The relevance of database marketing in the digital marketing age

Still striving to deliver the right message to the right person at the right time

As someone who has over 30 years’ experience of data-driven marketing I don’t see any great dividing line between the age of traditional direct marketing and the digital marketing age in terms of strategy. What has occurred is the introduction of new methods of analysis and communication which facilitate concepts that previously were difficult to achieve.

Techniques like Artificial Intelligence, Machine Learning and Predictive Analytics support concepts that have been around for all the time I have been in database or data-driven marketing, addressing the need to direct audience selection as well as personalization of the creative treatment, the expected outcomes and channels of response, leading to the construction of customer journeys and profitable relationships.

The adage ‘right message, right person, right time’ formulated back in the 1980s is as true a strategic objective today as it was back then.

The New Strategic Cycle

Traditionally, a campaign would be implemented using traditional communication media and this campaign would generate a response. The marketer had full control to the greater extent, of who would receive the offer and hence the response could be viewed as controlled as the respondents were in the main a function of the audience.

That response would be analysed, and the outcome would be used to fine tune the strategy and drive the next round of campaign communication (see Fig 1).


Fig 1 Traditional Strategic Cycle

Whereas previously the marketer could anticipate with some degree of control when a response would be generated, the digital world has reduced this element of control. A marketer can no longer engineer when a contact may wish to encounter their message, proposition or brand, or, indeed, come across it. Hence, there is a need to establish valid methods for acquiring data from customers and website visitors.

The digital marketing age has delivered not only ease of data collection and analysis in order to drive the commercial relationships and measure the outcomes, but an absolute need for interpretation of the data that finds its way into the database not only from campaign responses but additionally uncontrolled responses, enquiries and tracking from website visits, to establish its value to the strategy (see Fig 2) and provide information that will drive automated communication strategies through a set of strategic rules and a broadened selection of communication media, in real time.

Fig 2: The Digital Strategic Cycle

These new concepts also have implication on clustering and segmentation. Combinations of demographic and behavioural criteria have traditionally defined clusters. Once defined, these have been used to drive product development and marketing activity.


Cluster sets change their content and their direction; individuals join and leave as new information is learned about them and as the importance of the business rules inherent in the data relationships is recognised.

This means that the clusters must be recognised as volatile and allowed to be dynamic.

Their dynamism must be tracked, and the changes identified in order to keep the marketing strategy and communications schedule on course.

Marketers can find that markets for their products are dwindling, readership for their publications is flagging and response to their once attractive offers reduced. This may not be because the product is any lower quality, or the price has leapt; it can be that the customer profile once associated with that product or market has changed shape and has moved out of the target zone for that marketer. Has the customer turned 31 and so is no longer eligible for a Club 18-30 vacation? Has a child reached an age where its parents would no longer be interested in nappies?

These changes and new viable targets must be recognised to effect and maintain product and communications strategies. Early warning of changes in customer profiles can be provided from constant data feeds with triggers identified to drive personalization of proposition, communication channel and delivery.


Personalization is not a new thing. Forty years ago, the large direct marketers like Readers’ Digest and Damart were thrilling their customers and prospects by incorporating their names, their street or their town in the text of the multi-page letters they sent.

This basic personalization continues today, with marketing communications incorporating the same text variables. However, these days no-one is particularly excited by the fact that direct marketers can reproduce their name on the page or the screen, irrespective of how large the font size is.

Others go one stage further to annoy the recipients of their marketing collateral, by making either guesses or misguided assumptions about the recipients’ interests.

To personalise relevantly, you need to understand the context of your products and how this will fit with the context of your customers, unlike this Amazon blunder exposed by a customer last year on Twitter:

So, data-driven personalization should be a business priority. Research over the last three or four years has determined that conversion rates can be increased as such relevant marketing messages make it more likely that people will engage; the efficiency of marketing spend can be improved; and customer lifetime value will be increased.

Considering the context of your products and services must be matched to the context of your customers – their actions, transactions and engagement and combinations of how long they have been a customer, complaints, referrals or advocacy and such classic measures as Recency Frequency and Value score.

You must also match to their personal profile or persona. Rather than targeting just based on demographics, consider how people behave and what this tells you about what might engage them.

Demographics are largely static and may not always influence how or why people buy, but psychographic and behavioural personas can give insight into who does what, and why based on aspirations, attitudes, self-view, price sensitivity, journey stage, satisfaction and sentiment; create strategies that target each behaviour-based profile.

Try to introduce personalization in real time – driving dynamics in the customer experience. It has been suggested that not personalising in real-time is not personalising. We didn’t have the channels for communication or the technology for driving real time dynamics back in the 1980s, but such tools are available today.

Personalization should enable immediate reaction – like face to face, and so you need personalization technology that can understand, react to, and optimise customer journeys in real-time by applying data analytics to deliver the right message or experience to the right person at the right time – that adage that we worked to 30 years ago!

Dynamic content presented to customers can be achieved using machine learning that decides what the best content for each customer is, based on such parameters as purchase history, preferences, persona and browsing and buying behaviour, along with the customer lifecycle.

But don’t just consider what to personalise, but how to personalise it; use the data-driven personalization to drive the creative presentation in copy, images, format, offer and the response channel.

Make personalization an integral part of the experience but don’t go out of your way to push it in the face of your customers. Don’t do something just because you can, like the meaningless incorporation of the customers’ names and towns discussed earlier in this article.

Similarly, don’t flaunt to customers how they are tracked or the data you hold – I have seen this just unnerve them and so it has a detrimental effect on the experience. It has been suggested that your tactics should go unnoticed and create an effortless experience.

The best personalization is that which enhances the customer experience without them querying how or why, but also demonstrates an understanding of the customer and reflects the truth about them.

Good, reliable data is the fundamental ingredient

Poor data is the single most quoted reason for failure of such communication strategies. Ask yourself:

Do I have a data strategy?

Can I rely on my data?

Have I undertaken a gap analysis?

Am I maximising the touchpoints?

If the answer to any of these is ‘no’ then this is your start point.

So, in summary…

  1. Basic personalization tactics are no longer enough to engage members
  2. Using personalization intelligently is the best way to predict and shape behaviour
  3. Following the elements of data-driven personalization helps you develop your strategy
    • Embrace predictive analytics
    • Use data to drive dynamics in segmentation, offer, creative and channel
    • Establish measures that will feed the intelligence behind your personalization
    • Ensure your goal is based on improving customer experience and you’ll see increased engagement, retention, commercial success
  4. Ensure everyone has a single view of the truth through sound data management and governance

Know your members, grow your business

Whilst reduced discretionary spend continues to increase share of wallet and restrict personal subscriptions and cost cutting means less corporate money to pay for staff memberships there is an approach that can help maintain and even increase revenue levels.

It is essential to be in a position to address your own business imperatives like pre-empting churn by knowing which members are less likely to renew their subscriptions, determining propositions to help withstand cutbacks for your products and services or perhaps strategies for reversing the slowdown in recruitment of new members. Practically all will need direction on how to improve the effectiveness of marketing activity.

The solution lies in leveraging probably one of the most valuable assets of the organisation. By maximising the use of your database you will turn it into a powerful marketing tool. What you know about your members can help drive a better understanding of how they use and value their membership, leading to more relevant communications, better management of the relationship and increased loyalty.

Your organisation will have membership data. Some organisations will have membership relationship management (MRM) systems with all data in one place, but more likely is the scenario where some may be held in central systems whilst more is held in additional departmental data repositories like standalone databases or as spreadsheets. The object is to turn all that data into information and that information into knowledge or insight.

Using your database goes far beyond just direct marketing. The key lies in bringing together everything you know or could know about your members, prospective members, customers and other constituents, their activities, actions, purchases and behaviour. A single view can reveal who is active and who is dormant, who will only ever remain a customer for publications and training, who attends events and participates in branch affairs and when and to which campaign they have responded. It can also help determine who your next contingent of advocates may be.

Along with basic identifiers like name, address, telephone number, Email address and date of birth you are likely to hold qualifying data such as gender, occupation and professional qualifications. Ideally you will also have access to details of their behaviour, such as renewal history, event attendance, training and contact history.

This will give you access to the ‘who’ did ‘what’, ‘when’ and ‘how’ that go to deliver the profile of a constituent. There is no substitute for accurate data – analysis tools can’t compensate. Meaningful results can be achieved without sophisticated modelling but with high quality, robust and reliable data; it’s the quality that will determine how much of a grey guide or ‘black & white’ specific analysis can be reached.

It is essential that you assess your data before trying to develop any analytics. Inaccurate data will skew the results. If you cannot improve on the quality in the timeframe, then at least being aware of the inaccuracies will help in the interpretation of the outcome.

Building these profiles of your constituents will enable you to determine how long someone stays loyal and what might be done to pre-empt churn. It will facilitate the application of scoring and lifetime value techniques to help build predictive models and identify the communities within your membership base for targeted marketing.

This identification of communities can be viewed as segmentation“….the division into homogenous clusters which may be identified and marketed to with a specifically targeted communication strategy”. It is potentially more costly since it requires greater levels of management yet it promotes greater effectiveness through improved relationships.

Most people see the benefit of segmentation but the question is how should you segment them? By geography, by demographics, by professional qualifications or by behaviour? I would suggest it is a combination of all of them. The other dimensions to consider are concerned with how the constituent wishes to conduct their relationship with the organisation. I call it the ‘spectrum of engagement’ which runs from the apparently active member who says ‘send me every communication, proposition and offer that there is’ at one end to the member who appears distant or dormant at the other who says ‘I know where you are when I need something; don’t bother me’. The key concept to grasp here is that your most loyal member and potential advocate could exist at either extreme and, in fact, anywhere in between.

You need to know what is happening within your constituent base. Equipped with the kind of data mentioned above you can start revealing important measures. I normally start with trying to understand what touchpoints each constituent has had with the organisation. Taking each of the services or product areas your organisation offers count how many constituents have availed themselves of the combinations on offer. Table 1 shows an example of this.

Count of current members

Bookings (courses/ events)

Attended event / delegate

Made enquiry

Took exam

Book shop sales













































































































Table 1

This is the same concept that we have come to recognize when buying consumer products on-line; the helpful hints that tell us that people who have bought the book we are looking at have also bought these other titles, green socks and a set of gardening implements. It is known as ‘market basket analysis’ and is a totally valid model across any combination of goods or services.

Here we can see how many members have been active and which dormant. It also reveals some issues for further investigation like the 6,000+ members who took an exam without buying any course books or the 7,000+ who make bookings but never attend i.e. do they book on behalf of others?

The next most interesting insight is usually a view on churn (see Table 2). By matching the year of defection against the recruitment year trends can be identified. It appears, realistically, that churn increases the longer the membership period. However, what happened at the end of 2009 to make the number increase and carry this trend on into 2010 with 186 members not even completing their first year? This is where the penultimate data element comes in: tacit knowledge – the experiences and understanding that exists within the minds of the people who run the organisation. Using this to interpret the outcome of these analyses is the tipping point between information and true knowledge.

Years as a member

Set up Year



















































Number of defecting constituents

Table 2

Gaining insight into your members, customers and other contacts will help you understand who the best are, what opportunities exist, how to hold on to them and find more like them. Insight also provides the other side of the coin: who the worst are and what they are costing you in lost opportunity so that you avoid them and others like them in the future.

If you don’t have the insight into your members, how do you hope to manage the relationship?

What we have considered so far helps us understand the “Who”, “What”, “When” and “How” that were behind the members’ actions. The final data element addresses the ‘why’ factor – the additional dimension that establishes true differential between passive passenger and active member? This is the difference between reporting on the data and analysing the data, to reveal why a member remains loyal and leverages his affinity to the organisation in his working and social life.

To achieve this you must include a view of the attitudes and aspirations that drive their membership decisions. Known as adding the psychographic profile, such insight into members will:

  • Help identify and pre-empt their failure to renew membership;
  • Show opportunities for cross- selling other services and up-selling on what they buy;
  • Help you understand why specific propositions are successful with certain types of member or customer;
  • Reveal preferences, such as which types of contact will never become active members or participate at local branch level; and
  • Increase the effectiveness of prospecting.

You can achieve this through fusing the analysis of member and customer involvement derived across all of the points of contact with your organisation with targeted market research survey data to determine the true differentiators. The outcome can then be used to create communication strategies to ensure no opportunity is lost and the member is always directed towards the best next action, driving relevant, targeted communications.

Consider the scenario where information acquired through tracking response and behaviour is matched to the researched view of attitudes and aspirations to create member profiles. These can be used to determine how best to move the relationship forward or flag up potential danger signs. Does their most recent action (or inaction) indicate possible churn? Has their most recent behaviour been exceptionally different to all that that came before? Are they approaching a major milestone in their relationship with the organisation such as a threshold of membership level that requires examination or CPD audit and which has traditionally been a jump-off point for members like them?

Having insight into what motivates their behaviour can be used to generate a relevant communication or even direct that member to the right web page or telephone agent as part of a strategy to retain their membership or upgrade into a new level of membership.

Remember the ‘spectrum of engagement’. On the one hand, the active and the other the aloof or one could say there are committed members and those that participate on an adhoc basis. You need to know where each constituent is on the scale and it is the organisation’s role to determine the position and use that knowledge to drive the relationship, the regularity and content of communications and the propositions offered.

What you can determine about current members can also be applied to prospecting activities in a similar way, to help meet acquisition goals. By applying the profile knowledge to new applicants, the organisation can be better prepared for what kind of member the new applicant is likely to be.

Armed with this knowledge you can ensure your marketing strategy addresses these issues to drive the relationships.