Setting out a plan for using big data
With data now at the heart of more major marketing campaigns, how are companies collecting, organising and distributing customer information?
The panel (l-r above)
Branwell Johnson, acting editor, Marketing Week (chair)
Andrew Ananth, head of data, Carter Jonas LLP
Nick Blunden, global digital publisher, The Economist
Rachel Bristow, VP global media data and analytics, Unilever
Tanya Bowen, director, CRM loyalty and clienteling, BT Expedite
Simon Kaffel, global head of data and analysis, Zurich
Darren Robertson, digital communications data scientist, Action for Children
Nigel Vaz, senior vice president, managing director, Europe Sapient Nitro
Marketing Week (MW): Marketers are faced with ‘big data’ as the new buzz phrase. What does it mean to you?
Tanya Bowen (TB): The challenge for us at BT is not necessarily identifying what you can get out of big data because I think we’re pretty clear on that. It’s the cost it takes to aggregate that data so that it can become meaningful and married up with all the traditional data our retailers have.
Darren Robertson (DR): What is big data? It is anything you can’t put in an Excel spreadsheet. It’s taking unstructured and structured data and marrying them together to give us business insight. Government cuts have hit charities really hard and we have to get smart about how we get new donors on board, and how we keep them.
Andrew Ananth (AA): There has never really been anyone in property who has taken ownership of data. At Carter Jonas, I’ve got a mammoth task over the next year to do with data cleansing as well as trying to consolidate all the data in the different offices, Excel sheets and many different databases from an operational point of view.
Rachel Bristow (RB): Unilever has a big data programme that looks at bringing data in-house, making it accessible across the company and marrying different data sets. My role is to look at consumer data and how we get global governance and processes in place. It’s also about how we change our marketing structure and ways of working to enable us to increase consumer loyalty and drive bigger share growth than we already have.
Nick Blunden (NB): We talk about structured and unstructured data on a scale that can’t be managed by standard data management tools. Like most media companies, The Economist has made money over the past 169 years by controlling the information flow. It’s much harder to do these days so there’s a question mark about the role that we play in a world where data and information are so readily available. It impacts on our customer acquisition activities, our circulation marketing activities and our advertising revenue stream.
Simon Kaffel (SK): Big data is the latest phrase that management consultants in technology companies are using to say ‘buy more technology and services’. Data is data, whether there is a lot of it or not. You’ve got to get the basics right, do analytics, do reporting and generate insight that’s valuable and actionable.
Data has been captured for God knows how long but the people who are driving usage has changed
MW: What team structures do you need to perform data mining and then move that into actionable data?
RB: We’ve got inconsistent structures and that is a problem for me. Where do we start? Right now, we’re spending a lot of time understanding the big markets. China versus the US versus France - they’re all at different places. They’re all trying to do a really good job but, actually, when you look at the data sets we have, they’re incomplete. China is setting up the processes and is just on the journey of setting them up in the right way.
AA: You need data stewards to control the quality of data going in and to govern it. You need someone who can manage the CRM and then you really need someone at our level who can be the PR person for data, who goes to meetings to give advice and lay out strategy but who also has a team to implement it.
RB: Tools and systems are the easy bit. It’s the people we need. We don’t recruit for data analytical capabilities. We’re not looking for people who understand or who are ambitious about using data, it’s not part of how we’re training the marketer at the moment. We’ve got significant changes to make.
TB: We find that you can take advantage of point pieces of big data with existing resources that are in-house. When you have someone working with digital and social analytics and someone dealing with more traditional analytics, the strategy becomes disjointed. We encourage retailers to leverage existing resources to become involved in the new pieces of analytics.
DR: It’s very hard for businesses to get the right structure if you’re very established. We’ve been around for 146 years and there’s lots of data flying around the building. The people who have been gathering it aren’t necessarily data-oriented or even business-oriented. We’ve found we don’t have problems hiring people to do data. Most businesses have those people within their four walls. You just have to look for them.
NB: We’re relatively fortunate in that we have a central data analytics team that’s not huge but is full of very talented people. You can collect all this data but it’s the insight that’s really valuable and this is often country- or region-specific. If you have a centralised team, how do you leverage that value? You can have people sitting in London looking at this data but without the contextual relevance of being in situ where the data was collected, you lose some of the value. We have to find a way of getting people within our regions who are as passionate and skilled as the people collecting the data at the centre.
MW: Should data be transparent and available across the organisation?
RB: It’s about anonymising consumers’ personal data. If you do that, you should be able to give access to customer data because what you’re looking at is just information [rather than personal details]. How can you gather insight if everyone can’t see it?
Nigel Vaz (NV): If you’re in a bank or an insurance company, there’s a lot of regulation around PII (personally identifiable information). There are very specific guidelines around who has access to PII, in what capacity and at what level. Access and permissions is very much a function of the role.
SK: Visibility of data is key and the level of information we provide these business users with is incredibly important. A call centre agent will clearly need a lot of access to data. An analyst or a business insight generator doesn’t need to know someone’s name and address but they need to know about an individual’s behaviour. It’s all about the structure.
There is a growing sense that technology needs to accommodate the right to be forgotten
MW: How do you gather your customer data and are there any barriers to collecting it?
DR: We gather data from many different sources. We take over data sets from government so that we can push back to them to prove how we work as a charity. Getting frontline staff and volunteers to understand the importance of data and how to govern it to ensure that the data we do collect can be used is a challenge. There’s nothing worse than people collecting a whole load of data and suddenly finding out that a box hasn’t been ticked and it’s useless.
NB: Part of me says consumers are smarter and more sophisticated than we think and they get that this is a transaction and that we will then use their data in some shape or form. But on the other hand, you get anecdotal feedback that tells you they think they hadn’t realised X or hadn’t really thought about it in that context. We try to be as transparent as possible. We think about it as a value exchange - if we’re going to ask for data, then we need to be giving something back.
MW: Is it getting harder to get data out of customers?
RB: I think the law in the UK is going to change and you will have to be able to show exactly what data you have on a consumer. So whether you like it or not, a consumer will be able to control exactly what data you hold on them. We are going to be all over this because we want to have a really good policy. We want to make sure a consumer opts in rather than out.
SK: Lawyers need to make sure that the rules are in place and that we follow the rules. But the rules are not just there to stop us from doing the wrong things, they’re also about putting the best practice in place. It’s about enhancing trust and value. Someone will give us their data to get that value exchange but they trust you to give value to them. They trust you to ensure the security of that data. We become the guardians of that trust through the ethical use of data.
NV: Many brands don’t realise how hard we make it for consumers to share information willingly. The value exchange is one thing but the ease of gathering is another. We worked with a glass cleaning client, which wanted to know how long it took for a bottle of cleaner to run out and what it was used for. We got 15,000 households across a very large geography to opt in to help us answer those two questions by inserting a little chip at the bottom of the bottle that was activated as soon as it was swiped through checkout. When the fluid in the bottle dropped it sent out a message saying the bottle has been emptied.
A smaller group of people opted into a visual survey for the product. We inserted a pinhole camera at the head of the bottle so every time it was squeezed, it would take a picture of what it was being sprayed on. It’s a value exchange but it’s also ease. They don’t have to fill out forms, no one is sitting in their house to do ethnographic research and observe them. It’s easy.
RB: If everybody knows that the chief executive is supportive of data capture and analytics, it’ll definitely make a difference. You need to have measurements around compliance and data capture for those frontline staff. They need to be rewarded and recognised off the back of it so we talk about commission but it can also be soft rewards and recognition. The retailers who have done that have seen data capture go from something like 20 per cent to 78 per cent or 79 per cent consistently.
NV: Established industries such as finance or telecommunications have been using data and CRM for years very proficiently. These established organisations are being disrupted today because consumers move across the entire landscape in a very seamless way. Companies might have all my phone calls in a NCR Teradata warehouse that tells them how much money I spend but they don’t know that I’m browsing for a new phone on the website.
SK: Data has been captured for God knows how long but the people who are driving data usage has changed. Twenty years ago the IT teams had ownership. That focus has now shifted - marketers are driving the usage.
MW: How is digital data bleeding into the physical world?
NV: There is no division between physical data and digital data. Our latest insight from digital technology used in store was that most consumers in a supermarket, particularly in North America and Europe, simply ‘revolve’ around the shop. Ninety per cent of their time is spent on the periphery of the store instead of zig zagging through the aisles. When customers go down the aisles, they’re looking for specific things. All the thinking time seems to be spent in the outer rectangle. Those aisles are too narrow to hang around in and think yet so much marketing is done in the aisle.
TB: There used to be very segmented budgets at retailers, so anything to do with IT or technology went to the IT department and they budgeted for investment every year. Martech research recently showed that the IT budget is now fragmented and diluted. Now 30 per cent of the marketing budget is spent on IT. It’s interesting to see how the responsibility of outsourcing and delivering technology has changed.
MW: Can we trust social media data?
DR: We’re experimenting with using it to ask people for donations. For me something like LinkedIn as a source of social media data is brilliant because if I can get someone’s job title I can look at how much I should ask them to donate. One of my biggest concerns with all data, not just social media, is hype for vanity metrics. A Facebook ‘like’ to me means nothing.
If someone starts to complain about something, we can take them offline to talk to them and find out what their issues are. In the charity industry I think that’s really important.
SK: The whole social media element is operational. It’s very much a real-time piece that says, as soon as someone says something negative about my brand, how do I interact and turn them around so that it doesn’t balloon. When the O2 network went down, the company focused on social media and its management of customer expectations was fantastic.
NV: Big data sets such as Facebook and LinkedIn are bringing the worlds of broadcast and narrowcast together. These data platforms are now so big that companies can access an audience of more people than they could using a Super Bowl slot but in a targeted fashion. For example, targeting a 15-second video at IT managers on LinkedIn. This will become more interesting. I don’t think it’s tipped as a trend yet.
NB: It gives us a focus group on a scale that we’ve never had before. We have 3 million Twitter followers and 1.5 million Facebook fans. For a long time, if you asked us what the role of The Economist brand was, we would have said something like: ‘It’s about telling people who are extremely busy about things that they didn’t have time to form their own opinions about.’
Over the past few years, we’ve had to fundamentally rethink that because when you look at what is going on in Twitter and Facebook and how people are interacting, some of our most passionate customers disagree as much as they agree with what we write.
We actually found that The Economist’s role is not to tell people what to think, it’s just to ask people to think. Our UK ad strategy has changed as a result. We have a ‘Where do you stand?’ tagline that represents two different sides of a topic. This is a reflection of the way social media has changed and how we think about our brand.
MW: What new trends do you foresee in new data over the next five years?
AA: I’ll be intrigued to see if the idea of carrying your data physically on an implanted chip as Kevin Warwick [cybernetics professor, University of Reading] has suggested, takes off.
RB: I think consumers controlling who sees their data through technology will be interesting. We’ll have the tools to follow the consumer but they’ll have the power do decide who sees what.
TB: Consumers are going to get a lot more savvy and realise that there are a lot more ways of capturing data on them and they’re going to want some value exchange for that. There is also the growing sense that technology needs to be able to accommodate the right to be forgotten.
NV: In the next five years it will be about rewriting the rules about the basics of data: trust and security. But with a marketer’s hat on, the hardest job in data will be the right questions to ask. What you want is that piece of insight to generate value. The art of asking and framing the question is going to be the single biggest thing we have to innovate.
AA: Global data throws up lots more problems. In the Middle East people’s names change. In the past, I’ve had difficulty knowing who the customer is because his name is very different to the one in the public arena.
RB: The worst thing that data can do is follow the digital path where all these agencies and specialists have been created. I would hate to have a whole new set of people in marketing in five years’ time who are going to work with you and tell you what the answer is. If you’re a marketer, you now need to know and be excited about digital and have those skill sets and the same goes for data.
NV: I think siloing it is absolutely the worst possible outcome.
MW: Where should marketers’ focus be?
TB: It’s still data. Don’t get carried away with the hype. Ensure the data is shared across teams and embrace this. We don’t want someone in several years telling us what we need to do.
DR: For me it’s to continue to work on creating a data sharing strategy for the organisation. My toughest problem is who gets to see what data. I can’t show everyone everything - we’re talking about children’s data.
There has to be strong governance. But yet there are insights that are a necessity for other people in the organisation to see. How do I create a better sharing strategy to get better insights without compromising integrity, for example? I also need to ensure the systems we do build don’t lock us down to collecting more data in the future and integrate new systems.
AA: The focus should be on getting the basics right and protect the customer, their data and your brand.
RB: Marketers should identify what are the key insights and how to marry data sets across the organisation. Above all, don’t try to boil the ocean gathering everything possible and getting lost in it. Governance and privacy will be key.
NB: It’s about asking the right question and putting structures in place to own data within the organisation to reduce the risk of creating silos.
SK: I would focus on the value exchange and ensuring that we provide the level of service that our customers would expect from us. That will give us the ability to ask for more information.
It’s about getting the basics right and not getting absorbed in the hype. The future is another world: in five years Facebook could have disappeared replaced by something completely different. If we get data basics right now, when new ideas come in we’ll be ready for them.
NV: The idea that a digital fingerprint exists is something that we have to recognise more and more. Then it’s much easier to get privacy security governance and legislation right. Regulation, industry and consciousness have to cater for the fact that some industries and countries are operating data 5.0 and others are at 0.1 at the same time.