The winning organisations over the next 5 years will be determined by which are best able to harness and exploit data to their advantage. The winners will be organisations that operate closed loop insight driven processes that are aligned to their strategic business goals. The losers will either observe this data revolution without meaningfully acting, believing they can carry on as normal, or make only tactical technology deployments that are not sufficiently aligned to what really matters to the business. They may not also understand the need for a confluence of business processes, data science and technology. For the winners, these will be fully integrated and not abstract disciplines.
Successful data and analytics strategies and deployments will be transformative in terms of driving the organisation to perform to the best that it can be and in doing so making it a competitive winner.
Getting it right and making the most of your data and analytics investments uses the sort of proven IP and methods we have developed at my company. That said, there are some basic steps you can take to ensure that you are focussing on the right objectives and your data and analytics projects move the needle forward for your organisation.
What questions do you really need to answer? You might be awash with volumes and types of data, but that in of itself is not going to tell you what you should be focussing on, no matter how hard you look at it. If magical patterns appear by merely looking at the data they are likely to be nothing more than just interesting correlations and highly unlikely to be able to drive any meaningful improvement in your organisational performance. Always start by looking at your organisational strategy, the processes and goals that are designed to enable you to deliver on that strategy. Are these processes best suited to delivering your corporate strategy effectively and efficiently? Are the goals measured in a way that is meaningful to delivering that strategy successfully? By looking at your data and analytics needs through the lens of the business strategy you will more easily identify the processes and goals that matter, need to optimised and measured. First answer the ‘what’ and only then the ‘how’.
Explore the available data in a structured way (even if the data is unstructured!) – When you have reasoned and prioritised what things really drive your business, it is then time to look at the data that supports or is generated by that process, measure or goal. Your intimate knowledge of your business will help you here, as there will doubtless be more data and data points than you can shake a stick at. Use them all? Definitely not, as some data will be nearly impossible to harvest and process in a timely manner and then hardly move the performance dial at all. We have developed a 16 point template model that we use with our customers to comprehensively identify the data that makes a difference and is cost effective to extract and process. Generally though, whatever data you are going to use, it needs to be:
- Sufficiently comprehensive
- In the right place
- At the right time
- In the right format
- Where possible federated
Presenting the outputs – having neatly skipped over the analytical processing of the data, the technology, frameworks and algorithms (as I’ll return to these in more detail another time) – we will eventually end up with outputs that need to be presented as ‘actionable insights’. The more important of these two words I believe is ‘actionable’ and how we ensure somebody does action them. If it is a closed loop automated ‘if this, do that’ system or machine then presentation is merely machine code. For us humans it is much different. As individuals, we all act differently and we respond to different prompts to act in different ways. Some people, like dashboards, others alerts, others still audible or natural language cues. ‘Different strokes for different folks’. Modern analytics software has great visualisation modules, but do take the time to understand what stimulates your key performers to act. You should even consider psychometric analysis to ensure your people act as they should to ‘insight and alert’ prompts.
Resources and Talent
Extracting, preparing and analysing the data can be very labour intensive for anything beyond basic self service Business Intelligence. It is also not any old labour, so don’t think this can be done entirely by your existing Sys Admins. Have you already hired or plan to hire Data Engineers, Data Analysts, Data Scientists etc? If you have looked already you will have learnt that these skills are hardly abundant or cheap. Consider using talent from third party specialist organisations (such as ourselves at OCF DATA), as this provides you with flexible access to a varied range of skills without it necessarily being a fixed cost or overhead to your business as hiring for the payroll would be.
Leadership and Organisation
I have witnessed through the many analytics projects I have been involved with, that success and the realisation of the benefits is more than anything else determined by the commitment of the organisation. This commitment needs to come from the C-Suite, who set the organisational strategy that the data and analytics deployments will underpin and enable the organisation to excel. That is not to say that ideas and proof of concepts for data and analytics projects can’t come from departments, but becoming a winning data enriched and driven organisation requires sponsorship and day to day commitment from the very top.