What are your organisation’s goals this year? Increase revenue? Reduce operating costs? Improve brand recognition? Increase NPS? I’d be preaching to the choir if I said that data is critical to achieving these goals, not least just tracking the metrics regardless of whether you’re making data-driven decisions.

A Data Strategy will help you understand and communicate exactly how your data assets need to be managed in order to achieve your organisation’s goals. But it can be daunting to know where to start.

Getting started

A good start is clearly separating out offensive and defensive elements of your strategy as described in this article in the HBR. To get going, it can be useful to think of defensive considerations as risks and offensive as opportunities but it’s not always that clear-cut in reality. For example, a data breach has clear negative financial implications (regulator fines, downtime etc…) but could also very negatively affect your brand recognition ambitions which you may have been pursuing as part of your offensive strategy.

Capture risks and opportunities

Your organisation is unique. We can’t apply a cookie-cutter approach. However, there are risks that every organisation needs to consider. A stock-take of these risks, and any specifically facing your organisation, is a good place to start. Here are a few examples.

• Intentional data-breach (financial, reputation, operational)

• Accidental data loss (financial, reputation)

• Discriminatory automated processing (legal, reputation, financial)

You face some of these risks as soon as you create any records related to your business activity - as in the first two examples. Some risks only get “unleashed” when you pursue loftier offensive ambitions such as when you apply AI/ML to optimise your support process for example.

What does good look like?

Assume we want to proceed with mitigating all of these risks. Next, boil these risks down to the processes that could prevent them. Then, you will have a series of imperatives that you need to enforce in order to mitigate the risks your data use poses to the organisation.

• All organisational data is stored in an accessible format with appropriate security and privacy controls

• The lineage of all data assets is documented

• The lineage of all data assets is available on demand

You can then apply a similar process to any offensive considerations and align them to a goal. For example, take increased NPS. In order to measure and make data-driven decisions to improve this metric you might produce something like:

• Continuously ingest data from all the customer feedback services we use

• Make that data available to decision-makers

Through workshops and interactive discussions you might boil these down to the following imperatives:

• All customer feedback data must be ingested into the data warehouse on an ongoing basis

• Data from separate vendors must be unified and modelled according to the agreed domain model that makes sense for our needs

• Decision-makers must be able to define the metrics and self-serve on-demand

Next steps

There is a lot of detail that will follow on from this which will describe how you are going to pursue these objectives - aka your Data Strategy. You should seek wider input ASAP as you will quite likely need to get some hiring and procurement balls to get rolling. Fortunately, this approach creates objectives that are fully-aligned with organisation-level goals to maximise your chances of successful funding for the initiatives.

Go forth and drum-up support for fleshing this out into a cohesive Data Strategy! Hint, an audit of your organisation’s current capabilities is a good start. You never know - things might already be in a better state than you thought and some of the required resources might already exist!