Data is the clue that binds all the change together. But who owns the data. As digital transformation has shifted power from seller to buyer, marketers find themselves in an endless race not just against their competitors, but also against the expectations of their customers.
Customer experience now feeds directly in brand perception, putting the onus on companies to ensure they provide seamless, useful and often delightful experiences for their audience.
But to achieve this they must first master one of the most bedeviling problems of the modern enterprise – taming the ever-growing mass of data they hold on every customer.
Long before the Internet, companies were already struggled to present a unified front to consumers – and back then all they had to worry about was their own customer and product data. Now brands have to account for second and third-party data as well, much of which is unstructured, such as that from social media.
Mike Stocker, senior director, business development at Marketo describes first party data as that data that a marketer has collected through their marketing efforts. Then there are various sources of second and third party data.
“Second-party data could be data that is found on channels like major social networks where they extend their data to their partners for usage. Third-party data is data that is gathered from outside sources,” he says.
The challenge of course if for marketers to collect all of this data together into a more singular view of the customer to be able to provide a more individualised marketing experience, he says.
Mobility has only added to the complexity with geo-location and weather data now factoring into the mix.
At a high level a typically company has to consider:
- Profile data such as customer records
- Social graphs such as how one person is related or connected to another
- Transaction data, such as purchase history
- Event data, such as logs
- Social data such as posts on social media
- Web browsing data
- Customer call data
- Mobile app data
- Location data
- Video data such as video data streams.
Luckily, according to Trent Lloyd, Co-founder and GM Publisher Development at Eyeota, first, second and third party data is often the same.
“When a company offers its first party data for someone else to use, it becomes second party data – if they broker it through someone else, it’s third-party. But it’s still the same data. Traditionally the data has been behavioural and proprietary. But now more information is being shared – collected from various sources to create large sets that can be extremely useful for marketers.”
One common problem says Lloyd is that often companies find they simply don’t have enough first party data to be meaningful.
And first party data can also be highly restrictive. “It might describe behaviour on a brands own website but it may not stretch beyond that to explain the user’s natural behavior.”
That why, he says, there is such a massive increase in the use of third-party audience data.
Whatever the source, ultimately all this data needs to be brought together to provide as close to a single view of the customer as possible and to provide insights into how customers behave, and importantly what they might do next.
Apigee director, channel and alliances, Chee Keong Law says the holy grail every company wants to attain is a single view of the customer which they (hopefully) reach by combining some or all of the data points about that customer.
“It’s not just about having the data but stitching together the relationship among the data, such as time-sequenced event data of a customer.”
He told Which-50, “As an example, if a company is able to detect that a segment of users are having particular trouble with a website feature through website analytics, they could proactively reach out to those customers rather than wait for them to complain.”
In recent years Customer Data Platforms (CDPs) have emerged as a way to help marketers build a complete customer view says Eyal Mekler, Tealium Managing Director, Australia and New Zealand.
The CDP is primarily a marketing system that creates a unified, cross-channel view of the customer.
“Historically these were batch, back-office programs, but as the communication landscape has become more diverse, marketers are moving to embrace real-time systems,” he says.
“CDPs take data from digital channels like email, website, mobile, CRM, display advertising and e-commerce sites, and harmonise or link the first-party customer data from each channel so you have a better understanding of the true relationship the customer has with your brand. Some are also able to take offline channel data feeds like contact centre and direct mail.”
The fundamental benefit of a real-time CDP according to Mekler is that companies get the unified customer view they need to develop consistent, relevant messages across all channels and create a superior customer experience. “Without this kind of insight, customers may receive wide variances in messages depending on the channel, and brands have difficulty delivering real-time relevant messages and offers.”
In addition to the harmonising of data and the building of a master customer profile, a customer data platform enables real-time action in any of the technologies a marketer is already using.
Umporn Tantipech, Principal Consultant, Teradata described sophisticated data management capabilities as the crucial foundation for establishing and continuing maintenance of SCV that will drive greater business value.
“In today’s Big Data paradigm, a single platform solution is no longer workable, especially given rapidly advancing technologies such as virtualisation, WaaS and cloud-based solutions.”
With Big Data’s enormous and increasing volume, together with more and more real time velocity and a wide range of multi-structured formats, the best practice is to store and process data on the most appropriate platform according to the characteristics of data in use, he says.
Since moving Big Data around to the executing applications is no longer viable for many organisations, the “Unified Data Ecosystem” architectural framework and its associated Push Down Query processing capability must become the norm for getting different technologies to work together, according to the Teradata executive.
“A critical lesson learnt from pioneers of Big Data SCV is the need for low latency because of the transient context of many customer insights. Consequently, ability to integrate new digital data sources into a central customer data repository is necessary but, in itself, insufficient to deliver optimal business outcome and value-add.”
He argues that fast ingestion and refinement of any data types must be complemented with the ability to perform advanced, exploratory analytics at massive scale, and then be able to immediately be applied with the derived intelligence for better business decisions and customer experience at the moment of truth.
Kevin Ross, General Manager ANZ, SDL ANZ expresses some scepticism about just how well companies are exploiting the opportunities presented by big data.
“Big Data has not really delivered. Well, it has, but only insofar as collating historic customer records and company operational history. It has not allowed companies to truly extract operational competitive edge because of those insights,” he says.
“When we talk single view of customer, what we really mean is being able to pull together all of the multiple sources of customer data across an organisation into one cohesive view, in order to actually do something useful with it.”
According to Ross, these days it is much more important to be able to access that single business view operationally, and marry it up with the ambient data in that moment – and together this becomes a single contextual view of customer. “Long gone are the days when executives spent weeks putting together dusty reports for the board to consider at their leisure,” he says.
“We can use these insights to pull the relevant content from our multiple content sources and present this to the customer in real time. This does not have to be difficult. Technology today is capable of delivering this.”
Of course delivering seamless, single and integrated views of the customer involves a lot of pain, complexity and occasional structural upheaval. And it requires a commitment from the organisation to drive towards that single goal, eliminating silos and occasionally felling internal empires along the way.
Even within a single department like marketing, silos can emerge, as can agendas of individual sub groups that don’t always align to the overall outcome.
Tealium’s Mekler cautions that in order to achieve a unified customer view, all of the technologies and applications that make up an organisation’s marketing technology stack need to communicate, share data, and have the ability to fuel real-time action.
“This is achieved by building a vendor-neutral data foundation that sits across all of those technologies and brings together all first-party data sources, including web, mobile, offline, and IoT.”
So long, simplicity
“It used to be about having a really good CRM system,” says Stuart O’Neill, sales director, SAP Hybris.”However today that’s not enough”
Customers, he says, don’t want to be managed. “They want to engage with the brand or entity and they want to do this at any time of day or night. And they are no longer passive recipients of services. “They want to determine what and when they experience something.”
“To provide this organisations need to go beyond CRM systems and integrate this seamlessly with a quality commerce platform, a great marketing too, profiling tools and sales and service tools.”
According to O’Neill, “They all have to play together as one seamless team of customer technologies.”
The rise of the DMP
As Which-50 reported in November, despite their fairly recent provenance, Data Management Platforms (DMPs) from companies like AOL, Lotame, or Oracle’s Blue Kai have caught the imagination of marketers around the world. A study by ExchangeWire and Oracke found that 90% of marketers reported their DMP implementations were successful; and that, 91% of marketers saw positive ROI from their DMP at 12 months.
AOL’s Senior director client strategy, Nikki Rettalick says DMPs play a strong role on two fronts. Firstly they offer a safe house for sensitive customer data.
And secondly they can ‘anonymise’ that data assigning a persistent ID that is non personally-identifiable information (PII). This can then be shared among marketing and technology companies. These partners can then target and personalise marketing messages and measure the customer journey for that particular customer.
“What often gets in the way is the ability to sync technologies to accurately identify the same user profile.”
And, Rettalick notes, “Being able to identify and measure user’s mobile activity is also key to seamless data management.”
Cameron Strachan, APAC Principal Solutions Consultant, OMC, says Customer Data Management technologies mostly fall into those that collect data, those that store data, those that provide an interface/ability to query the data and those that action that data for a purpose for example a marketing or service request.
He notes three in particular:
- Collection methods could come from web forms via a CMS or marketing automation system, DMP, tag managers, web analytics. The purpose is to purely collect data in the most efficient way. Many platforms both collect and store data, which must then be taken into context when a company uses multiple platforms.
- Databases store data – usually companies have multiple databases because most operational systems store their own data and additionally there are analytical and marketing databases created for specific purposes. As part of a marketing or analytical database, best practice is to have some sort of ETL (Extract, Transform, Load) and Single Customer View processes to ensure that the collation of data from multiple places is accurate and unified.
- Interfaces are used for specific purposes, such as marketing automation, reporting, dashboards and analytics and service or sales requests. The data in these interfaces should be connected in an intelligent way via a central database or ETL/SCV process so that different interfaces do not show conflicting information. Strachan says problems often emerge when companies use multiple platforms to achieve all the purposes required. Platforms generally do not interoperate so additional process must be put into place to ensure accuracy.
“These could be real-time API calls to synchronise data, ETL and SCV processes to clean and identify duplicate data. API in particular have emerged as a critical technology to help companies deal with disparate data flows.
According to Apigee’s Chee Laws “We see that data is the new currency. Take Uber for example. Every car ride results in new data for Uber to collect and this enables Uber to act as a powerful data platform. For example, Uber uses that data to run programs with a hotel chain such as Starwood, so with user consent, shares ride information with Starwood.
“Starwood understands more about the travel behaviour of riders from airports and may offer discounts or promotions to entice these riders to stay with Starwood. Uber essentially monetises its data. This data is collected via apps through APIs. Uber can then open up some of these APIs for third parties to be used for a fee.
Nike is another example of a sports apparel company leveraging its platform, in this case Nike + which has data on millions of athletes. It shares this data via APIs with selected partners to create new innovative apps, benefiting its users and adding even more data to its Nike+ platform, says Law.
The other role of APIs is acting as the “last mile” of connectivity, enabling actionable real-time insights through push notifications or alerts.
“Without APIs to connect analytics results to a user’s mobile app for instance, companies will not be able to take advantage of data in real-time. APIs is the oxygen that enables the digital economy. You don’t see it but it is everywhere and is essential to the functioning of the digital economy.”
Adobe’s APAC director of marketing Paula Parkes takes a wide view and says that any technology that supports digital experiences and the customer journey needs to integrate to form a single customer view. “Technology might include web design, analytics, mobile app development, digital documents, stock imagery and e-signatures.”
“Operational silos are a significant barrier to successfully using data and insights to deliver improved customer experience. Incumbent marketing solutions and dis-jointed integrations continue to be a barrier to holistic ROI insight.”
Alignment across an organisation will foster agility and nimbleness and drive business growth, says Parkes.
But technology is not the only impediment to better customer data management as Parkes explains. “Skill shortages also continue to be a barrier for marketers,” she cautions.
Each year the APAC Digital Marketing Performance Dashboard has found that a lack of skills is preventing marketers from broadening their digital marketing strategies beyond some of the more standard metrics such as website performance and click through rates.
“Important indicators such as churn rate and customer lifetime value remain a low priority. Further investment and better skills will allow marketers to leverage much more from the data to drive business results and demonstrate ROI.”
The buck stops where?
Another vexed question companies increasingly need to deal with concerns internal ownership of data. Simply, who owns the data. Functional heads like CMOs and Sales Directors typically controlled their own little data fiefdoms in the past. Or worse, they didn’t. CIOs set rules and guidelines often driven by compliance guidelines from risk managers.
Increasingly though companies are looking to centralise control of data under a single, or at the very least, a select group of executives.
Oracle’s Strachan, says the issue of ownership depends on the scale of the company. “Some companies employ a Chief Customer Officer or Chief Data Officer that specifically handles concerns around the companies’ data. However, typically, the owner should be split between marketing, IT, analytics and legal.”
He suggests that the reason a marketing department may not fully own the customer data is that mostly data begins out of operational realms – that is collected because of a sale or a query – therefore there may not be the consent to market to that person.
“IT has a vested interest to ensure that the systems are correctly collecting and storing data. Whilst analytical departments should have a total view of customer across all touch points. The legal department also needs to be involved because the way a company holds and uses data on people must adhere to local law.”
Tealium’s Melker on the other hand argues that ultimately the “owner” of customer data needs to be Marketing and perhaps the newer role of Customer Success (CCO Chief Customer Officer).
“Marketing typically owns the communication and engagement with customers, and as such needs to own the data. The CIO should be in an “assist” role to the CMO and CCO by providing the technologies and tools to enable collection, compliance and integrity of the customer data.”
Teradata’s Tantipech, meanwhile, suggests that a partnership between CMO and CIO is ideal where business initiatives drive the data that should be deployed to support analytics and marketing interventions, he says.
“Simplistically, the delineation is where marketers specify the desired customer experience plus the applicable business rules. IT works hand-in-hand with stakeholders to deliver the optimal balance between functionality, business value-add and total cost of ownership.”
To realise high performance, this partnership will also need to work within an ‘Agile’ delivery model instead of the traditional ‘Waterfall’ method. The effort required for this organisation alignment should not be underestimated but substantial rewards await those who can successfully execute on the theory.
SDL’s Ross argues that data management needs to be a core consideration for everyone.
“CRM, ERP, eCommerce, Optimisation capability, Social tools, Analytics and campaign orchestration tools and even BPM tools really drive and require access to customer data. Does it all need to be in one single source? Not always.”
He says it may be that the IT group is responsible for putting the platform in place to deliver a single view. “But these days CMOs are tooling up appropriately for their functions, and the key is that they are able to access and manipulate that data using their tools.”
Warren Billington, managing director, APAC, Signal shares this view. “Signal believes that any executive who touches the organisation’s customers – from the CIO to the CMO – has a role in helping manage customer data.”
Ultimately, he says, as data continues to become more important to enterprises, more companies will designate a Chief Data Strategist who will be focused solely on deriving value from data.
*This feature is a chapter of ADMA’s Arm’s Traders Almanac, a collaboration between ADMA, Which-50 and B&T. ADMA is a member of the Which-50 Digital Intelligence Unit. Membership fees apply.