Global Marketing
Management Consultants
Global Marketing
Management Consultants
mobile-logo
Global Marketing
Management Consultants
Top

Managing Marketing: Data In Media, Marketing and CX Management

Steve_Sinha
28
Page Views

Managing Marketing is a podcast hosted by TrinityP3 Founder and Global CEO, Darren Woolley. Each podcast is a conversation with a thought-leader, professional or practitioner of marketing and communications on the issues, insights and opportunities in the marketing management category. Ideal for marketers, advertisers, media and commercial communications professionals.

Steve Sinha is the Chief Operating Officer of the Australian Alliance for Data Leadership. Here Steve shares his views of the role and challenges of using data to inform media, marketing and customer experience management and how the hype and reality gap exists not because of technology but the slow pace of marketing transformation.

You can listen to the podcast here:

Follow Managing Marketing on Soundcloud or iTunes

Transcription:

Darren:

Welcome to Managing Marketing and today I’m sitting down and having a chat with Steve Sinha who is the Chief Operating Officer at the Australian Alliance of Data Leadership. Welcome, Steve.

Steve:

Thank you for having me, Darren.

Darren:

Well, it’s been a long, long time. I think we met soon after I started TrinityP3 or P3 as it was in those days back in 2000.

Steve:

Absolutely, it would have been around that time. And since then I have been thoroughly appraised and pitched by you sometimes for better and sometimes for worse.

Darren:

Yes, you’ve been on the horizon, part of the ecosystem especially in regards to media but in the last few years your career has expanded beyond just media hasn’t it especially with this role in the AADL?

Steve:

Yes, indeed it has. But people talk about moving out of parts of Comms into Data. Media is a pillar of Comms whereas data has always been at the heart of it.

I always encapsulated my media career (and I wasn’t unique in this) as helping marketers understand consumers. Obviously, there was a strategic and analytic rigour as part of that, distilling insights etc, but the data was at the heart of that.

The data at the start of my career was highly primitive but it was the same thing. It was what do we have that allows us to take objective decisions?

Darren:

That’s a really interesting point. The only thing I’d challenge you on is at the start of your career at the start of the 21st century for a media agency the idea of data was the subscription surveys like the Morgan’s and the Nielsen’s, maybe some first party data that the clients may expose them to, and then maybe some bespoke research.

We’re talking maybe four or five sources. You’d have to admit now that when we talk about data (especially data analytics and the insights) often they’re talking about 100s of sources of data, massive amounts of information that make the decision-making a lot more rigorous don’t they? Compared to four sources; I checked Morgan and Nielsen and that says that this is our audience.

Steve:

Absolutely.

Darren:

Has it changed much in media agencies?

Steve:

Yes, it has to a degree but as you start to map that out (going back down memory lane; Morgan, Claim Data, Call Research, Focus Groups) but interestingly they look like blunt tools now but that was your palette of data from which you drove strategic decisions.

When I first joined OMD when James Greet was putting together that forward-looking structure (rapidly copied since then) we were one of the first early adopters who had an econometric modelling division so I became a big proponent of that and tried to promote that to clients early on.

And when I was exposed to that I saw the power of data and modelling but early on probably the biggest barrier at the start was clients going, ‘hang on I don’t know if I’ll give you permission to do this, you’re a media planner’.

But you’d go ‘but this is taking media planning to the next level—you’ve always wanted to get to a high level of understanding of the macro effects of ROI so surely you should be happy we’ve now got this in our toolkit’.

And in those first couple of years there was certainly some resistance with people going ‘I’m not sure I can give my media agency permission to delve into this’ even though you’ve hired non-media people and specialist data scientists to help with it.

And then over time it became absolutely accepted into the media and broader marketing toolkit and marketers and agencies will have data scientists at the heart of what they do. That evolution has been interesting.

Yes, the depth of data you have now where you are able to look at, manipulate, evaluate in the fraction of a second really big data sets is much more powerful but the analytical and strategic process to get to powerful transformation answers for businesses is still the same.

You still have to take a source of data and find the most compelling things that can lead to a transformation but throughout my experience in marketing there’s never one answer. There are lots of potentially right answers but you’ve always got a finite budget so you always try to go with what is the best answer.

If you’re at a higher order of things invariably you try to go what is the one thing we can change tomorrow (maybe it’s talking to a new audience that are more likely to buy our products more often or higher value) that will make the biggest immediate difference?

When you listen to really good data scientists they talk the same way. They go, ‘to get a foothold in an organisation we try and find what’s the right thing to change first. I want a success and a number.’

Darren:

I absolutely agree that when you’ve got huge amounts of data there are so many things you could find but you’re really wanting to find what’s the biggest problem and the biggest opportunity that I can focus on first otherwise it’s literally a flood of choice of things you could possibly do.

I want to go back to the amount of data because a lot of people talk about data; almost all conversations in marketing and everyone will have some sort of data analytics and insights capability because it’s expected.

But there’s a big difference between having a data analyst work on four or five sources of data and working on hundreds of sources. I caught up with Martin Cass out of New York MDC Media Partners and he was telling a story about asking one of his data analysts ‘what data do you need?’ And the guy just looked at him and said, ‘just bring me everything and I’ll tell you what I can use’.

Remember when Quantium first started and they were positioning themselves as being so much better than any agency because they not only had all of the same data agencies could subscribe to, but they also had a whole lot of credit card and electronic fund transfer data, then Woolworths came along and all that data was poured in there.

So, this idea that the richness and validity of the insight is almost directly proportional to the number of sources and the size of the data set that you’re working with. It’s still an issue isn’t it because a lot of people working in data analytics are still being starved of getting their hands on rich sources of relatable or valid data?

Steve:

Yes and it’s interesting (I don’t mean to be controversial) but that Quantium model was really scary when it was first mooted because you’ve got access to a couple of pukka first-party data pools and you’re putting yourselves up as the world’s best modellers.

There were probably nervous moments when we were up against them back in the media days because they were going to change the world.

Darren:

Except that the world seemed to catch up with them very quickly.

Steve:

The world caught up very quickly and a couple of times I was exposed to the actual working of that. They never drove some of the deeper insights I thought they might have done. With a combination of first-party data sets and absolute econometric modelling expertise we can really come up with some bigger and better marketing, product, price, and audience decisions.

And they didn’t get beyond the things media people were doing – now here’s a bit of flighting – but it became every time we model something if you put your money in the cheapest cost per 1,000 you’ll sell more. So, in a way it always came back to put all your money on regional TV.

So, some sharper analysis over the top; it was a good toolkit and they’ve got a great business if you’re working on some marketing problems and the quality of data is really important. You think all data is equal but there’s premium disclosed data you can buy and really good data sets and where you got your data from could be as murky as how much you paid on the way through.

If you were a marketer there are things you would buy from Quantium in terms of data sets that are premium and robust. But somehow we were expecting that model to take over at least part of the world and it never quite did. I wonder whether that was just the power of the analysis over the top. They had all the machinery.

Darren:

I remember one of the issues was the fact that while the sources of data they got were addressable as in geographic,  individually addressable without actually identifying the individual, they knew the streets that people lived in and things like that. I remember conversations around the fact that very little media data is actually at that level.

Like it’s very hard to get media data at a postcode level or a street level right? You get TV by metro regions and it might get split down into northern Sydney and eastern Sydney but you don’t get it at that level of it’s happening in this street, 90% of people watch this programme or listen to this radio station at this time or read this newspaper or magazine.

Steve:

You’re getting really fudgy geo-graphics.

Darren:

There’s this disconnect between you can have all the data in the world. You can have millions and millions of data points from 100s and 100s of sources but when it actually comes to now being able to address that from a media perspective it has always been flawed until now. And now we’re starting to see the beginning of addressable media; it’s starting to become more of a reality isn’t it?

Steve:

Absolutely. I think you’ve hit on the nub there. Theoretically, data sets in the last thirty years you’ve been able to see a lot but then how do you translate that back into a media or marketing plan that’s going to make a great difference. I’m sure using these things as well as I have in recent years—someone says, ‘we’ve got all this data so we can move to a much more precise intent behaviour-based audience and that’s going to drive everything we do in real time’.

And you’re really looking forward to this and you turn the page over and they’ve applied that to about 8% of their plan in digital where they’re kind of doing that. And you ask well how did that apply to the outdoor or radio. I’ve been following all the connected TV developments very closely and the Upfronts last week (the numbers supposed to be higher now based on a conversation I had last December) but even so Lion have got to 7 million addressable individuals through logins on the connected TV.

That to me becomes a really exciting number.

Darren:

Well it’s got mass—individuals at mass—and you can start to look at their viewing profile based on their logins.

Steve:

So, as a medium to large scale marketer do you now get to a place where you can have a really tight intent-based audience and all sorts of intent based behavioural definitions. But I could spend 90% of my money and not 8% with the rest, a kind of blunderbuss approach with a blunderbuss delivery but I could do 80, 90, 100%.

What say I duplicate everything I can get on Fairfax—around 500 million uniques a month—where’s the number is it 10? For most target audiences we can dive into a live pool of 10 million people; it’s very compelling. Are people going to buy this next week or are we after people that are lactose intolerant—if you then go and analyse 10 million people we can pick up a bigger pool of these people.

Is that enough for us to go let’s spend our money there for four months? And does that make up every sales metric we need?

Darren:

There are a couple of points there. The first is what number, 7 or 10 million is duplicated because there are going to be a huge amount of those individuals. Even creating logins; there’ll be people who have two or three logins on their iPhone, tablet or smartphone, they couldn’t remember the passwords so they created a new login and a new email address.

The second thing is that what we’ve seen is that marketers and data analysts will start with a pool of data and start to segment that down to find particular valuable groups of people in that particular data set without actually thinking of those sets as being media addressable.

You’ll end up with segments of people that do this and this—largely behavioural because a lot of the data has come from online where it’s more about behaviour rather than demographic that have this particular behaviour profile and then they turn to the media and go we want to buy this segment. And the media goes, ‘it doesn’t align to any way that we segment or sell media’ because media is still largely segmented on demographic profiles.

I know they’ve done some psychographic but largely it’s still grocery buyers with children under 15 and things like that rather than a psychographic or behavioural model.

Steve:

Absolutely. Taking that Channel 9 example—anything still back on the main network is still tied down to a demographic. I don’t know how that duplication will work out but the base at the moment is 7 million and you can spread out – it won’t be 12 but whether its 8 or 10 I’m not bright enough to work out but whatever that number, with that pool you can go let’s start with that, that becomes compelling.

Because if you go on a classic always-on reach build rather than saying it sounds great but ultimately 60, 70% of my money is in a pretty broad demographically bought mass medium and the cherry on the top is this intender I get for 20% of my money I get through my digital stuff.

You might be able to flip it on its head and go because I’ve got a pool of 10 million people, it might be on this intention audience whatever it is I might be hitting as few as 30%, maybe as many as 90% depending on how narrow or mass that audience is and the ability of that media to hit it.

And you might be able to go well I don’t really need too much beyond that—we’ve done 8/10ths of the job through that and everything else is just a bit of an Ehrenberg-Bass hit everyone—that becomes a really interesting experiment.

Darren:

The problem with media, even digital media, is the huge amount of wastage because even when you go online with a programmatic algorithm to target a particular profile you’re still relying on the quality of the publisher’s data about identifying that.

I know there’s talk about being able to target individuals but then there are a whole lot of issues around privacy. The day you can sit there and target Steve and I know his email address and I know where he lives but having that level of data would have to come with having your permission.

Steve:

I think that’s a really interesting topic, probably something I’ve been thinking about in my more recent roles at AADL and so forth. ADMA has participated in some global research into consumer’s relationships with data and the value of data and so forth.

There was an update that came out in April (comes out about every two years) and you could pull out Australian’s attitudes to data as well as a global panel. If you look at that and at what some people are doing in response to GDPR—how do you make this all work. In a theoretical way we all want it to work. I think part of this will be encouraging consumers to share their data.

So how can I be really clear that I’ve got you, Darren, and all your right logins across the top box on your TV, the iPad and your iPhone and every device you use I’ve got it covered and can I identify the devices you give the twins. I’m really clear on making sure they don’t get served adult 30 second ads instead of kids ads. So, how can I do that? It’s probably time that people motivated you to do it. If we can know you’re on these we’ll reward you.

Darren:

I worry that as an industry we’ve missed the boat. There was a point maybe 10 years ago where the industry should have stepped up and said, ‘we will have an exchange of value; you give us permission to know who you are and contact you and we will make sure that you are not spammed.

Because what’s happened is with my email address (my business email is my primary email address) every day I get emails from organisations that have clearly bought a list of CEOs and are just emailing me all sorts of things. And for me to sit there and go unsubscribe, unsubscribe, which I sometimes think just validates that it’s a real email address so rather than being unsubscribed I then end up on more lists. Now, I just hit junk filter, junk filter.

Steve:

Exactly, especially when it’s a company that’s told you for 15 years—I’m not in your audience.

Darren:

We’re all so burnt out by the distrust of the fact that our data for ten years has been used against us—to then turn around as an industry and say, ‘we want your data so that we can do a better job for you’.

When I turned 50 (I wasn’t married) I got spammed with Russian brides and Viagra. Neither of those were things I was interested in but at the most rudimentary data profiling they went ‘wow 50 year old man, not married; he either needs a wife or chemical help to maintain an erection’. And that was it.

Steve:

It must have been around 6 years ago and I started getting those random emails about Russian brides even though I was married. What’s all this? I think it’s auto-generated spam. It hasn’t been prompted by anything I’ve looked at.

Darren:

You can see what I mean. Why didn’t the industry actually take the opportunity to promote this? There’s a lot of talk about Blockchain but one of my favourite ones is that they will build a blockchain where you will be able to put all your personalised data, all of that information, all of my email addresses, the devices I use, my accounts I subscribe to, all my passwords but I will be able to because it’s in a Blockchain to release it to people selectively who have access to it. And I love that idea.

Steve:

That’s fantastic. I agree with you wholeheartedly about everything you’ve mapped out. Again we’re going to get into conversations around CX and the merging of marketing and CX so how our whole industry can be talking about this but not thinking about the effects of constantly spamming people and annoying the hell out of them.

The two don’t go together. I agree with everything you said but I think an optimistic thing is have we missed the boat so is that the current reality? Are there problems with it? Yes. Have we missed the boat? Not necessarily.

And the only reason I say that is in this research into consumer attitudes to data—basically it segmented the Australian population into three segments based on their relationship with data. And the largest segment, which is 54% of the population are a data-educated and data-literate segment.

That’s not overtly skewed by demographics. It is quite broad across all age groups etc. That group will say now, ‘I get there is a value in my data, I understand that, I’m open to it, my only beef though is the value exchange doesn’t yet work for me.’ I know there is a value exchange. I know why you want it. I know it helps you sell more products and I’m not averse to that. I’m waiting for you to make me an offer.

That’s the only reason why I say the horse hasn’t bolted. I think if all this stuff happened now or if brands now were very overt about I know you know we’re going to use your data to put the right products and services in front of you but I’m going to make it worth your while. Yeah, 54% of the population is open to that conversation right now.

Darren:

That’s why I think the voice activated search that Google, Amazon, and Apple have got into is really interesting because what you want as a consumer is for it to predict what you need before you even know it.

So, when you say, ‘I feel like pizza tonight’ not that it recommends the pizza that the advertisers been paid to recommend but has actually profiled you to know which of the pizzas on offer is the one you prefer or is most likely to satisfy you. And people talk about being able to achieve that except that the commercial model seems to be we’ll just recommend the brand that’s paying us the most to promote them.

Steve:

Exactly. I think that’s what we all want but how do you get to that curated model?

Darren:

And the argument is that pizza is pizza but that completely destroys the idea that brands exist because brands are the things that differentiate one product from the next and it stops them being commoditised.

Steve:

It comes back to the research we’ve seen and the Spotify model and others—Apple is probably a good example because it’s a premium brand and people pay silly money for it.

If you took Apple customers as an example, across the whole world of walled garden of curation we can offer you would you be prepared to pay a bit more and we’ll turn off paid recommendations and give you a genuinely curated service with our amazing algorithms. There might be group of people that are up for that.

Even if it only turned out to be 10% of all Apple users worldwide (you’re hoping it’s more) that’s quite a valuable group of people.

Darren:

Could you imagine in a world of transparency when you say ‘Hey, Siri, I’d like a pizza’ and Siri comes back and says, ‘We’ve been bid $7.15 to recommend Crust, $5.20 to recommend Dominoes’? Choose the one that’s paid the most for your service or custom.

Steve:

Would it work now? Try this at home. If you went, ‘Siri, who’s paid you the most to offer me a pizza?’ Would it give you the truth? Can Siri and Alexa lie or do they only have to tell the truth or are there filters in there?

Darren:

I think they just filter out and don’t give them the data that would be required to do that. But you mentioned customer experience and if anyone has a right or is someone I’m willing to give my personal data to its brands and companies that I do business with and especially financial businesses.

You would have to say banks and financial institutions have a huge amount of data on me and ask for data all the time if I take out a mortgage, want a loan, or credit card there’s a huge amount of data that I have to provide.

I won’t mention which bank I’m with but I have three different customer Ids because I’m a business customer, personal customer, and a home loan customer. In your experience in the world of data how do you think that works to their single view of me as a customer?

Steve:

The bank I’m with, the same, there’s a business account and some personal accounts but that ratchets up in the same environment so assuming they can easily work across both, can’t yours do that?

Darren:

No, every time I log in if I use the wrong customer ID I get the wrong perspective. It drives me crazy if I call them up and they ask for the customer ID I have to remember what product it is so that I give them the right ID for the right product.

This is one of the big four banks. We’re talking about premium banking in this market and in the 21st century they’re struggling with giving me a single view of me as a customer. And my previous bank had the same problem as well so two out of the big four struggle with this.

Steve:

So me as Bank A, I’ve got a customer ID I put in and then on the website I can ratchet between mortgages, personal accounts, and business account.

Darren:

Once I go in I can navigate around.

Steve:

I’m assuming within that architecture they would have a one customer view of me. They should be able to capture that.

Darren:

Except if I phone their call centre I have to pick the right customer ID so they can then get the right view of me.

Steve:

That should be easily solvable. Marketing is land grabbing to customer experiences if we’re talking to you in a web environment we can understand simple user interfaces. If I change that clicks go up slightly or bounce rates or whatever, which gives people permission to take control of that.

You would think for any brand worth its salt that’s an absolute laboratory where if I haven’t got one customer view or anything happens along that online journey that loses the customer, frustrates or annoys the customer that’s easy to fix.

And there seems to be a world of data scientists out there that help with all those things.

Darren:

The role you’ve been in recently there’s a lot of talk about data-informed customer experience and the role of data in creating customer satisfaction. How far do you think the hype is compared to the reality because there are so many places where you see people talking it up and yet the actual implementation feels like it’s a decade behind?

Steve:

Yeah, I would probably agree with you. The hype is probably still out the front. Theoretically the tools are all there and the disciplines are mushrooming so that theoretically you should be able to achieve it. But are 100% of brands achieving it? No.

Going back decades, with the available data, theoretically everyone is data-driven and you take marketing and business decisions based on the objective analysis in front of you. If I went through every marketing and Comms job I’ve been involved with and I presented someone with a recommendation based on the strategic and analytical view of the data in front of me how many people have gone objectively agree or have got some different perspective.

Or how many people have then taken a decision subjectively on gut feel?

Surely in the last four or five years every single excuse to take subjective decisions has gone because we have the data and the tools. How many people out there are still making subjective decisions? Too many. You’ve got all the tools, technology, and skill-sets available but you’ve got to start from a place where you have to be led objectively.

If you’re still in a mindset where you’re still subjective, the minute you’re presented with an objective ‘do this’; ‘our company’s not set up for that, it doesn’t work with us, we’re the exception’ then it all falls in a hole.

I think that reality could be achieved but it puts more and more pressure on people to go ‘we’ve taken every peg out now; you have to have an objective approach to everything that’s shown to you.’

I’ve heard a couple of talks from people at the Iconic and in some ways their approach to this makes a lot of sense. Again, they’re a pure play online retailer and they seem to do things very well and they’ve built a really good business. Their view seems to be you can get absolutely buried in data.

Your starting point is what are your business objectives or what’s the question we need to answer? So when they put a project to their data team—they don’t do that. They work backwards—what are we trying to answer from a business perspective? And then go, ‘O.K, it’s this, do that’.

Darren:

Which is a better approach, get all the data and then work out what the problem is that we’re going to solve. What’s the problem or opportunity and now how do we cut the data in a way that actually gives us a solution or an insight into the solution?

Steve:

That’s a really good way of approaching it but it’s a direct answer to your point that we now have mind-numbing levels of data. And again you can be in a Facebook environment or some complex business and you can look at millions or billions of rows of data every day. You can get utterly blinded by this.

The antidote is to not run everything you can and plenty of businesses do this and probably those banks are the same. You couldn’t run all the data on a daily basis to see what’s happening; it would take you forever to run it. So, what’s the question; let’s configure our data to answer that. You need that focus.

Darren:

The really smart organisations and especially the smart marketers are the ones that are adapting from the gut instinct, intuition model to the rational data-driven model by saying, ‘an informed decision based on data is still processed through the instinct or intuition’. And the reason they do that is that an informed decision is one where the data informs it but it still allows for what’s called the human condition.

So, I agree. If you can at least define the problem that you’re going to solve upfront then the data is there to hopefully find the solution. But, Steve, I’ve just noticed the time. It’s amazing how time flies when you’re having fun.

Steve:

It does. I really appreciate being on. I might add just for the listeners, because this was an audio and not video we’re both naked.

Darren:

Terrible. Thanks for popping by. I do have a question for you. We hear about all the people who are doing data well but can you share with us someone who you think has really cocked it up?

Media continues to be the single largest budget item for most advertisers. But media has changed significantly. Find out about our media solutions here

Want more articles like this? Subscribe to our newsletter:

Fill out my online form.

Darren is considered a thought leader on all aspects of marketing management. A Problem Solver, Negotiator, Founder & Global CEO of TrinityP3 - Marketing Management Consultants, founding member of the Marketing FIRST Forum and Author. He is also a Past-Chair of the Australian Marketing Institute, Ex-Medical Scientist and Ex-Creative Director. And in his spare time he sleeps. Darren's Bio Here Email: darren@trinityp3.com

We're Listening

Have something to say about this article?
Share it with us on Twitter, Facebook or LinkedIn

12 Shares
Tweet
Share12
Share