Researcher says proper data analysis aids enterprises

By Standard Reporter

All business managers know that growing a successful enterprise rests on proper management, guided by ability to position oneself in the face of an uncertain future.

Business research is key in closing this gap as we make informed decisions.

However, more critical is the ability to correctly analyse the current information.

From customer surveys that help in understanding customer demands and guiding product development, to opinion polls that are powerful enough to steer a country’s decision-making process, the value of research cannot be underestimated.

Balancing research accuracy and cost is the cornerstone of successful research.

Mr George Waititu, the group managing director, Synovate Kenya (formerly Steadman) understands well the importance of accuracy in any research.

"To start with, the study design should have a proper understanding of the survey objectives," explains Waititu whose company has conducted major surveys in the country and the region.

Mr Waititu said errors related to study design as well as those likely to occur in implementation of the study, have to be properly managed.

Synovate uses an automated data capture system which has superior capabilities in handling and processing complex data structures.

The same system is also used for statistical analysis and modelling for all survey data, said Waititu.

Best approach

This is only an example; companies doing in-house research should handle the data with the same care to ensure that results are error-free, says Norman Kamuya the Operations Manager at SPSS East Africa.

The aim is usually to make accurate conclusions and business decisions.

"Potential users of an analytical tool will need to develop a systemised and structured approach to data analysis," Kamuya said.

"Well organised, accessible data is the foundation of effective data mining," he said.

A financial institution for example, could easily have multiple data files for a single customer, each recording the customer’s information a slightly different way – in business lingo, known as a product-centric approach.

In research, a customer-centric approach is preferred, in which a customer has a single record.

"Ability to link data files intelligently is key in building data structures that can be easily analysed," says Kamuya who is also an analytics expert.