got enough data to teach you what you can, if you’ve got enough data to teach it, teach it on your own what you know is right. But I think this is it’s very interesting, very interesting problem because this machine is learning you’re not programming it, so to speak. You’re not saying you know, if you see A say A – you’re giving it a bunch of data and asking you to come up with some some stuff and if you fed it, Wikipedia, it’s gonna think it’s just gonna recite Wikipedia back. I think there was some, you know, people look at, I guess, some of the, the AI systems we have, and they think, Well, why can’t these things be as smart as check up? Why can I ask, you know, Siri, these questions and it can give an answer. And of course, it’s perfectly possible that Apple could toggle that technology into Siri or Amazon into into Alexa. But then those things probably just start spouting nonsense – well, not spouting nonsense, but it’s like talking to Dave down the pub. Dave has heard some stuff and he’s just gonna recite it back again. You know, there’s there’s no guarantee what they’re going to tell you is right at all. Yeah, it’s very interesting problem. And it’s not these things are just smart monkeys or very smart monkeys or very smart humans. They still are reading stuff and composing their responses to it.

27:49
But as a business, how do you nail your colours to a particular mast?

27:54
So that’s what I mean I don’t think technology at the moment exists for AI to run a business. But what you do usually use it tactically. just to talk about you know, generating copy, for example. So rather than asking a human to write, copy, you can ask AI to write copy for you. And then you can ask a bunch of variations in how to go so you’re using the AI at the edges to execute campaigns or execute decisions but equally, you’re monitoring it. We’ve been using it for ages you know, when you use when you place PPC ads, you ask Google to try a bunch of variations and then zone in on a better one. So it’s just using, you know, it’s the same sort of thing and increasingly sophisticated versions of that. You’re not, you know, just getting an AI and saying, you know, run my business. That would be very nice. And it makes amazing decisions, but we’re not there yet. And we have a menu still the pipe data to it.

28:47
Yeah, it says, it seems like it’s going to be quite an interesting next few years to see how this settles down, if it might do in two years. Who knows! One of the things that I think is a fundamental question that I would ask is, who or what department should have the responsibility for the quality of the data? Because, obviously, in a B2B organisation, sales say, well, this data is wrong because marketing sent it to me and marketing say, No, the data is right, you just haven’t been filling out the CRM systems and it’s always been this kind of battle between the two. And it’s now we’ve got this more formalised process around data. Where do you think that responsibility lies?

29:31
The number of, in particular, sales systems I’ve opened up and picking up a rock and seeing the creepy crawlies scurrying out of there. Sales guys usually don’t keep great records, but that’s because they’re doing sales rather than just you know, being very nerdy and typing everything in correctly. And that’s probably correct. The truth is, if everybody understand the value of data and data feeds back to them, then people will see the value in putting good data in. And it comes back to that, you know, Rev Ops, that data team structure. They’ve got to have some ability, some agency in being able to enforce data, data rules, pull out, my data is bad, and then you go back down to the teams, I guess, to the leadership is going to be made clear that data is really important and let’s collect data correctly. But you know, companies are structured, you know, with boards. Boards, like to look at reporting and if the data feeds the reports for all individual, you know, team leaders, and they should make sure their data is collected and is presented in a good way. Data hygiene, it’s important for everybody in the business and in everything they do.

30:43
So, let me take that thought a little bit further, because I think that many many moons ago, the marketing director or marketing never had a seat on the board. And it’s only a relatively recent, I’d say the last 10/12 maybe 20 years but that’s become more frequent. We’ve got VPs and marketing, etc, etc, as CMOs, but there doesn’t seem to be anybody that’s sitting there with the data responsibility which seems like a huge responsibility in itself, to make sure that what they’re doing is the right sort of thing. Do you think there will be somebody who’s sitting on the board?

31:20
Well, there is role now is quite becoming more common or Chief Data Officer right. So this is becoming more, more prevalent, but, you know, larger businesses, really. To me, if you if you had to put it under someone, it will be under the finance operation, the finance side of things, because ultimately things tie down to revenue, you know, if you, you can tie everything down to revenue and if you get the revenue numbers, right, and you get the number of sales, right and the number of customers right and all of those kinds of things then the rest of it, it’s like a really solid foundation. And finance people tend to think of the business in the correct definitions and structures of data

31:58
More rigid, let’s put it that way

32:02
More rigid but how do I explain? If a marketing person – will be thinking in terms of you know, subscribers in a contact list whereas a finance person will go Oh, actually, I need to report on active customers, and how much money I’m making and all of that. So actually the way they think is the business model that drives the business. And everything can actually hang off that nicely. And if you actually enforce it – the business everyone’s actually thinking in the same, the same terms. It’s really the one where they may not be using the data to marketing is a really powerful use of data. Really as operations is a really powerful use of data.

32:44
In product development is

32:47
Yeah, it’s super important. But actually, from a business point of view, the business model behind the business is how the business works. If you can get everyone thinking like that, and the data can be structured like that, then everything becomes a lot easier. Quite often as well, data falls under the CTOs responsibility. But data is not really although it’s very technical, and there’s engineering in it, actually, the more important thing is it’s representing the business and the decisions of businesses needs to make. It’s not a technical exercise that there is technicality to it. That’s my view, but there are many others.

33:25
I’m sure there will be there’s there’s never one particular view that everybody attests to, which makes life quite interesting. Gerry, thank you very much for your time. I think this has been a fascinating conversation and will continue to be a fascinating conversation as artificial intelligence rolls out. Furthermore, and I’m sure it won’t be just talking about the data although that will be the fundamental platform of which everything is built upon but the effect of artificial intelligence marketing on sales on, you know, stock control is just phenomenal. And it’d be interesting to see where it takes us in the next year. Well, even three to five years I think things will change dramatically.

34:08
Three to five months probably!

34:11
Well, things do move quickly. I’m not hoping they don’t move quite that quickly in being able to do things because the whole idea of managing that change. You can’t continue to be in a state of flux all the time. You’ve got to be able to start doing something for a period of time and in the past that may have been five years that you’ve kind of set your business plan out. But life is changing at a pace now it must be very difficult for those people that are running businesses to understand how and when to use artificial intelligence to be able to make things work for them better.

34:45
Yeah, yeah, I think so. And I think if you think it’s the fundamental thing about how to maximise it is – we have a data maturity model, which we work to here, which is you know, firstly your data scattered all around and you bring it all together. Once you’ve done all the basics, there’s some basic hygiene you need to do in your business which is actually listed as getting this you know, unified set of data upon which the whole business can run. When you’ve got that, using these tools, these new tools will be there, artificial intelligence will become a lot simpler. One of the things we’ve we enable, or we go through with our customers is migrations from one email platform to another, you know, someone will start off with MailChimp and then they’ll go to Klaviyo then Dotdigital and Braze as the business grows and have different demands, and every single one of those if you don’t have all your data together as a major engineering project, but if you do, it’s really easy to plug and play. The same is true with using these these new artificial intelligence technologies and new connectivity between businesses – if you’ve got your data straight, then actually you can try them out and I think that’s to me is key is making cost of experimenting and cost of trying as as low as possible. You know, if every single time you want to try something, it’s a major engineering project, you’re never going to try it. Because the risk of failure is too high was it’s really easy – well, it doesn’t matter if you fail. If you fail, you’ve learned something and the cost wasn’t very great. Anyway, so that means you can be more responsive to the new stuff as it comes out.

36:16
Gerry, thank you very much for a fascinating 40 minutes or so it’s been it’s been a great insight, and I’m looking forward with a little trepidation with regard to what the future holds for us. But thank you very much for your time.

36:29
Pleasure. Thank you very much.