big-data-better-data-white-paper-final

Big Data or
Simply Better Data?
A Short Guide to Big Data
Project Implementation for
Heads of Marketing

| Big Data or Simply Better Data? ©ZetaInteractive 2015
Contents
03 Introduction
03 Definition of ‘Big Data’
05 Collecting the right data
05 Engaging the right stakeholders
06 Turning Data into information
07 Turning information into optimised marketing programmes
09 Turning optimised marketing programmes into contextual
dialogues
10 Business Intelligence and performance measurement
objectives
11 Summary and conclusion
With key points summarised below each section

Introduction
Let’s face it the phrase ‘you need to implement a Big Data project’ is scary. It usually
sends Heads of Marketing like you or I, running for our budget spreadsheets, trying
to figure out how to cover the cost.
So is it all really worth it?
Well... yes. If approached in the right way, the results can be dramatic, enabling
your business to make better decisions. And better decisions mean increased
sales, higher open and conversion rates, stickier content, the opening up of new
opportunities, greater operational efficiency, cost reductions and reduced financial
risk.
Which of us wouldn’t want to go to their Board of Directors with that?
But the key – as always – is ‘Preparation, Preparation, Preparation’, so think of this
as a user-friendly guide, offering some welcome pointers to ensure your project is a
success.
Definition of ‘Big Data’
What does big data actually mean?
‘Big Data’ is simply a term used to describe the creation of a large dataset or Single Customer View.
This dataset is often consolidated from many sources across a business, and is then exploited to reveal
patterns or trends, especially relating to customer behaviour and interactions.
This then enables you to send targeted marketing communications to your customers based on any of
the attributes available in your data, or trends that are identified.
In essence, businesses embarking on big data projects are trying to extract value from their data. By
being able to access more tracking points you can determine customer behaviour more accurately.
The insights gained from this behaviour can be used to improve customer acquisition, up-selling and
retention rates, or providing better management information to understand how it is performing
against specific key performance indicators.

The recommended approach to embarking on any ‘Big Data’ project can be described in these steps:
What are the various data sources available
across your business or third parties, that need
to be included?
How reliable and accurate is the information
these data sources provide you with, and how
likely are they to change?
What value does the data offer to your business?
e.g. Communication triggers, specific KPI’s or
response and interaction tracking
How does the data link to other data? e.g.
Customer ID, Email Address, IP Address,
Mobile Number
How does the data need to be presented to meet
your user needs? e.g. Rolled up, aggregated,
structured, or categorised
Are you able to prioritise those that will offer
the biggest impact? e.g. Drive strategic
revenue, or more accurately measure results
Where is this combined data needed to be
exposed? e.g. Call Centre, CRM, Business
Intelligence, other?
Will all those data sources accessible via web
services, or automated incremental data feeds,
and has IT provided timescales for provision?

Collecting the Right Data
Big data projects fail for a number of reasons – inconsistent data capture, data accuracy, resources,
budget, consolidation or storage – but all too often, it comes down to whether the project was trying
to solve the right business problem from the outset.
So rather than focus on “Big Data”, refocus your thoughts on “Better Data”... on pulling together
the right data and making it actionable. Think about the data you need to drive your marketing
programmes and make yourself relevant and timely.
By using this approach, you still see a 360˚ view of your customer, but will make the data accessible
in a way that helps you drive relevancy when targeting your communications, without the need to
wade through an enormous dataset looking for value.
The same rule applies to using the data to fulfil your management information and reporting needs -
it’s about identifying how you use the data, not taking months to learn something from it.
Key Points
1) Always define what you want your Big Data project to deliver beforehand.
2) Focus on what data will give you the best results, rather than what data you can have. That
way you will see more immediate insights and results, rather than trying to search for the
insight first.
Engaging the right stakeholders to
ensure successful delivery
Marketing usually own these projects, so typically the marketing team will begin by defining
requirements, but input is absolutely required from other key stakeholders to ensure it is successful.
Involvement from the IT department is essential for a successful implementation and these and other
stakeholders such as Board Directors and Heads of Customer Service might have opinions over what
the definition of success will be, so it’s important that buy-in is sought early and definitively.
From a technical perspective, it is vitally important to understand what resources are available within
your IT department, and the level of skills you have at your disposal.
Sometimes different people own different solutions even within departmental functions, so several
resources may be required.
In addition, third party sources like Web Analytics data, are often owned by marketing and
implemented, managed and maintained by the IT department, or even an external agency.

A project steering committee should be established where plans can be presented,
and key data sources uncovered as early as possible, so you can structure roles and
responsibilities, ensure those involved have an opportunity to input their availability,
set timescales for getting hold of data, and help to define realistic timelines for the
project – or more importantly, highlight those areas which need more research before
they can be included.
Key Points
1) Set up a project steering committee that meets on a regular basis.
2) Identify all possible stakeholders, their departmental objectives, success criteria
through the steering committee and gain consensus on project goals.
3) Define resources, timescales, and identify risks and mitigate for them.
Turning data into information
Often data consolidation is the (relatively) easy part. Determining methods that can
best exploit the data, and turn the data into information offering real insight into
your customers and prospective customers can be the real issue.
For example: the use of predictive analytics.
What mid-tier marketers really need to know are, for example, their acquisition
and retention rates, the offers consumers are responding to, what subject matter
or content they are interested in, and if and when they purchase products. And
whilst sophisticated Web analytics tools are available, the obstacle with predictive
analytics is not the availability of easy to use modelling tools; it’s the limited
availability of skilled analytics professionals to pull the insight out of the data. The
fact is that large enterprises routinely have in-house analytics teams, where-as mid-
tier companies can rarely afford to fund these positions.
Organisations that do not have the skills to perform this level of data analysis and
segmentation in-house should start by ensuring their big data project has out-of-
the-box embedded predictive models.
Cost-effective and easy-to-use, embedded predictive models will help organisations
turn data into actionable information. Pre-built, industry-specific models are
available already and provide compelling, cost-effective alternatives to employing
a team of analysts. Predictive models - such as likelihood to convert, time to next
purchase, likelihood to attrite and next best product - can provide the most accurate
and up-to-date customer understanding.
Descriptive models, such as RFM (recency, frequency and monetary value) and
customer categorisations such as new, lapsed, lost, reactivated, core-stable, core-

grow, core-decline, core-best can help the marketer to better target and time offers and
messages.
Key Points
1) If you can’t afford to employ in-house data analysts, ensure your technology supplier
provides you with out-of-the-box predictive models, so you can be up-and-running and
making ROI as quickly as possible.
Turning information into optimised
marketing programmes
So you have your data, and you’ve identified customer segments, and you have your models
to enable relevancy and timing of offers to customers. Now it’s time to look at the customer
lifecycle.
Typically, any Customer Lifecycle involves some variation on Acquiring, Growing, Retaining and
developing Loyalty and advocacy with your customers. Beneath these strategies are initiatives
that will help you underpin successful marketing programmes.
For example – let’s take a look at the role of Retention within the customer lifecycle.
It’s usual to receive emails from companies telling you how much they have missed you. This is
typically a triggered campaign, defined by predictive analytics that put you in a ‘hasn’t engaged
in a while’ category. Put simply, you were a good customer from the businesses point-of-view,
but you are now overdue to make a purchase according to your previous buying history. Other
tactics are more real-time, such as marketing to customers who abandon web purchases. The
targeting of messages can be achieved both on site – for known / unknown visitors based on
browser inactivity; the mouse passing over the browser address bar; and also on close of browser
– with a follow up email (often referred to as remarketing) or SMS message post leaving the
website (known visitors).
Think about those terms again which depict how typically businesses want to improve their
marketing plans: acquire customers, grow their value, retain their custom and develop loyalty.
Think about your current marketing strategy and the campaigns that you run today.
Are you covering each of these effectively?
What are the grey areas you are not fulfilling on? What programmes could you create from your
new dataset that bridges the gaps?
What additional programmes could help to grow revenues?
Technology plays a part in this process, as limitations around customer touch points and channels
will impact your ability to deliver messages and communications, irrespective of having the right
data available.

Every comprehensive email tool today will enable dynamic content / offers, and the more
advanced enterprise email solutions will enable testing on a sample of your campaign audience,
then after a designated time period (e.g. hours, days) automatically roll out the winning
combination to the remaining customers. Double-check that the solution you’re looking at will
do this as not all will.
Another component is how customers access communications.
With a significantly higher proportion of emails opened on a mobile device, as opposed to
a desktop, marketers need to also consider creative design. “Responsive design” refers to
designing emails that render well when they are opened on a mobile device. It gives brands an
opportunity to optimise the user experience.
The impact of responses can be dramatic. With some businesses seeing average click through
rates increase by as much as 178% over non mobile friendly emails. Again, your existing email
service provider should provide you with the capability to preview your emails in a wide range of
desktop, mobile and web browsers to ensure that your emails render correctly.
Key Points
1) Identify the gaps in your Customer Lifecycle, and ensure the dataset you are pulling
together is going to help you fill them with proactive marketing campaigns.
2) Plan for Evolution. Your project is going to provide you with enormous scope to test and
trial on-the-fly like never before. Ensure you have the right human resources and mindset
to seize the opportunities.
3) Build mobile into your plans.

Turning optimised marketing
programmes into contextual
dialogues
Let’s turn to other data techniques that can be used, leveraging data not necessarily stored in your big
data pool, but data that is driven by technology solutions.
Marketers now have access to various channel tools that can help increase offer and message context
to make them even more attuned to their customer.
For example: mobile devices.
With the introduction of so many apps, a new and ever growing marketing channel is available with
the ability to push notifications directly to mobile devices. More advanced push capabilities involve
rich in-app messaging – essentially creating a mobile email inbox – with notification of new messages.
This approach effectively removes the ‘interference noise’ from your competitor’s, so it’s easier to get
your message acknowledged, and get your brand messages recognised.
Beacon technology has also aided in deploying more location based targeting techniques, enabling
marketers to detect app users as they enter a location such as a hotel or store, and pushing tailored
offers to them. So this is incredibly important for hybrid bricks and clicks retail marketers.
Open-time optimisation is also a growing phenomenon with advanced marketers.
Open-time technology tags images, and these tags can be dropped into email templates used to
send your communications, so the email shows the most appropriate image and offers based on the
customer’s location, and the device and app they are using.
Examples of how this can help include: Providing only offers which are in stock / are available;
Providing a link to iTunes to download the app (as it recognises you have an iPhone); nearest hotel
or store to their current location, plus a load more methods to contextualise the message for the
customer adding more value for them in the information that you are providing.
Real-time decisioning (sometimes referred to as ‘in-the-moment’) technology can also help turn every
customer interaction with your business into a unique sales opportunity, or even an opportunity to
provide even more value and enhance the overall customer experience.
These technologies typically provide “data connectors” which listen to your various customer
touch points, such as your web traffic, social media posts, etc. and then leverage business rules
and predictive models (as described earlier) to decide how to respond to the enquiry. The decision
typically results in an action, such as a tailored offer presented on the website, or perhaps firing off an
email to the marketing manager notifying them that there has been a negative Twitter post that needs
addressing.
These solutions are not typically driven by an underlying database, however, they are still driven by
data, connecting to the various data sources required to make the decision.
Your big data project may provide valuable historical personal information, or even what the last
marketing communication was that a person received, whether they engaged with it (i.e. opened or
clicked an email) or how long it has been since they last made a purchase.

Key Points
1) Identify whether contextual marketing is something that is right for your business, as the
importance to incorporate into customer dialogues vary by market.
2) Real-time marketing can be costly, so be aware of the importance of real-time versus the
value it will bring. Ask yourself if you will still get a return by driving campaigns out within
hours, or is it just a nice-to-have?
Business Intelligence and
performance measurement objectives
The previous sections focus on the marketing program side of leveraging big data, but what if your
focus is on business intelligence and reporting, either instead of marketing?
Business intelligence big data projects put an emphasis on measuring performance against goals and
establishing accountability for reaching them. As with marketing-led projects, the goals and metrics
have to be aligned with internal stakeholders, and definitions agreed. For example, commercial
metrics might be interpreted differently between sales and finance teams. Throw marketing into this
mix, and you have an additional three types of users: strategic, tactical and operational. This might
impact the data required to support them and should be discussed as early as possible in the project
to ensure the right data from the right systems is incorporated to meet their goals.
Similarly, the data accuracy and quality assessment side of any Business Intelligence project is equally
as important. You need to consider bringing data in, in a structure that is not too complex for users,
and designing the dataset to focus on high priority metrics that will deliver the most value.
As with a marketing led project, it holds true that if you introduce too much complexity or too much
data, it remains exactly that, data and not information.
When delivering a project with different departmental requirements, think about creating alternative
views of the data. This will enable simplified snapshots of the data, tailored to the user audience, and
enable them to address their core needs, accessing key data without a steep learning curve.
Key Points
1) Consider internal stakeholders requirements for reporting, and the different use such reporting
will be used for, as well as differing views over the meaning behind metrics.
2) Structure the data to simplify the user experience and avoid the noise of irrelevant or low value
data.
3) Consider multiple views to serve different purposes, whilst still providing the level of detail
necessary for measuring performance, or operational benchmarks.

Summary and conclusion
Big Data is a broad topic, but is now a generally understood concept that Heads of Marketing
acknowledge as holding huge potential, however, the challenge is to translate aspiration into
operational reality and ensure new data sets work hard for Marketing Departments and Brands.
There is no magic wand, but with the right support and commitment marketers are well placed
to use the simple frameworks and methodologies outlined in this white paper to lead a big data
planning process which is grounded in the realities of their customers, products and brands to:
• Establish stakeholders, create a big data steering committee
- Understand departmental data requirements and definitions of metrics
- Understand departmental perceptions of a successful project
- Determine required resources and timescales for access to, and provision of data (both
internal and external sources)
• Understand and identify core data sources
- Establish their value and abilities with regard to your objectives
- Audit data sources to assess quality, accuracy, cardinality and population
- Prioritise those that will deliver quick wins, deliver the most value or develop a phased
approach for bringing them in further down the line
• Relate the data to your internal strategies
- Align with improving marketing programmes or management information reporting; clearly
defining customer lifecycles or core business metrics
- Leverage data to drive additional value from data, using predictive analytics to identify
opportunities and core customer segments
• Reconcile your data with technology capabilities, and determine how it
can be used to deliver additional value to your customer, and your business
- Sensing and testing offers via different channels or touch points
- Contextualising customer communications, to ensure timeliness and relevance
- Consider future channels, or touch points with which you can improve customer experience
Like any CRM and customer experience strategy development, a Big Data project has to be a
multi-disciplinary, business-wide undertaking but marketers must lead the way. Not because of
any political or organisational hierarchy agenda but to ensure serving the customer better lies at
the heart of a company’s strategy.
The aim is to empower Heads of Marketing to lead the Big Data strategy and planning process.

Foundations of Excellence in Leadership, Innovation & Service
15 years | 3 continents | 9 offices | 700+ Employees | Servicing 100+ clients | 20 languages
Zeta Interactive is a leading Customer Lifecycle Marketing
company that uses big data and proprietary technology to
acquire, engage and retain customers.
For more information on gaining competitive advantage from
big data please contact Zeta Interactive on the below number.
Ready to talk? Contact Zeta Interactive today:
Tel 01642 808888
Email enquiry@zetainteractive.com

big-data-better-data-white-paper-final

More Related Content

What's hot (20)

Viewers also liked (12)

Similar to big-data-better-data-white-paper-final (20)

big-data-better-data-white-paper-final