Debbie Oates, Principal Consultant, Analytics, Experian Marketing Services
Developing a data strategy is nothing new – many organisations have been focussed on the topic for years. However history shows many organisations struggle to get the right level of engagement, leaving such initiatives stalling with business functions still questioning the accuracy of data underpinning day-to-day decisions.
The addition of the Big Data effect adds even more challenges:
- organisations that previously had limited information on consumers now have many opportunities to collect and leverage data, but to make the most of this data they may need to recruit different skill sets and embed different business processes;
- even those who were traditionally considered data rich are having to evaluate how new data sources can be integrated into their business in a cohesive and useful way.
One thing is for sure, there is now a general consensus that data is a most valuable asset that, when farmed in the right way, will deliver organisations increased opportunities and insights to provide a stronger competitive advantage. Bearing this in mind, how can organisations develop a strong data strategy that resonates through the business to achieve cross functional buy in? A pragmatic approach is required to avoid data paralysis, one that focuses on business benefits to prioritise critical activities versus nice to haves.
Here are the 6 essential steps for defining a Big Data strategy:
1. Start with the business vision and understand the supporting initiatives required to achieve this.
This is a fundamental step that many organisations appear to skip, delving right into the detail rather than taking a high level view of what they are looking to achieve with their data. These high level vision statements vary across organisations and sectors although reoccurring themes stand, such as ‘knowing our consumer’ or ‘developing increased customer advocacy’. Once this has been agreed focus on the success metrics. Each organisation will be likely to have at least 3 or 4 supporting initiatives; these are likely to include obvious objectives such as increasing subscriber sign up, active customer base or increase product purchase. Again, these basic steps ensure that the business has a clear view of what it’s trying to achieve before embarking on such a project, thus obtaining key stakeholder buy in. Further down the line, referring back to these can provide clarity as to whether a task is in or out of scope of the project.
Having agreed on the basics review what the business needs to do to achieve them; this is likely to touch on:
- Current data collection mechanisms and marketing opt ins to ensure maximum customer contactability;
- Developing increased customer knowledge across demographics and previous interactions with the organisation;
- Leveraging advanced analytics to increase relevant personalisation and contact strategies to positively impact customer behaviours;
- Ensuring all customer touch points are aligned to deliver relevant and consistent messaging;
- Increase ability to react to behaviours in real-time as the consumer is engaging with you;
- Track impact of changes over business as usual.Where possible even high level estimates of potential returns help grab attention.
Having developed the vision it becomes much easier to review exactly what data will be necessary to underpin the strategies.
As you are defining the information then it is useful to assess whether you have highlighted items that address every part of the customer journey – the data an organisation requires to achieve an effective sales conversion from prospect to buyer is very different than managing an ongoing relationship with a loyal consumer.
Concentrate on a wish list regardless of whether the data is currently available – remember to include data that could come from external sources as well as your own internal sources.
Linking back to the overall vision should allow the business to assess which information is priority – data items that cannot impact any of the overall objectives should be disregarded at this stage.
This is important as now more than ever there is an abundance of data.
Once decided on which elements of data are fundamental, organisations need to review their current data assets. This should cover the following areas:
- Data coverage: where are the gaps in data items coverage – can these be plugged by enhanced data collection or utilising external data?
- Marketing options: assessment of current processes to assess if they are leading to optimal contacts and whether there is a case to try and enhance current opt-ins across different channels.
- Data quality: check how clean the current data is, assessing volumes of duplicate records across individuals and whether data items captured across different source systems have been captured to consistent documented rules.
- Data redundancy and governance: if data is collected at multiple points, review which data should be taken from what source as data is integrated.
- Meta data capture: if there is intelligence embedded into data items, such as source codes or campaign extracts, is the underlying data captured in an easy to access way.
- Data Linkage: different sources of data may operate on different references, e.g. email subscriber lists compared to website transactors. Organisations need to assess how best to link this data together to meet their objectives, again this could be enhanced by considering external data assets.
- Data capture: current processes and usage across different business units.
- Documentation: reviewing the current data assets will likely uncover a lack of data dictionary documentation. As part of the process rigorous documentation should be adopted to ensure that time is not wasted.
Define and prioritise what processes, resources/skill sets and technologies need to be adopted within the business to firstly get the data into the required state and then to analyse, deploy and measure ongoing impacts. You will probably need to adopt a staged approach based on an informed view of what will drive most business benefit. For example, is there any point getting distracted on how best to integrate social data if you core customer data or transactional data can’t be trusted?
Data should be seen as an asset so any data strategy initiative needs high level sponsorship and ongoing commitment to ensure that decisions are underpinned by accurate data.
It is not a stand-alone project but should develop over time as the business objectives change or new data becomes available – thorough ground work at the start of the project will allow an easy way to evaluate new requirements.
Getting a second opinion:
Planning data strategy projects can be complex and it is very easy to become inward looking. To move through the data strategy process more rapidly it is often useful to get an external view to ensure that wider business considerations are brought to the fore.
Areas to consider input are:
- support on scoping out the initial vision and input into the roadmap development;
- analytics to underpin any business case required;
- data cleansing and enhancement – including providing linkage across channels;
- assessing Single Customer View, marketing data mart build and execution technology to support the relevant access and insight capabilities.
Regardless of whether you canvas opinion wider than your own organisation, adopting these 6 best practice principles should help add structure to any project and avoid data overload.