Issue #30 – Standardising KPIs to Enable Data Effectiveness
Why you need to ensure your data activities are actually delivering value! And that starts with the KPIs!
Read time: 10 minutes
In a Data Ecosystem first, this article was primarily written by Mark De Jong, the co-founder of POINT05, a data & AI consulting firm in Amsterdam. Mark has a ton of experience in everything data, from data strategy to building data platforms, dashboarding and data products. He wanted to do an article on KPI standardisation and methods to achieve that because this is one of the biggest problems he sees data teams facing. Check out his stuff and enjoy the article!
Raise your hand if you have standardised KPIs across your organisation?
Nobody? Yup, not surprised…
Many companies have a hard time creating a standard and unified framework for their key performance indicators (KPIs).
Different source systems, countless data formats, siloed departments, multiple methodologies to calculate the same KPIs. All these things plague the ability of organisations to measure what they do.
Especially in data!
Furthermore, data teams are scattered due to organisational structures limiting the process of standardisation and data governance or quality procedures meant to circumvent these problems are usually immature or non-existent.
And there is limited drive to fix this. Pressure from investors or the C-suite for quick actionable insights lead to time spent on short term urgent topics by the data team instead of focusing on a more long term standardised KPI framework.
So what will we do if we can’t standardise what we measure? Doesn’t this negate all the work we do to be data-driven? Well, hopefully this article can help!
Consequences of Misaligned KPIs
Usually, when you work in an office, everybody speaks one language relatively fluently.
Why? Because you need to understand each other to get stuff done.
And yet, companies tend to forget this when it comes to their metrics. These misaligned KPIs can hurt them in many different ways:
Collaboration & communication – Different interpretations of the same KPI frustrate meetings and limits collaboration between departments. It enforces the influence of silos in an organisation.
Decision-making – Unstandardised KPIs can hinder strategic decision making. Inconsistent metrics lead to poor, misinformed, or contradictory organisational decisions, slowing overall progress.
Resource efficiency – Resources such as budget, personnel, and time may be measured and acted upon to support individual department goals rather than organisational priorities, leading to wastage or inefficiencies.
Reduced financial performance – At one of POINT05’s clients, there were three different versions of the actual spend of a department for a multinational organisation. Depending on the dashboard (management overview, department view or finance view) the numbers would be different. How can this organisation optimise and improve financial performance without a consistent view of true spending?
Company performance deteriorates when there is noise around KPIs and their definitions. Standardising and structuring KPIs is an exercise every organisation needs to conduct at some point because these issues will eventually rear their head. And this means every organisation—from the small to large global ones.
Addressing the KPI Conundrum
So what can we do about it? How can we get Jane in Marketing and Fred in Finance to agree to KPIs so the data team can finally understand what is happening and report it to the Executive team?
Luckily, we have a four-step process for you:
1) Assess the As-Is state
The first action you must take is to identify the currently used KPIs, their definitions, source systems and methods of calculating them. A few pointers here:
Interviewing employees or managers on definitions and ways of calculation is not enough. Deep dives are required to understand which field comes from which source system and how the transformation and calculation methodology works
Identify the frequency of usage and the responsibility for the KPI. This means who owns the measurement methodology, the collection, and eventual targets for the KPI. Those people have to be bought in and champion the new definition!
Finally, determine the importance of the KPIs. Is each KPI really a key performance indicator? Is it essential to the organisation's success? Less is more here
2) Create the To-Be state
Based on your initial As-Is investigation and exploratory interviews, a proposal for new KPI definitions and ownership can be crafted:
The newly defined KPIs need to be centrally aligned and approved at the top level, then implemented top down to secure the standardisation this project aims to achieve
The creation of a comprehensive KPI dictionary or manual that includes all standardised KPIs, their definitions, calculation methods, and relevant guidelines needs to be created. Trust me, this documentation will actually be used and referenced in your organisation!
3) Plan the change
If you fail to plan, you plan to fail. Communication is crucial when rolling out new methods of measurement! So make sure you have a clear plan and aligned stakeholders:
Understand any technology / system changes that new metrics will impact. Plan for potential technology changes (say if a source system is replaced) to avoid redoing this type of exercise again!
Create a communication plan that caters to all stakeholder groups. Individual priorities are often unique even if KPIs are the same, so management needs a different approach to convince factory workers vs the marketing team vs finance vs other business stakeholders
Start with the most important Executive-level KPIs (usually financially focused) as the departmental targets/ metrics should ultimately influence these
Take into account key external stakeholders; changing KPIs could mean changing financial ratios which banks, external auditors or investors need to be involved in
4) Implement the change
Here we say two words teams hate—change management! To make this work, you have to get your hands dirty and put the plan into action:
The relevant KPIs and their definitions should be added and tracked in a simple Executive dashboard. This allows users to directly check the definition or data sources, helping them trust the metrics
Speaking of change management, it is highly important to facilitate workshops and training sessions to teach stakeholders how to use the KPIs/tools and how they ladder into one another. Numbers will change, and people may not trust that change, which can cause a lot of questions/ confusion. So anticipate that and nip it in the bud!
All KPIs must be recalculated backwards on historical data to maintain comparability over previous periods. Having this functionality increases its relevance and the ability of people to use the data effectively
The Supporting Role of Data Quality & Governance
KPI measurement is useless if nobody trusts the data informing the output.
Therefore, Data Governance and Quality are critical for successfully standardising KPIs. Implementing these helps ensure consistency, accuracy, and reliability across all levels of the organisation.
Let’s start with Data Governance. This domain/ practice establishes clear rules and protocols for business and data stakeholders to work with data. This helps to eliminate inconsistencies that can undermine KPI comparability. A well-structured data governance framework ensures that standardised KPIs are supported by reliable, trustworthy data and that the data is worked responsibly. Surfacing these things fosters transparency and accountability throughout the organisation.
This leads to Data Quality. High data quality guarantees that the information used to calculate KPIs is accurate, eventually leading to more reliable insights and decisions.
In this case, garbage in is garbage out - without consistency and accuracy of data throughout the data lifecycle the KPIs will also be inconsistent or inaccurate, and nobody will trust them.
Data Governance and Quality work hand-in-hand, contributing to the standardisation of KPIs. This primarily concerns the need to create controls for quality and clarity around definitions and lineage. A few examples:
Data quality standards ensure inputs to calculate KPIs are correct
Data contracts collaboratively establish agreements to validate the input of KPIs
A data catalog or lineage tool contains the definitions or traceability of KPIs
This newsletter has talked about this before, and there are several approaches to take to improve your data quality.
The Helpful Aid of KPI Trees
At this point, you probably get the need to standardise KPIs and how to do it. But are there any tools/ frameworks to start doing it?
Luckily, we’ve done this for countless clients at POINT05 using a great framework called the KPI Tree.
Not sure what a KPI tree is? Allow me to explain and check the picture below:
“A KPI tree is a visually, easy to understand overview of all individual KPIs and how they link them together. The KPI tree is designed to show how different KPIs relate to each other and how they collectively contribute to achieving high-level goals within an organisation. It helps employees understand how the business works and how they contribute to overall business success.”
Let’s walk through an example of this framework. The picture below shows how you may do this for the revenue and cost sides at a high level.
In this image, you can see the relationships between the KPIs. For example, on the revenue side:
A low EBITDA is due to a lower revenue
A change in revenue could be caused by a lower number of paid orders
This may be due to a decrease in recurring orders
Recurring orders may be influenced by a churn in customers
If you wanted to keep going, churn could be connected to further low-level KPIs
Overall, we can connect the higher churn causing the lower EBITDA.
The KPI tree contains all the key metrics (KPIs) and their linkage in a hierarchical manner. From this overview, the responsibility for a KPI, calculation method, source systems, frequency, etc, can be added in a table form, creating a KPI dictionary.
For proper implementation, the metric tree should guide the structure of your dashboarding environment and be prominently placed within the dashboarding environment. Some UI/ UX tips to make this happen include:
Using click-through functionality to go to a detailed KPI dashboard from the KPI tree
The tree should be visible for the entire business to avoid definition discrepancies
Departments can create individualised versions for their key metrics (but this should ladder up to the organisation’s view)
At POINT05, one of our clients struggled with misaligned reports and KPIs from different departments. By setting up a KPI tree and standardizing their KPIs across the organisation, their ability to manage the organisation’s performance increased drastically.
Understanding and properly measuring their targets gave them a leg up to hit them, which led to additional VC investment.
Final words
Many organisations don’t have a proper framework for KPI standardisation.
It requires self-awareness that the measured KPIs aren’t consistent and the impetus to do something about it. Then, you need a data team willing to work closely with business departments to standardise these KPIs.
Standardising isn’t that hard. Just remember the four steps:
Assess the KPI As-Is state
Create the KPI To-Be state
Plan the change
Implement the change
Underpin this process with foundational elements like governance & quality and use frameworks to get a better answer. In the end, this activity will save you countless headaches moving forward.
Next week we will continue the analytical topics, getting into the tangible art of dashboarding. More specifically, how do you wireframe and construct a dashboard strategically? I’m tired of people jumping straight to Tableau or Power BI without thinking about what this dashboard means. So tune in next week for a deep dive into dashboards!
Thanks for the read! Comment below and share the newsletter/ issue if you think it is relevant! Feel free to also follow me on LinkedIn (very active) or Medium (hopefully increasingly active). See you amazing folks next week!
 | A guest post by
|