Data is gold in our increasingly digitised world, just as the value of gold is only realised in the refinement process. Data needs to be refined to unlock its real value.

Unrefined data can damage businesses, their competitiveness and ability to capitalise on opportunities. Good quality data, which is refined, can be leveraged to improve competitiveness, decision-making and profitability.

The pace at which data is being collected and stored is unprecedented, and this will only continue to accelerate.

Modern organisations expect Data to drive innovation, progress and competitiveness, however Data is only as good as its quality.

Poor-quality data can severely damage a business’s ability to make good, informed decisions. This has a direct bearing on performance, resulting in lost revenue and missed opportunities, possible reputational damage and increased operational costs trying to deal with data errors.

Beyond this, poor data quality may well lead to misguided strategic investment decisions. It is abundantly clear that businesses must prioritise high-quality data.

So, how do businesses end up with poor-quality data? Human error, outdated systems, inconsistent data-entry protocols and a lack of data governance lead to duplication, inaccuracies, inconsistencies and conflicting data sets. Without proper data governance there is no standardised process for maintaining high-quality data.

Maintaining good, clean data requires implementing essential key performance indicators (KPIs). These are: relevance, integrity, completeness, uniqueness, timeliness, validity, accuracy, consistency, accessibility and reliability.

A good data partner will assist an organisation with tracking these KPIs on an ongoing basis to maintain high-quality data.

Relevance is crucial as it ensures that data aligns with the context in which it is being used. Irrelevant data can clutter the analysis process and hinder effective decision making.

It is advisable for companies to consistently assess their data collection standards and clearly define their data needs. Furthermore, eliminating unnecessary data is equally important.

Integrity plays a vital role in fostering trust and compliance, encompassing practices such as data encryption, access control measures and regular integrity audits to detect any breaches.

Completeness ensures that all necessary data elements are present, which is essential for analysis and informed decision making.

This involves mandatory fields in data entry systems, conducting audits to identify any gaps and automating the process of collecting relevant information.

Uniqueness evaluates whether there are any duplications within the dataset, which can impede analysis and lead to inefficiencies.

Organisations can mitigate this risk by leveraging de-duplication tools, establishing protocols for data-entry procedures and conducting audits to identify and eliminate duplicates.

Timeliness reflects how up to date the data is. Outdated data may result in missed opportunities and flawed decision making.

Validity ensures that all collected data adheres to specified parameters and formats. Invalid information can introduce errors and distort interpretations. Implementing checks and utilising machine learning can enhance the accuracy of entering data.

Accuracy pertains to how the collected data mirrors reality. Implementing cross-checking mechanisms, using authoritative data sources, and regularly verifying data against external benchmarks are crucial for maintaining data accuracy.

Consistency speaks to the uniformity and reliability of data, across datasets and systems. Discrepancies can lead to confusion and undermine confidence in the data.

Developing data governance frameworks harmonising data across systems and utilising master data management (MDM) solutions can enhance data consistency.

Accessibility relates to how readily available and easily accessible data is to authorised users. Inaccessible data may cause delays in decision-making processes and impede operations. Implementing user protocols for accessing data is essential for enhancing data accessibility.

Reliability ensures that the accuracy of data remains consistent over time. Performing assessments of data quality, adopting maintenance practices for managing data and promoting a culture of responsible data stewardship are essential for upholding the reliability of the data.

To address dirty data and build trust, organisations should:

• Implement Data Cleaning processes – Regularly clean the datasets by eliminating errors, duplicates and outdated information using tools designed for this purpose.

• Standardise data entry – Set guidelines for entering new data to maintain uniformity within the database. Make sure to train your staff on these guidelines and implement data validation rules to enforce them.

• Enhance data governance – Establish a comprehensive framework for data governance that includes standards for data quality, policies and procedures. Designate data stewards to drive data quality and ensure compliance with governance protocols.

• Leverage technology – Make use of data management technologies such as master data management (MDM) and data integration tools to maintain consistent and accurate data across different systems.

• Promote data literacy – Educate employees on the significance of maintaining high quality data. Foster a culture where everyone takes responsibility for ensuring data quality.

The pursuit of high-quality data is an ongoing process that requires a strategic approach and commitment from all stakeholders.

Organisations can build a robust data quality framework by focusing on data quality KPIs, while implementing best practices such as data governance, automation, training, regular audits, data integration and a culture of continuous improvement, will help them significantly improve the quality of their data.

The African continent is poised for economic growth, rapid urbanisation and increasing incomes with its population expected to reach about 1.7-billion people by 2030, making it crucial for businesses to fast-track data strategies if they are to remain competitive and benefit from these developments.

A major challenge in achieving these goals is that data and its associated processes have historically been seen as an IT function; while the IT department might have the knowledge and technical expertise to execute data strategies, many fail to deliver a return on investment as the initial process of defining the ‘what’ of the data strategy is not business-led.

Often, data strategies that are led by IT departments focus on the ‘how’ – the technologies required for big data, data cataloguing, data governance, data integration, data lakes and metadata management – without an initial and ongoing investment into understanding ‘what’ data is needed for the strategy to be a success.

Ultimately, there is a mismatch between the execution of data management and what the business actually wants, resulting in various departments disregarding the organisation’s technical data strategy in order to satisfy their data needs. Counterintuitively, this leads to the creation of data silos, a lack of data integrity, multiple versions of the truth and eventually a data strategy that fails.

Instead, what is needed is for the data strategy to align with business strategy, and for it to be able to evolve in line with changing internal and external environmental factors. This is usually only possible when an organisation’s data and information strategies are led by senior members who understand the business and where it wants to be in the future, and are ardent proponents of using data to continually inform business decisions.

By definition, a data strategy has to have impact across all levels of the organisation; yet, we see instances where data is available, but is not used to its full potential because the business people don’t know how to use self-service business intelligence (BI) tools to guide decision making. This is a common pitfall of having only the IT department lead data strategies, but it also points to the requirement for data literacy across all levels of the business.

Does this mean that all senior business executives in an organisation should be data scientists? No. On the contrary, all that is needed is for them to be able to use the self-service BI tools and formulate their questions about the business in a way that allows the actual data scientists to manipulate data and get the answers.

One way to get this right is by having a data team that brings together business analysts (who are responsible for various business functions), supported by data scientists and developers who can assist with complex transformations and creating predictive algorithms.

Not only does this ensure consistency in an organisation’s data strategy, but it also means that the business’s unique data requirements are met while IT best-practice is also adhered to.

The continent is blessed with a natural abundance of talent and potential, and having data strategies that are driven by a combination of business and IT – with the weighting of efforts allocated differently at various stages of the process – will be key to unleashing this potential. Meanwhile, this combination will also enable organisations to perform data projects that are directly linked to business goals and empower business people to make data-driven decisions on a daily basis – helping drive the competitiveness of businesses in Africa.

It should be noted, however, that there is no one-size-fits-all approach that can be applied in different countries around Africa, especially given that some challenges are unique to the continent.

There are also issues such as data sovereignty that need to be carefully considered; it is best that organisations work closely with a partner that has on-the-ground experience as well as a deep understanding of the conditions in different regions across the continent.