Using data to validate your business instinct and maximise successful decisions

Every organisation, in every sector, must use data. The value of using data to inform strategic decision-making and monitor day-to-day activities is irrefutable and undisputed. Whether your goals are to increase revenue, find efficiencies or enhance your brand – decision informed by data will mean you can do it better. Why then do so many organisations not do this?

We typically see organisations making two types of data mistakes:

1. technical – the underlying data foundations that must underpin a high-performing data culture are missing or low-quality
2. human – individuals and teams within the organisation use data poorly, whether this means pursuing vanity metrics, becoming slavish to dashboards without deriving insights, endlessly producing hypotheses without taking action to test them, or having data dictate rather than inform their activities

There are also legal and regulatory traps to avoid along the way, that are fallen into too easily and too often.

In this blog post I set out our checklist of good data practices to put in place, to ensure data becomes the rocket fuel your organisation needs rather than a drag.

1. Audit

Ask yourself: what have you currently got? A clear understanding of existing datasets (substance, quality, format), storage (location, access controls), flows (integrations, manual, API access), use cases (who, what, when, where, why?) and policies is a critical pre-requisite to everything else that will follow.

We usually find, when undergoing this initial audit, that no single individual has anything remotely approaching a holisitic and clear view of these matters within our clients’ organisations. This means we usually find something, somewhere, that was not previously properly understood and considered.

2. Objectives

Ask yourself: what are you trying to achieve? Data for the sake of more data is superficially attractive but not really a very good use of funds and effort. Tie your data objectives back to your overall business goals. This means that in some way your data objectives should contribute to growing your brand, increasing your revenue, or decreasing your costs. Setting out with clear intentions is really important, and bonus points are available for setting clear success criteria that ensure you hold yourself to account and can measure access or failure properly.

For example, a well-defined data objective might be: use data to…enhance our understanding of our customers/users. This example might tie to overarching business objectives, which might include improving customer experiences (brand and revenue), launch new products or services (revenue) or reduce waste (costs).

3. Infrastructure

Ask yourself: what do we need to put in place to best harness data? An end-to-end data architecture can be complex. It will include:

• collection and ingestion – where are you getting the data from? This will likely be a combination of first and third party data, and depending on the size of your organisation will likely include both API pulls and manual uploads
• storage – where are you keeping it? Increasingly we see medium to large sized businesses setting up data lakes and warehouses, as common repositories that can be utilised organisation-wide in a scalable and secure manner
• processing – what are you doing with it? What processes are being implemented to transform raw data into something usable and useful?
• analytics and visualisation – so what? Effective visualisation and exposure to often non-technical stakeholders across the business in a well communicated and accessible way is critical
• security – how are we keeping this secure? Well-designed access controls, clear and well communicated policies, third party protections and support, and well-prepared recovery and incident response plans are of vital importance

4. People and Processes

Ask yourself: what are we trying to achieve? Too often, individuals and teams either operate without enough data, or drown in an abundance of it. Be crystal clear as to what data you have, what its limitations are, and how it and the derivative insights can contribute to your overarching strategy.

If you are a senior stakeholder, role model references to and critical analysis of data. In practice, this means saying things like, “We know [X insight] because of [Y data]. My confidence in this is stronger/weaker because of [Z context]”, rather than “Okay, so [X insight]”. This helps to ensure that, in the 90% of the time that you are not in the room and your direct reports and junior colleagues are conducting these exercises without you, they will do so properly and holistically.

5. Governance and Compliance

Ask yourself: what are we allowed to do? Data is powerful but can be a massive liability for businesses, too. Data breaches can lead to costly business interruptions, embarrassing reputational damage and, in the case of personal data, potentially game-changing fines and sanctions.

The most effective internal policies and procedures are those that comply with legal and regulatory requirements in all relevant jurisdictions. They also, almost as importantly, are effectively communicated and well-understood at all levels of an organisation. Too often we see one or both critical requirements missing, to the detriment of the relevant organisations.

There is a lot to think about when thinking about data strategy and best practice. None of the steps are independently complex per se, but they do require a deep understanding of an unusually broad set of principles.

There’s support available if you’re not sure where to start, are stuck, or are not yet achieving the results you want. Get in touch with me direct at Sian.Rodway@mdrx.tech to discuss how to use data to make decisions and empower your teams.