Home > Poslovno svetovanje > Področja dela > Operativno prestrukturiranje > Bjorn Bloching, Lars Luck, Thomas Ramge: In Data We Trust, How Customer Data Is Revolutionizing Our Economy

Bjorn Bloching, Lars Luck, Thomas Ramge: In Data We Trust, How Customer Data Is Revolutionizing Our Economy

Computers know us better than we know ourselves. Or at least they are often more reliable than we are at saying how we will behave in certain situations. All this is possible with growth of data. The smartphone removed distinction between offline and online world. Translating our day into a stream a data. We are entering into data-driven marketing business models and this demands evidence-based management. And since we live in a flat world, where technology, processes and business models are copied mercilessly, it is even more important to create value outside products and to be a front-runner in usage of modern evidence-based management. This will lead closer to customer via data. Companies with offline backgrounds can learn a lot about understanding customers from the digitally driven newcomers.
Three groups are important for success of data-based marketing: workforce, marketing service providers and partners and customers. Digital natives get that easier and they have more insight what is logically possible and economically sensible.

At the heart of grocer’s business models is a value of a customer in the profit which can come from a long-term customer relationship. This was a base on which corner stores where made and personal relationship with each customer was created. Customer life time value – total value of a customer over the length of the business relationship and customer equity – the value of the customer to the company’s future revenue should be main KPI for many businesses.

According to Robin Dunbar, most people can maintain relationship with about 150 people, but someone’s head is a cognitively limited CRM system. So, development started from small store, to mass marketing and hypermarkets, to professionalized marketing, use of CRM, online marketing to integrated, data-based marketing. With growth of mass marketing, brand awareness rise. But traditional marketing has problem with ROMI – return on marketing investment. It measures activities and not effects. Every retailer knows that less, but more focused, advertising would be better for everyone. But out of fear that the “advertising pressure” from the competition could be stronger, everyone carries on as before.

When mass marketing upgrades back to single customer relationship, that can only happen if you use all available data points. Contract data, coupled with usage data, face recognition and the evaluation of user’s Internet browsing behavior. When we started data related projects, it is important that we ask the most important question first. What key data is needed for the company to generate profitable growth. This is often asked too late. Those projects should not be only technology driven. Sales and marketing people have to learn to collaborate better on data-based marketing activity. On the other hand, developments in technology are enabling great laps forward in big data analysis. Technologies like in-memory applications (temporarily pull data from fragmented databases into a separate level and analyze them there with real computing power) and Hadoop-based solutions (high capacity generated by linking many hard drives) are helping in this. Datameer is a company that thanks to their open-source software, Hadopp, they can promise to structure and re-structure data – for example social media data – in almost any quantity according to whatever parameters you wish, in a short time.

When pushing customer to cooperate in this data acquiring activities, we should be careful not to push too much. Usually customers don’t want to be part of more than three bonus programs. Generating added value for sales and marketing department from complex data sets remains a difficult but important task. That is why sales and marketing need to be involved into all steps of process, technological also. Good systems prove themselves by their ability to filter out of large amounts of data the few really important bits of information that are relevant for marketing and sales activity. Many customers say that individualized communication is a relief. When it works, advertising becomes information – which is when it is effective. That is exactly the point of targeted marketing.

With rise of data-based marketing approaches, many borders are disappearing, one of them is difference between online and offline retailers. Real potential for growth is in the convergence of e-commerce, social shopping, online payment systems, cloud applications and the very physical shopping world in city centers and shopping malls. Goods follow data. The key to success here is that marketers compare individual customers’ characteristics with the logic of known consumer segments and derive their product suggestion from combination of the two. Behaviour beats context. Companies target groups in their marketing efforts. Real-time analysis will become the killer app. The fine art of marketing in a cloud economy will consist of hybridizing data from consumer segments, individual behaviour and location.

Marketing guys can analyze everything: from values, needs, preferences to price elasticities. User data are usually scattered around different databases. In order to link them intelligently, we need time, intelligence and computing power.

Two sort of data matters:
• data about behavioural drivers – demographics, psychographics, needs
• data about the customer’s actual behaviour – information search (passive and active), response behaviour to marketing, advice, conclusion, usage, buying additional products, complaints/customer services, recycling, purchase history.

When lookin at behavioural data, we need to observe also things that are not seen – like competition purchasing, complaints to friends.
Goals that are underlying data-based approaches to market potential are:
• Obtaining an improved understanding of customer value through whole lifecycle – CLV.
• Obtaining whole market picture.

CLV – Customer potential (gross profit) x share of wallet x duration of customer relationship + value of recommendations – cost of keeping customer loyal. Customer value is calculated from revenue (discounted on time basis) minus the cost if acguisition and retention (discounted on time basis). We need to take into consideration also cross-sell and up-sell potential and ability of customer to bring new customers through recommendation. Researching customer behaviour is not a black magic. Companies need to analyte customer interactions all along the customer journey.

Proper framework for marketing activities on individual basis is:
• Identify the market (potential) on the level of the individual customer.
• Increase share of wallet of the individual customer:
o Sales optimization in customer existing product category – up-sell
o Ousting the competition
o Substitution of spending on other products
• Customer retention – increasing length of customer relationship
• Increase customer recommendation
• Lower marketing cost, increase ROMI

Getting to know customer better, predicting demand via »predictive modeling« that uses internal and external data, optimizing logistic, range and price with the use of acquired data. This is what turns retailer into what Tom Davenport calls an »analytical competitor« – a markete player who builds his strategy and growth on the intensive use of data. In the era of ubiquitous computing intelligence and an economy based on cloud computing, the marketing mantra will be to use customer data wherever it is. And customers will also appreciate the strengths of all sales channels.

One of good examples of data usage is German family company Dodenhof. They have large shopping center. They have 450.000 card holders. They create 140 mailings each year. Their response rate is 20%. 8 % conversion ratio. Their consistent approach brings response rates that is better than industry average by factor of eight. Some of good examples are also in internet banking companies. Another industry is air travel, where they bring loyalty cards, started to segment users and offer better deals for most profitable customers. But is was automotive industry that introduce segment oriented offers first. Some of industries build on binary code like telecoms and IT can have large difficulties in handling data.

Some of potential areas where better data handling can bring enormous benefits to telecoms are: data-based optimization of end device subsidizing, data-based focus on sales offensives, data-based customer retention in prepaid markets. Some good practices are to allocate marketing budget based on different market potential. In prepaid retention one project look for patterns in churn of prepaid user and see that if new numbers that belong to competitor is starting to appear in customer network, chances for churn increase and if customer is getting one call dropped per day, they will churn 100%. In industry where customer retention costs are high and margins are sinking, marketing should be seen as customer service.

When introducing data-handling projects it is important that users, mainly marketing and sales people understand technology behind it. If they can see positive cost-benefit ratio in project, they will support it. If you crunch the data, you will survive in market. Whoever want to be successful in the future has to link data through all the channels. Socialized media analysis is a great opportunity for personalized online advertising, but it is even more valuable for business models that involve personal contact with customers. In indirect sales (B2b and B2C) use of data-based marketing is beneficial for both manufacturer and traders and it is important that they cooperate and share data.

To make sense of data, we need to use randomized controlled trials (RTC). They are perfectly designed to set simple question and receive simple answers, in order to provide causality. That is why they are so important for data-based marketing. RTC has become a standard tool in the optimization of website functions and in online marketing. They can also serve as great tool for investment justification. But not a lot of companies use them. In study by Mark Jeffery this number is estimated to 30%.

Change in marketing approach is change from audience reach to expected return. First attempt to calculate ROMI was made two decades ago by Philip Kotler. Ten year later Guy Powell popularized the term. Challenges in calculation are big. One of them is that turnover is hard to pin to some of marketing efforts, since customer is influenced in many ways. Advertising effectiveness can be measured if we define KPI and work consistently with control groups from the same segment. Some of the KPI are:
• Activation rate – proportion of target group that bought at least one product.
• Sales uplift – difference between control group and target group.
• Incremental revenue – turnover effect comparing target and control group.
• ROMI – incremental margin to cost of marketing activity.
To have proper use of control groups, they have to be representative and big enough.
Evidence-based methods are important and can bring more understanding into other areas too. Areas like pricing policies and channel choices.

One of the main transitions that companies are working on is how can data-rich companies become customer-focus organizations. They should combine value and need oriented segmentation with transaction data. They should move decision authority from buying departments to sales. They should reallocate resources for customer analytics. If companies decide to move into directions of changes, they should be careful of their structure. Structures block changes. Change managers know that only too well. And sometimes structures also disable data. If they use outside help, consultants should accompany the strategy implementation. And of course, data-based marketing can be the decisive factor in this transformation. It is important that transition begins with quick success, which increase the desire for more.
The intensive use of customer data does not happen from the bottom up. In customer-data-driven organizations, the executives are the drivers of analytics applications. They have to pick up technical knowledge so that they can ask the right questions and quantify progress with the right metrics. They are present in team meetings and keep motivation up when implementation difficulties arise. Two-thirds of marketing decision-makers admit that they do not have people who can interpret complex customer data sets. Sometimes management capacities are also obstacles in development. But with growth of cloud-based solutions and development of available software solutions companies can get on customer-centric journey quicker and easier. Centralization of analytics processes brings improved data consistency. Introducing data-based marketing can serve for structural re-organization too. They could bring some changes to recruitment and can be supported by good communication and use of best practices. Sometimes new incentive calculations are introducing, which are measuring not how much activity is put in, but results in comparison to do-nothing scenario. CEM (customer experience management) field is introduced.

When looking for new partners and better customer engagement in marketing activities, new goals are set:
• Connect
• Create
• Communicate
• Commerce
One good example was McDonald’s campaign where users could create their own burgers online and the best one – was introduced and rewarded. In future few companies will design their own campaigns, but they will want to steer the segmentation more directly as well as individualized the message.

Some interesting changes in this field are coming from change of data collectors to data buyers and sellers. MediaSift, Gnip and Sysomos are young companies that have set themselves the task of filtering out of the data streams of social media the most economically relevant information. Sysomos finds out for companies in real time what the Twitter-sphere thinks of product. MediaSift goes a step further. Marketers can ask which male Twitter user of a certain age from a certain town is likely to be interested in a certain product.
With data intensive activities, trust is important for partnerships. Retailer are afraid that manufacturers would go directly for customers if they would new who they are. So, for projects of sharing data to be successful manufacturer need to provide retailers with answer on what is in it for them.

Changes will also affect work of marketing agencies. The classical agencies are still the best at creating brand claims and can conjure up the most emotional images. Agencies with a technical background, on the other hand, are masters of one-to-one communication but often lack conceptual originality, precise language and quality graphic design. One of those new type companies is MIT spin-off Locately. They are attracting customers with vouchers if they share their mobility profile with them. For their clients they combine data from mobility profiles with Nielsen data and client’s own campaign. They are still not collecting data about exact customer purchase, time and location of purchase.
One of promising new approaches is called foresight thinking. It aggregates customer data in the form of consumer biographies and works out which products someone who today is 2 will probably want to have in 10 years. Another field is visualization of data. Information visualization is information compression.

Growth of data-based approaches will put more pressure on handling privacy issues and building proper transparency and trust. Customers need to be informed beforehand and possible even be asked to give their agreement. In order to get customers approval for use of data, companies will need to show clear benefits to them. Marketing need to be open about their handling of data and communicate its advantages to customers. In transparency-benefit matrix we have: unwanted spies (low on both), tolerated advertisers (high transparency, low benefits), tolerated spies (low transparency, high benefits), partners (high both). Marketing will develop into sphere, where it could be offered as a service. In order for that to happen, the following elements need to be present: relevance, frequency (proper), added value (special offer for each customer). In 10 years’ time question will not be which brand come to mind first, but which companies do you allow to use a lot of your personal data.

In development of regulation, we have different approaches. One idea is informational self-determination which basically means that the adult user decides which personal information about him is stored for what purposes. But the problem of abundance of all this is, that as a consumer, we have completely lost track of our opt-ins. Government should be smart when implementing data protection laws and regulation, since too much regulation can stop development. It is not a politicians task to create new law for every new problem. Balance in this situation and trust will be developed like in every relationship, if you believe that your partner is listening to you and try to understand you, that will increase your feeling of goodwill. If you have the impression that you are being spied on, then the relationship is over.
New deal on data should be based on: data security (data should be stored safely), transparency (customer want to know what kind of data and for which purposes are stored), added value (for customers) and proportionality (only data that really bring value to customers should be stored). Regulation, customer acceptance, certification and marginal utility will further strengthen these four pillars.