Build a training approach to suit your business needs, your calendar and your skill gaps
Making sense of digital data
18 May 2012
The vast amounts of digital data organisations have access to can act as a barrier to real insight for marketers and data experts alike. There’s a temptation to just report on the easiest pieces of data and monitor those features that grab our immediate attention.
However, it is the consumer underneath all that data that is making the decision to engage with and choose your brand, product or service. So, to truly make sense of digital datawe need to put the customer at the centre of all our activities whether we are analysingsocial media data or a customer database.
Digital data - advantages and challenges
Digital data provides bucket loads of immediate data. We can start to wax lyrical about “big data”, knowing everything about our clients and customers, anticipating their every need. Getting messages, to the right person, at the right time in the right manner.
Gone are the frustrating delays between execution and report, gone is the difficulty of linking outcome to communication, gone is the need to make educated guesses.
However, it soon becomes clear that these promises rarely match up to reality. Response rates stay stubbornly low on the whole, apart from the few exceptions that generate lots of publicity. Customer satisfaction levels fall, as expectations rise about what is possible.
We realise now that out of this huge volume of data that’s available, we have to begin to make judgements about what is useful and predictive and what we should ignore.
Focus on the untypical prospects and consumers
The most useful way to forge a way through all this data is to keep looking to create summaries of the behaviour of individual consumers and/or businesses. This may consist of the data within an existing customer database, but will also identify when these customers have become a Facebook friend, signed up for email communications or made a request for information from a recent online landing page.
There is always the danger that our focus is distracted towards separate easily countable activities, for example web page visits, cookie histories, email responses. These become the basis of our reports, selections and decisions. Take some time to collate the various activities, touch points both positive and negative that your consumers and prospects participate in.
Benefits of an RFV/RFM approach
My own experience keeps on returning to a very old method of sampling and investigation. In whatever format I am working, my first approach is to look for the last time a consumer made a contact of whatever type, the number of contacts made by that consumer, and what impact and financial contribution that particular consumer has had on that organisation.
To those who have been around a while, they will recognise this as a slight refinement of recency, frequency and value /money (RFV/RFM) analysis.
Focusing on RFV/RFM has two benefits:
1) Asking the question forces data to be merged together at least during an analytical phase. As a result, it can highlight some curious inconsistencies within a data-driven organisation’s infrastructure.
2) It can demonstrate that consumers often do not behave as we anticipate, but rather engage with organisations in different sequences and for different reasons.
For example, online activity could generate an additional visit to a store, but subsequently generate an online purchase via a consolidated site. Similarly, when we see a low response to a communication it doesn’t necessarily mean the content or marketing device was flawed, but that we’re looking for the response in the wrong place.
Tracking consumer behaviour has always been a challenge. Today’s consumer has many different opportunities to surprise us by taking a variety of routes. Overwhelmed with all this data, we instinctively favour information that confirms our belief, leaving ourselves blissfully ignorant of other options.
It’s an example of a common psychological behaviour called confirmation bias. However, if we focus on the consumer as a whole, examining any response we can find, we’ll be able to cast the net wider and explore other impacts beyond the preferred route.
Joined-up approach to digital data
How can we assess the value of each separate click, call, visit or “Like”? We can achieve this by focusing on individual behaviour across various touch points, rather than each activity in isolation. Organisations can begin to see the benefit of managing the different digital-based activities alongside the offline ones that still take place, taking account of the position of the activity in the buying cycle.
This is often referred to as “funnel” management. The “funnel” here describes how you start with a large number of contacts showing casual interest and, as desire and commitment increase, the number of consumers reduces. To provide an objective way of comparing insights and effects across a wide range of activities, it is advisable to manage your reports, and decision around the inferred counting of individuals.
In the above example this means calibrating website visits to the likely number of separate individuals. After all, an individual may have multiple access points, (eg smartphone, tablet, home and work PC) to research and compare at a each level of the “funnel”. At each stage of the funnel, it is useful to identify the likelihood that the individual will proceed with a subsequent purchase. As a result, the benefit of website engagement could be compared and contrasted with the incremental impact of a Facebook consumer indicating a “Like”.
Each one of these touch points indicates a different level of engagement with the brand. A comprehensive approach to this might, for example, demonstrate that 1,000 Facebook Likes are as effective as a single inbound phone call. The hard work required to deliver these insights provides an objective way of comparing across different channels, media and stages within the consumer sales cycle.
The following principles become clear:
1. Consumers are the source of decisions within the sales cycle, so assume they have the initiative and analyse the results on that basis.
2. The same consumer will appear in a number of different guises at different stages in their own buying cycle.
3. Lots of consumer online activity does not result in a sale. Increasingly, consumers are researching and comparing brands and reputations online.
4. Identify where consumers are engaging with your own unique brand characteristics rather than riding on the growth or otherwise of your particular category or sector.
A final thought...
If in doubt, ask yourself why you’ve decided to focus on this data set over another. Remember, “not everything that is counted, counts and not everything that counts can be counted” (Einstein). Equally, “if you don’t measure what you value, you will end up valuing what you measure” (Anon).
Tim Drye, director, DataTalk