Building a Data Analytics capability

Developing a robust data analytics capability within a business, requires a number of important steps that need to be undertaken before embarking on any development programme. There are a couple of key points you must consider when starting this journey.

1. Start with the end in mind

The first and most important step is to understand clearly what the end goal is, engaging the business to understand the strategic imperative.  Many businesses spend too little time on clearly defining what they want to achieve from an analytics capability, agreeing a clear scope and developing a realistic road map. Without this, any project is likely to become unmanageable and result in providing more confusion than insight to the business.

2. Small data - Big data

A further mistake many make when developing their capability is to try and bite off more than they can chew. Often firms are sucked in by the dream of plugging into the ‘Big Data’ revolution and then discover that they are unable to cope with the scale or complexity of the data that’s produced or imported, or what to do with it! Ideally a business should start off by understanding the data sources and information they already have available in the business and the quality of that data. Starting with ‘small data’ first is the most effective route – by starting small, the process is more manageable and can provide an early insight into the quality of the data that already sits within the business. Should there be concerns in this area, there are a range of tools that can be used to both monitor and develop the quality of data and provide an external benchmark if needed. If the data entering the system is poor, then any analysis conducted on it will also suffer a similar fate. Big Data techniques for working with ‘dirty data’ are effective and well understood but when working with relatively small numbers of records (loans or other financial instruments rather than web searches or website hits) then these techniques are less robust due to the significantly smaller sample sizes available once the data set is split into meaningful subgroups. 100,000 records is a good size for a loan arrears subset but a tiny number for Big Data.

3. Strong foundations

Before a business can properly build a data analytics capability, there are two key areas that need to be considered:

1.      Technology

2.      People

Clearly a robust technological solution is required in which to order to collate and manage data within the business. Choosing the right system and tools is key and they don’t need to be the most expensive or bespoke – scalability and fit for purpose are the key drivers here: it’s vital to ensure that a system can both cope with the data that is currently produced, but can also grow with the ever increasing amounts a business will produce or import in the future.

Having the correct knowledge and skillsets in place is also vital. A balance is required here between data/IT skills, analytical/statistical skills and strong commercial/business skills. If the pendulum moves too much in any one direction, then the analyses can become an academic exercise, lack true business insight or the core statistical skills can be lacking to work with the data correctly, producing spurious results. Ideally a good mix is required, and this can be rare to find in the same individual or any one team.

4. A data driven organisation

A solid data platform and analytics driven from it should become the engine of the business, not just supplying routine aggregated data, but providing true commercial insight in order to drive effective decision making. A key question for a business is often where the analytics capability should sit – it can be shoehorned into finance or IT functions based on the data sources or technology requirements, but ideally it should sit as a ‘shared service’ function. It then can pull data in from sources from around and outside the business and provide insight to underpin decision making.

About Target

Our experience with business process engineering, including data positioning to facilitate management insight into core performance drivers, is augmented with experience in re-engineering our own internal MI platforms to facilitate state of the art data mining techniques for the benefit of our customers.

We can support your business requirements with our data analytics services.