Ben Barrass is the group head of data at Centaur Media, a provider of market intelligence and publisher of publications such as Econsultancy, Design Week and The Lawyer. He is responsible for planning and execution of the data strategy across all the digital subscription products the company runs. With over 19 years of experience across a range of digital business operations, he is a marketing technology expert.

Coming up in the episode:

• Ben gives the perspective of a data professional in a central team within a large organisation containing different businesses that service very distinct audiences. (1m 35s)
• Moving print-based publications to digital successfully (3m 17s)
• Standardising and harmonising services, reports and removing silos of knowledge (4m 40s)
• A new way to think about what data we collect and whether it is worth sending an email to the people on your email list that never open it. (9m 23s)
• How to use data to help people that have been in the organisation for a long time (14 m 3s)
• The importance of building data collection in a project from the outset and the dangers of relying on ‘phase two’ of a project (20m 52s)
• The increasing importance of data literacy in an organisation (25m 53s)

- Hello, and welcome to The Red Olive Fibonacci Podcast.

The podcast all about data and analytics,

where we hear from leading specialists

and get their take on the industry.

I'm your host, Nicky Rudd.

Today, I'm joined by Ben Barrass,

Group Head of Data at Centaur Media,

a provider of market intelligence

and publisher of publications such as, Econsultancy,

Design Week, and The Lawyer.

Ben is responsible for planning

and execution of the data strategy

across all of the digital subscription products

that the company runs.

He has nearly two decades of experience

working in a range of digital business operations.

He's a marketing technology expert

and a specialist in using data

to transform digital processes

in both large and small business environments.

Our conversation covers the importance

of building data collection in a project from the outset,

why it's crucial to really understand your audience,

and a new way to think about what data we collect

and whether it's worth sending email

to the people on your email list that never open it.

What are we waiting for?

Let's go.

(upbeat music)

- I started off as a service side developer,

a back-end developer back in the day,

and got more into analytics,

and controlling data flows between systems,

and various sorts of integration work.

So when I left,

worked in a small company as a temp director,

pretty much every single aspect of the business

was data-led in some way.

It was effectively an affiliate marketing organisation,

so all the revenue came through digital tracking

and everything else.

So, my journey into data was more led by

the need for me to make stuff work.

So, hooking up various types of systems,

and then making sure things spoke to each other,

and the clients and suppliers and contractors

all knew kind of what we were billing

or why we were billing,

and make sure the right messages

went out to the right people.

My experience then mainly got involved with

making sure everything was hooked up and working together,

and the various robots of data

were connected into the right systems.

And I moved on to Econsultancy

very much along those same lines,

joining their marketing team to help really drive

their marketing segmentation and messaging.

I remember Ashley, on my first day at the job,

one of their events, he announced me to the room,

one of his speeches he was doing,

they've made their first data hire within the organisation.

And I think it was telling for Econsultancy

that they saw the need to put their first data hire

into the marketing team.

I wasn't joining finance or sales or backend devs.

It was very much focused around the marketing efforts.

And from there, I've moved from marketing

into sort of core central systems,

and then trying to organise

what is effectively a transitioning publisher

from an old sorta print model,

into a digital transformed online content

subscription business,

and all the associated systems and services

that go along with that.

- What's the steps that Centaur has gone through

in moving from that sorta traditional print publishing

into a fully digital online publisher,

and also events and trainer as well, I think.

You do training as well, don't you.

- Well effectively, all of those businesses,

the events, the training, you know, the webinars,

the reporting, all comes from valued content.

So, whether you're a publisher

that sticks words out on print,

or you're a publisher that sticks words out on the screen,

or talks words at events, or, you know,

it's that expert knowledge and understanding

that people are coming to you for.

And I think, to some degree,

the business to business market

obviously is transitioning more easily

than maybe the B to C market will,

because obviously you've got very distinct niche audiences

that are coming to you for that kind of expertise.

The whole sort of digital transformation experience

obviously isn't limited to publishing,

and has been part of the core ethos of the business

for the last few years,

in the way that it's helped other businesses

through their digital transformation efforts as well.

So it's been a really interesting learning curve

to see how different industries

are tackling some common transformational problems, really,

that I know that we've gone through,

that you see them facing.

And it is sometimes difficult for our business.

And as much as we talk about how people should be doing it,

and then behind the scenes,

you do see some of the situations where actually

that's quite a challenging thing to do for organisations.

- With all of the publications

that were in Centaur's portfolio, if you like,

how did you go about it?

Have you sort of taken it

so that you've taken certain publications

and then moved them on,

or has it been a company-wide thing?

What are the steps that you've sort of taken

to get to where you are now?

- A lot of the transformation

has been mostly around standardisation of services.

So, the portfolios,

obviously we're all effectively individual businesses,

all run as individual businesses in their own way,

have their own reason for doing their own events,

their own magazines.

And I think the way we've approached it

is through a certain level of standardisation.

So especially around the marketing technology,

make sure we're using single services across the business,

and harmonising a lot of data systems and integrations.

It's about making businesses aware

of the opportunity they've got

from the services that are available to them,

by creating almost like a standard approach

to how they can then manage their business online.

Some businesses transitioned much better than others.

We've gone through a kind of

a phase withdrawal of the print publications,

a more sort of effective push of online events

and face-to-face type expertise,

sharing, networking type events,

and really a big push around

bringing our corporate clients into an online environment.

So, getting people more used to changing their habits

from expecting to pick up

a corporate magazine on a reception desk,

through to actually them logging in,

or viewing a set of blog articles,

or something on a phone on the train

on the way in in the morning.

And it's that sorta habitual change, I think,

that has been slow and gradual

over the course of the sort of last five years,

where we've been able to sort of,

not slowly substitute one for the other,

but as people get more used to consuming content online,

we've ridden that wave

along with the sort of the audience that we've got.

You can offer more value services online

than you can in a magazine.

And it's about trying to supply valued content

in a mechanism that works for consumer

that maybe a magazine can't.

There are certain publications that,

I can't say should never transition,

but you look at Creative Review, for example.

It's a beautiful magazine.

It's beautifully printed, the artwork's amazing,

it's for a creative audience.

You can only supplement that experience online.

You can't necessarily replace

that tangible feeling of holding that beautiful thing,

if you see what I mean.

It is easier with a smaller number of publications.

We were 50, 48 brands or something along those lines.

We've cut that down to two main portfolios now,

which we can really support.

- That's a massive task, isn't it? (laugh)

- Yeah, so I mean, from a data perspective

on the transformation side,

there's a whole lotta work

around just managing those portfolios,

and sort of standardising stuff,

making sure everyone's got the associated data they need.

But there're also a lot of massive similarities

in the way the specific B to B audience operates,

how you target them from a marketing perspective,

how you understand what the market sizing is.

From that side of things, the whole B to B data environment,

I can't say it's easier,

but (laugh) you've got some pointers

with some fairly standardised variables there

that you can work with.

- Obviously, when the publications were print,

there was a subscription model.

With the online, how many questions have you asked

of the people who are subscribing

about what they actually want?

How do you manage that?

- The registration systems

that we've got across the various products,

it's a balance of getting a level of information out

that you'd need in order to be able to operate the service.

By operating the service, I include marketing.

It's not ops, though, it's about delivering value content.

So, there's certain standardised fields

that you kinda need to have available,

understanding someone's level of seniority, for example,

or, you know, the actual company they work for,

the industry they're in, that sorta stuff.

That's all fairly required.

But at the same time,

we've shied away from asking more questions than that.

The reason for that is

we probably became a bit more sophisticated

in the way that we understand user behaviour

behind the scenes.

So the idea was we could trim down our data collection forms

in order to supplement it

with specific behavioural information.

I think, to some degree,

the issues that we have around data collection

tend to be ownership of forms.

Is it the product teams that own the forms,

'cause they wanna churn the people through the forms

as quickly as possible with as least barriers as possible?

We try things like progressive profiling,

you know, other types of ways

of extracting other types of information out

during the course of a visit.

The move to first party data being powerful,

it's not news to us.

We've always had payables, gateways,

soft or hard payables to force people to pay

or to register before they can view our content.

So for us, we've always dealt in first party data.

We do very little in third party data.

But the ability to be able to

really judge someone's intention via their behaviour

then requires a massive amount of work behind the scenes

to ensure the content they're viewing

is appropriately tagged

so that you can really draw

some strong conclusions for that.

If you've got an editorial team

more and more favouring the idea of

a sort of a transparent value exchanges

or zero party data approach,

where you just literally go and say to someone,

(chuckle) well, why do you wanna use us?

What is it we can deliver for you

that's actually gonna be of value?

Okay, we'll just stick to that,

and we won't do any of the other stuff around the sides.

If you tell us you're not interested in it,

we just won't do that.

I think just how you convert that

into some sort of tangible data collection thing that works,

it'd be very straightforward

for us as a data team to implement that.

It would be very hard for a business

to turn down the opportunity to sell someone something

that maybe they've implied they weren't that interested in.

I think that's probably

the hardest battle across the business.

It's like you send an email, right.

(laugh) What have you ever seen email open rates

that vary massively away

from what generally they always are?

You know, if you look at, what's it,

25% open rate on an email,

you send it to 75% of the people that aren't interested.

But yet you've got 100% of a list.

You're gonna use that list every time.

And every time, you get 25% open rate for it.

(chuckle) It's like, well, why are we sending it

to the other 75% of people all the time?

It's 'cause the business can't let go of the fact

that their list is this big,

and they have to use their list.

Because if they don't use their list,

they're not exploiting the value of their data.

When in reality, 75% of the people,

they just don't care less anyway.

- I was gonna say,

it's obviously there's changes in cookie law and, you know,

how you can re-market stuff and what people are looking at.

And that's sort of a whole intent data.

Is that something that you are looking at

as a business for kind of future,

as in what you could predict somebody would be interested in

to make sure that you're tailoring that experience for them?

- I think that intent data will be more important

when we look to associate with new markets.

It's difficult to some degree in some of our markets,

because we have got great coverage.

They're not big pools.

You can pretty much identify

all the people that are in the pool already.

You know, the account managers know the market.

They know the people they're dealing with.

They know why person X isn't buying.

We're not talking trying to sell 3 million widgets a year

to 10 pence each, or whatever.

It's high value sales, high value corporate subscriptions

to big companies who work in established markets.

To some degree, we're always fishing in the same pools.

And that's only gonna work for so long.

When we have exhausted those pools,

that's when we look to extend out into other areas, I guess.

That's when we start to really leverage

on building out regionally diverse audiences,

changing our editorial content approaches,

looking to find lookalikes,

all that kinda stuff will start to really come into its own.

It does to some degree in some parts of our business,

and those parts of the business tend to be the ones

that are more e-commerce focused at the moment.

Whether they're selling events, or training products,

or something like that,

that are separate to sorta maybe a corporate subscription.

We'll pick out specific product areas of the business to do,

not like stunk works, but you know,

sort of adopt certain levels of new approaches,

different ways to approach new challenges.

And they'll take it on in a little structured way

and then bring that back to the business as a case study.

That's kind of how we adopt new technology now.

So, if one part of the business starts to use something

and it really sort of works,

we'll then introduce that out

to other parts of the business,

or maybe look to remove

a couple of existing ways of doing business

and replacing it with a new structure

that's been proven in one smaller area of the business.

We can be a lot more agile that way.

We can resource for it a lot better as well,

because you don't then have a cross-portfolio project group

trying to administer, you know, an 18-month project.

You're dealing with a couple of brand managers

and a product team, or, you know, a marketing team.

Everyone's getting involved,

everyone learns how to use the products.

You can really make some changes quite quickly that way.

The approach to marketing segmentation (chuckle)

is kind of my background, more than anything else.

All the old sorta standard marketing approaches,

clustering, you know, segmentation analysis,

all that kinda stuff,

they're the same as they ever have been.

There's different tools and things

available to do 'em all now,

but that kind of approach and things

is very much the same as it always has been.

In a business that's so led

by people that've been in the industry for a while,

people that understand the customers so well,

the role of data to challenge their belief,

if you're coming at someone

with a fresh report I certainly don't know about,

it's very easy to gain their trust and bring them on board

with what you're trying to tell them.

If you're dealing with someone

who's been an account manager in the industry

for 10, 15 years,

and they've spoken to all the people,

and they're very clear on why people buy

and why they don't buy,

what role is there for data?

What can you offer them

that potentially they don't already know?

So, you're kind of setting yourselves up

almost like an antithesis of their belief

in order to try and push them in a different direction.

You're almost getting involved in psychology

in order to do your day-to-day work.

You kind of have to position everything that you're doing

in such a way that it allows people

to change their worldview.

- It's quite interesting,

because when we're talking about kinda subscription models

and stuff like that,

you're always thinking about prospects or customers,

and it being an external thing.

But it sounds though that actually

you're using your data strategy as well

to actually help and change behaviour internally as well.

Would you say that's true?

- Yeah, I would say that's probably more true

than it is us providing our data

to assist the external part of the business

and the customer part of the business, in as much as,

you know, data's the only way

you will change people's (chuckle) minds.

With digital transformation,

data transformation is always a subset of it.

It shouldn't be, but digital transformation,

all the gloss gets put into the website,

the process, the conversion rates, the e-commerce,

you know, all the flashy stuff,

and the, you know, we're gonna be remarketing this,

and we're gonna be, you know, using LinkedIn,

and it's gonna be great.

You're gonna see our adverts everywhere and all that stuff.

And then when it comes to actually

the nuts and bolts of the data transformation,

people forget that actually,

there's a very different set of skills

involved in a business with regards to how it uses data,

how it understands data, and how it processes data.

At Centaur, we're really, really lucky,

'cause everyone promptly brought into I want data.

You know, I know I need this data

in order to make these decisions.

It's just making sure that

the systems are there in such a way

to be able to deliver that information to them.

Because what you do tend to lose quite quickly,

without all of that really boring stuff,

to most people, not necessarily me,

but all that stuff around governance, trust, reliability,

you know, being able to put a number out to the business

and stand hand on heart and say, that is true

that is accurate, that won't change,

you only need a few disasters along the way,

where, oh, sorry, no, that's wrong. (laugh)

Can we have that back, please?

We've gotta do that again.

Or, you know, something else happens

and maybe you lose a bit of data here, there,

or everywhere where something wasn't quite tracked properly,

or, you know.

Without that sort of governance

and that rigour around the team,

all of a sudden you lose a lot of trust and reliability.

It sets your data transformation efforts back massively.

And I think that's probably the biggest challenge,

that kinda trust, building trust within the business.

Being there when they need you in such a form that allows,

you've gotta effectually be a mind reader.

You need to know what they're gonna be asking for

in six months time,

so you can make sure that it's available for them

when they get around to needing that piece of information.

- How important is it

to have the visualisation of that right,

so that it's accessible for everybody who needs it

do you think?

Is that a big part of a project, do you think,

of that transformation?

- Yeah, no, it is definitely.

I think we've been through various phases of this.

So, you know, your initial phase of migrating

from massive Excel sheets of information

that you're sending out, and going, look, isn't it great.

And everyone's looking at an Excel sheet going,

no, I have no idea what you're talking about.

Through to some standard visualisations,

where people go to, you know, a handful of times,

see things are going in

kinda the direction they want them to do,

and then never revisit them again,

to maybe trying to get it to that next sort of level

of how do you actually present that information

back to people to keep them interested

is a massive challenge.

I know a lotta companies talk about things like,

you know, lunch and learn sessions,

and getting the analyst out into the business.

We're not blessed with a massive data team

of thousands of analysts and, you know, massive budgets.

So what does make the difference

is just making you publicly available to talk through stuff,

to almost be the voice of the report,

and being able to recall that information

on demand (chuckle) in a meeting at the drop of a hat.

To be fair, lockdown's been a blessing for that,

'cause you can have a million windows open, right?

(laugh) And you can just be, like,

how many subscribers have you got on product H?

You'd be like, yeah, of course we've got 647 as of last.

You know, you can't do that in a meeting, really,

you'd look like a fool.

That sort of data literacy,

that understanding of what your business is

is something we're really trying to promote.

We call it know your business.

But it's just a one-pager, all the top numbers.

You know, how many subscribers have you got,

what's your average renewal rate,

what's your newsletter sizes, that sorta thing.

The most basic kind of information possible.

So I kind of go on (chuckle) the basis of

what can we do to make things easier.

So, when it comes to the analyst,

I'm trying to get those guys to be embedded within the team.

So when the project teams are starting to look at stuff,

they're infusing that understanding into the product

or project delivery.

They're bringing back the understanding

of what the business is actually trying to do.

And we can then make incremental changes to what we're doing

in order to make sure that we can answer

some of those questions,

or at least expectations are being met across the business.

That you're gonna be fine doing that,

but you're absolutely not gonna get

any more information around there

unless you wanna put this massive project in

and do this piece of work.

And we do do some fairly wide-ranging technology changes.

We've done quite a few over the past few years.

And they're not a lot of fun.

They tend to suck up all the resource.

So what happens then is you get this void of

I can't do anything now because the central teams are busy

doing this particular type of technical implementation.

So the business almost has to stop

and wait for that to all finish.

And then as soon as that finish,

you get the tsunami of data requests come in,

which you're not prepared to deal with

'cause you're still mopping up

from the end of the last project.

That's again where trust and everything starts to fall down.

So pretty much, and this is where my background

around business and marketing kinda come in,

is I tend to sit more as an interface to the business

than I do as an interface to the data.

Is translating what the projects and things are doing,

what volumes we have,

what approaches we can take

in order to answer the questions.

- Yeah, it sounds like,

with some of the things that you've mentioned there,

that that has been a hard lesson to learn,

but you've been in that situation,

and things have gone pear-shaped,

let's not go there again.

Has it been a baptism by fire over the last five years,

would you say?

- The baptism of fire is more around

understanding the prioritisation

of where you sit within a business.

Now, I had this thing where, you know,

phase two never happens in a project.

So, when you're walking through something

and they're like, yeah, that's great,

but we'll put that in phase two.

We originally had it put in phase one,

but we're not gonna quite hit that,

so we're gonna move that out.

That tends to be more of the analytic type jobs

in favour of the front end working properly.

Obviously people have to buy stuff,

people have to register, and that user experience,

which is absolutely right.

The frustrations tend to come from the fact that

you always feel, to some degree, second fiddle.

I think it's

you know you're gonna get asked questions on this later,

and you know you're gonna have to say,

I'm sorry, we can't deliver that because it's actually,

it wasn't included as part of the project,

or we weren't able to get that in on time.

- I was gonna ask a little bit about shoring up.

Sometimes you haven't got the information,

because either it's gone into, you know,

sort of it's the next phase

or you get to the end of it and somebody says,

well, we really could have done with that.

And how do you go about

scoping out a project as best you can

to make sure that you have, you know, shored that all up

so that you've got as much information as you can.

Is there a process that you go through to ensure that

when you're scoping out those individual projects,

that they are gonna deliver on what you're gonna need

or what you're gonna be asked for?

- There are control mechanisms

and things in place with project management techniques

that allow you to ensure that

certain things get picked up at certain times.

And I think data is no different.

It's the same as any other part of the business

trying to deploy something,

making sure that business requirements are met

from an early time.

That being said, I'd be a complete liar if I said

that that happened in every single case.

(laugh) Works in a very similar role will understand,

the day before something goes live,

you get the email that says,

oh, we're launching this tomorrow.

And you're, okay, I know I'm gonna be asked

for some stats on that.

What have you put in place?

Yeah, phase two.

Okay, so I've got no data for it.

Everyone's gonna start looking at me like I'm an idiot.

I can't answer any questions on what you've done.

That's just the way of life.

But by pushing people out into the product teams,

you know, you get those early warning signals.

That's all you can do.

Just make sure you're engaged.

And I think if someone's putting a new product in,

it's gonna be based on a business goal, right.

There'll be a very specific forecast

that we generate for that.

Centaur doesn't spend money on stuff on a whim.

If it's gonna throw some cash at something,

then you can believe there's a very good reason behind it,

why it's being done.

So, by engaging earlier in that process on any new project

or any new piece of work that gets done,

that's probably easier.

Because people wanna know

whether their business goals being hit.

Because ultimately

that's how they're gonna increase their budgets next year,

that's gonna lead them to spend more money on marketing,

that's gonna grow their own side of the business.

So, they are more likely

to engage with an analyst earlier on

and say, what does this look like,

what do I need to have in place

in order to prove that what I've said there works.

The harder bit is on the established businesses,

where they're just making incremental changes.

And not in any way blaming dev teams,

but quite often, analytical implementations are thrown in

with third-party services, heavy JavaScript based stuff.

Developers don't like the fact that

you've just installed 50 or 60 different types of service

or tag that runs alongside

their beautiful hand-coded JavaScript.

So, you know, slowing stuff down,

making stuff potentially fail for some reason,

generating errors, all that kinda stuff.

If you don't have a relationship with that team,

then it's very easy for them

to just start not engaging with you.

And then that's when things will sort of start to fall down.

So it's about sorta solid project management stuff,

making sure you're engaged and involved

in the business as you should be.

But then relationship building,

being out there and chatting to everyone,

making sure you know what people are trying to do,

understanding why they're trying to do it,

more than anything else.

- Would you say, internally and externally,

that data is the new marketing currency?

- The new currency is almost driven by the velocity

at which point you can get that information now.

I used to work on clustering exercises,

maybe take a couple of weeks to run.

You'd put in something, you may need an analyst then

to go and spend some time looking at it

and figure it through.

And tools and techniques have changed now,

so you can run almost in constant, real time,

leads going classification

of every single individual person that comes your way,

rather than trying to then take something offline,

work out how everything fits together.

That's where the difference is.

I would say it's less data is the new currency,

and it's more data literacy is the new currency.

When marketing automation came about,

I think it was, to some degree, everyone thought,

oh great, I'll automate this.

Automate equals less work, right?

So, not so much to do.

Whereas actually automation means

massively increasing your level of data literacy.

How many marketers have ever been

on the data literacy course?

How many marketers understand what their data literacy is?

Do they really understand

the intricacies of all of the various ways

that a marketing technology platform could be used

and harnessed, you know.

Centaur is massively guilty

of buying into some big, big packages,

and not necessarily then following up with

using all the benefits of those packages bring.

So, I think data is the new currency

in as much as it's the understanding

and the use of that data

within the technology that is now available to people.

If you can find those people,

those people that really understand that,

then there's some really

forward-thinking marketers at Centaur

who are quite happy to pick up on projects

that should be data team projects.

They're quite happy to pick up

and run with that kind of information in that.

They're asking very specific questions of data.

They know where things are falling over,

and they're pulling you up on it

as if you should be the people

that are supplying this kind of information

in that way to this product.

And that in itself makes life a lot easier for a data team.

From a data literacy level,

if you've got people

that are able to talk that kinda language

and engage with you,

they can really share with

what the business is trying to achieve,

and ultimately how they want it to work

within that piece of technology,

then that makes life a lot easier.

- That sorta brings me on nicely

to a sort of question that I would like to ask,

which is if somebody's trying to get into the industry,

what sort of skills or experience

do you think they should have?

Or what kinda things would you be looking for?

- We've looked to put in place a number of hiring strategies

across for the last few years especially.

I think we've seen, everyone's seen

a scarcity of resource across the business.

I think the key thing, when it comes to skills,

if we're looking to hire very specific skillsets,

we have to round that out quite a lot.

We're looking for actually

what are the transferable skills.

What are the industry areas

or types of businesses and things like that

that people may have worked in before

that might have experience that is relevant to us.

We've brought in maths grads in the past,

and that's been really useful.

They're quick learners.

They don't necessarily know the ins and outs

of all the pieces of marketing technology

and how to get the data out,

or, you know, they're blank slates effectively.

Might've done a bit of coding or something in university.

Have equal success with people that are sociology graduates,

that have got a really nice way of talking to people

and getting information out of people and are keen to learn.

You know if someone's quick, if they've got it,

if they understand.

They're good attention to detail, they're willing to learn,

they're interested in the industry.

You figure it out.

You kind of sit down and work out what's happening.

You know, you then look at what the causes of that could be.

Then you look into the sort of the issue

that's surrounding it.

And then, you know, you look to put in

some sort of resolution.

It's that kind of mindset that you're looking for

more than anything else.

Because ultimately, skills can be trained.

If they're not interested, they're not gonna bother.

You're just wasting everyone's time, at that point,

- It's people are passionate about data

rather than just thinking of it as a day job.

- The people that are good,

they're still talked about at Centaur.

You know, these are grads

that came and spent two, three years with us.

Did some stuff, really made an impact, and now have gone on,

I mean, they're (chuckle) earning more than I am now,

but you know. (laughter)

Sort of three years later.

But, you know, they've got that passion for it,

but that's what shows through.

So, they're remembered by people.

One of the managing directors

in particular parts of the businesses,

they've gone on and done something else somewhere else.

You know, you start to get those phone calls come around,

oh, we're looking for someone,

or, you know, we've got an opening for this particular role.

Who are you gonna recommend?

Oh, we had that grad in.

They were amazing.

You know, it doesn't really matter

if they didn't have the direct skills,

they picked up this thing really, really quickly.

I'll happily recommend them to you, you know.

And that sorta thing is invaluable.

I think, for any graduate, that networking side,

listening to what the business wants,

not trying to over-deliver on everything,

just getting known for a few core things,

and then, you know, you're sorta set for life,

practically, in the data world, aren't you?

That's it for them, I mean.

- It's a smaller industry than people would learn,

isn't it? (laughter)

- Absolutely.

- My last question, just before we finish this,

I just wondered about the sort of moving to the cloud space.

What are sort of the big business drivers

that you've seen for that?

How do you even empower an organisation

to kind of really embrace that?

We're all working in the same direction.

- I don't know why anyone

would ever wanna see a server room again.

(laughter)

It's allowed us to start to really kind of experiment

with different techniques,

and the things that we can then potentially

bring into production.

So some of the stuff that Red Olive are working with us,

around data cleansing and processing,

where we're trying to really make sense of

the first party data that we've got, for example.

We'll use Red Olive to help us to create

a sort of an end-to-end process for that.

Cleansing, working through data in a sort of data pipeline.

It's fairly standard stuff, right?

There's nothing particularly exciting about that in itself.

But having the availability of the sort of AWS tools

or other providers tools

allows us to maybe do some of the stuff

that was manual in a different way.

So, we're looking at potentially using some AI approaches

to even data cleansing.

You know, handling our job title classifications.

We've used it quite successfully

for classifying editorial.

So, training models to understand

how to bucket certain pieces of content,

so that you don't have to then worry about

editorial teams doing what they should be doing

from a data perspective.

And for being able to sort of organise

and manage most of the metadata

that sits behind the information that we gather.

Using these sort of AI type machine learning tools

allows us to really sort of move away

from those manual jobs.

- Would you say, is that the thing that's most exciting

about kind of being able to do stuff

that quickly with the AI space?

- I think so.

My background is not analysis.

The kind of using machine learning

for generating certain types of fancy analysis

or stuff like that, I have to leave that to the specialists.

You know, I'm a man of certain age.

I have to sort of respect the fact that

I've only got a certain amount of capacity

to really get to grips with this stuff.

But when it comes to the nuts and bolts stuff

that I have been doing over the last 20 years,

that I have experience in,

data cleansing, integration work, all that kinda stuff,

then finding opportunities in the cloud tools

to really sorta leverage

how that makes my life a lot easier,

then I'm very interested in.

- An interesting take from Ben there.

Join us for the next episode

of The Red Olive Fibonacci podcast,

where we'll be joined by

another data expert sharing their thoughts

on the latest trends in AI and big data,

along with some great hints and tips.

Make sure you subscribe to The Red Olive Fibonacci podcast

from wherever you get your podcasts,

to make sure you don't miss it.

That's all for today.

Thanks for listening.

I've been your host, Nicky Rudd.

See you next time.