Step 4: Operationalise
Updated: May 30
Determine the data, tech and resources required to support the insight
Data can be a powerful strategic tool but it isn’t magic - it requires a strategy based on your business ambitions to determine what you need to deliver value for the business.
Having worked with many SME’s over the years, we have developed ORION, a simple 5 steps to help you develop your data strategy.
By following the ORION 5 steps you build a plan and organise what needs to be done to support your business ambitions using data and insight, as well as identifying the additional support and skills that may be required.
To get the most out of this step, Operationalise, we would recommend reviewing the first 3 steps so you are in a good position to have identified the strategies that will support your business ambition and have got a clear view of the insight you need to be focussing on.
If you haven’t read these steps here is a brief summary:
Step 1: Opportunity determines the business needs and ambition, the importance of creating a common agenda for your data strategy and the role of a cross-functional team.
Step 2: Realise focuses on using data to plan the strategies you are going to put in place to deliver against the business ambition, establishing the key questions you need to be answered to support your plan.
Step 3: Insight identifies what insight is required to support the strategies, working closely with the data analytics and scientist teams to ensure that the business context is captured in the requests.
To learn more about these three steps, visit the blogs on each area on our website.
The Operationalise step is about constructing an operational plan that will identify the tasks required to deliver against the insight that has been identified in Step 3: Insight.
Even though ad hoc analysis will always be required, it is beneficial to have an operational plan to improve access and quality of the data that is being produced in the business in a systematic way. This will address any gaps there are in the data landscape and to accelerate the use of data to support the business ambition.
These tasks will require a combination of analysis, model building (machine learning), data assets, technology (processes or additional infrastructure) and resources required (data visualisation, analyst, scientist, engineer, architect etc). Some of which will be easily achieved but others may require a longer-term plan.
The operational plan will enable you to deliver insight today whilst building the infrastructure for more transformational insights.
The main aim is to establish the effort required to deliver the task whether additional any additional costs or investment is required. This will be used in Step 5: Now! to help prioritise the tasks.
Knowing your data landscape and ownership
Before building the plan it is helpful to understand what current state of your data is across your organisation.
It is common for data analysts or data science teams to spend a large proportion of their time manipulating data in order to prepare it for model building and analysis, before actually getting started with the analysis.
Some of the main reasons for this being the case are due to :
Limited understanding of what data is available across in the business
Poor data quality, duplicate records and missing data
Poor integrations between systems requiring a manual matching exercise and process
Lack of data; requiring data enhancement or a data capture strategy
Lack of data governance to determine the definitions and accuracy of the data
In our experience, these challenges arise due to the lack of data ownership within the business, where it’s no-one's job to manage the data value within the business.
IT will often support the infrastructure and the flow of data throughout the business but are less concerned with the quality and business use of the data.
This can have an impact on the quality of the output of the analysis, consistency of the insight and significantly increase the time required to produce the results.
So, we would recommend always having clear ownership, roles required and critical path dependencies agreed for each task.
Being realistic about the timings of tasks
Some of the tasks within the plan may be relatively easy to implement, for example, a data audit to provide a better understanding of the data landscape could take 2-3 months, but, developing a master data layer requiring integration between multiple systems could take over a year.
Data always is messier than anticipated, ensure you have the right project team to accurately assess how long something may take.
Things to consider when building the plan:
1. What analysis and models are needed to be built
What is the output that is required, ie dashboard, infographic, presentation or rules?
Does the insight use cases require a one-off piece of work or an on-going requirement?
What kind of analysis is needed?
2. What data assets are needed to build the models
Have you got a data dictionary across all the platforms/databases?
Are there consistent data definitions?
What data, reporting, insight and processes already exist?
What technology/platforms are you using?
How are technology/platforms integrated and are they fit for purpose?
Do you need to source new technology to support the tasks?
3. What resources are required across the business
What resource is currently available?
Where are there gaps?
The role of in-house vs. external experts to get processes underway and train existing individuals or teams
Are there other stakeholders that you need to get involved?
4. How is the project going to be managed?
Who is going to own the project?
What data governance is in place?
What are the dependencies between projects
5. What investment is required?
What additional investment is required, tech, licenses, resources etc
Predominately based on the requirements from the previous task
6. How long will it take?
Is a short term or long term project?
Is there different walk, crawl, run phases for the project?
Capturing these components of the project enables you to move forward with the execution and implementation of the tasks.
Our Operational plan template here can get you started.
Congratulations, you now have completed the fourth step of building your data strategy.
You will now have a better view of the tasks that are needed to be done across the business to deliver against the insight you have identified.
The final Step 5: Now! Will focus on prioritising what to do first based on the effort and cost required versus the gains expected that have been identified in Step 2: Realise.
We Are MoJo is dedicated to developing data strategies for SME’s to drive business growth. We can facilitate this approach by providing regular access to data strategy leadership.
Get in touch to find out more - Tash@wearemojo.co.uk