By Andrew Jackson, General Manager – Asia Pacific, Concep
The data analytics revolution is here. Used meaningfully, it can help transform your business in how it operates, organises, creates value and manages resource.
To help make this happen, it is senior level managers that have to grasp the nettle and get their hands dirty. A key reason is that the sheer complexity of the field means senior managers often leave it to their IT department.
This article looks at the critical elements that are essential to the success of making data analytics relevant for your business. People tend to think about data for data self – that is, just seeing interesting data sets, patterns and trends. However, this is just a means to an end. Data analytics is about providing the underlying answers. This will mean different things for different professional organisations – it’s about identifying the data points that mean the most to your business.
Ask the Right Questions
Most businesses just ask the more obvious questions: how can we improve productivity? Or how can we reduce cost? It’s better to delve deeper and get answers to the underlying things that need addressing: productivity per worker or cost reduction for a particular process. It’s about aligning important functions and domains with your most important business cases. And it’s about using these particular examples to gain meaningful insight. Vague questions lead to vague answers.
Start Small, apply Big
Start small. It’s easy to get seduced by the big picture. Companies can use data analytics to identify small points of difference with which they can then take advantage. The impact of big data isn’t some big bang, but a series of small incremental steps that, collectively, help improve the whole organisation. It’s what cycling and F1 teams refer to as ‘marginal gains. The benefits of these small changes can multiple exponentially when combined across bigger, multiple processes.
Think Outside the Box
It’s important to ascertain data from all its myriad of sources. Too often, data teams think only about their ‘pure’ data and ignore inputs from sources they think of as poor, outdated or inconsistent. Much of this ‘soft’ data needs to be taken seriously and added into the bigger picture of data analytics – email engagement is one example . Indeed, soft data can provide a pivotal role in areas such as sales forecasting or establishing relationship capital. To build a strong data model, companies should identify the source of every input and score its reliability. This form of risk management can then help management make critical decisions – even if the data is less than certain.
Seeing the Bigger Picture
Companies often fail to see the bigger picture. They are too busy drilling down into a single data set in isolation - but fail to see what all the data sets say when they are combined. This kind of error often happens when a company is trying to solve a particular technical problem - such as a reliability or quality issue. Whilst they might analyse the performance of a particular facet of the business, the actual solution might be in another area entirely. This can only be identified and solved by looking across the business as a whole.
A key insight might only be gained when data sets are put together side by side. An overlap or common issue then becomes clear. These insights become even more powerful when the number of data sets analysed grows significantly. However, there is the danger that this quest creates even greater complexity that can actually inhibit how data analytics is utilised. To tackle this problem, management should undertake a holistic approach to analysing their data. Effective interpretation means data cannot be run in silos. It means the outputs won’t work in real world conditions or, worse, the conclusions might sit unutilised.
Planning and Purpose
For data analytics to be successful there has to be a plan of action. Observe, Orient, Decide, Act – the OODA loop – is one such plan of action developed by a military strategist that can be applied to data analysis. You must Observe external data sources; Orient thinking based on the data and previous experience; Decide what direction or idea to adopt; then Act on it. This plan is used by many leading organisations to test assumptions, process new information more accurately and help them react more quickly. This means tracking and monitoring data to identify key patterns and moving quickly to take action if data points suggest a process is ‘off track’.
Make Data Articulate
Data must be articulate and self-evident to those outside the technical department. There’s no point in spending considerable time and effort developing a great algorithm that gives great results but is, essentially, long-winded and badly presented. This means that the end product won’t be utilised if people don’t understand it. It means the interface between the data and the user must be intuitive and easy to use – otherwise it is doomed to failure.
The Right Team Ingredients
Building a multi-skilled analytics team is like preparing a satisfying meal – you need the just the right balance of staff, skillset and passion to achieve brilliant results. This team includes: data scientists, IT engineers, data architects, user interface developers and designers. There also needs to be ‘translators’ – often senior managers - who bridge the technology and business gap to make its application real world. However, these specialist skills are often limited in availability, so look within your own organisation to recruit and retrain people with the necessary skills.
Make sure that the data analytics discovers insights that are practical for your business. Already analytics is moving out of its specialist field and is being used by data savvy staff as part of their everyday routine work activities.
Adoption is the Best Policy
There’s no point having great insight from your data analytics if the results aren’t embedded within the culture of your organisation. This means adopting them for real world processes and daily work flows. Everyone must sing from the same hymn sheet – from the most senior managers to front-line personnel who deal with customers on a daily basis. Only by having commitment from senior level influencers, will data insight be translated into competitive advantage.
By starting small, applying big; asking probing questions; seeing the vision across the enterprise; acting with planning and purpose; and pulling together a talented team, you can make data analytics truly work for your business.
*Inspired by an in-depth McKinsey whitepaper: Making Data Work for You