Why data is so important for self-organizing teams
Written by Bart Jonk on 27th October 2020
At Devhouse Spindle (and our sister company Voys) we organize ourselves through Holacracy. This is an organizational model without managers and without functions, allowing people to get the most out of themselves in an environment in which they gain trust, responsibility, and decision-making authority. It is a far-reaching form of self-management.
In an organization like ours, having metrics is of the utmost importance. In this article I’ll explain why we can’t do without. We will explore:
- Why is it important to have metrics?
- How do you choose good metrics?
- How can you facilitate working with metrics?
- What are the consequences of manual data collection?
- Is the investment in time, people and infrastructure worth it?
Why it is important to have metrics
Where are we going?
One of the questions you have to deal with as a self-organizing organization is “how do we determine whether we are still doing the right things?” In a classic management hierarchy, the manager often has management information, and is usually also the person who collects the data.
Holacracy shifts the authority and responsibility to collect and steer to the teams themselves. As a result, you have to determine with each other, as a team, whether you are still doing the right things. By looking at the right information – metrics – in your meetings (and outside of meetings) and being transparent about them, you prevent walking into the abyss while self-managing.
Part of the build – measure – learn cycle
When Holacracy does its work, you see a cycle emerge: building, measuring, learning. The measuring step is an extraordinarily important, but often forgotten, part of making changes. What does work, what doesn’t work and why? Metrics help to answer these questions.
Being transparent allows others to help
It’s extremely helpful to have metrics that objectively disclose the status of your plans, or that indicate the purpose of the team (circle in Holacracy). At Spindle the metrics review is a standard agenda item at every tactical meeting (for explanations of all terms used in Holacracy click here).
Lead links have the accountability to define metrics for the circle and every circle member has the duty to track and report on any metrics assigned to their roles. Rep links have the accountability to provide visibility to the super-circle into the health of the subcircle, including reporting on any metrics assigned to the subcircle.
This transparency makes bottlenecks visible to all those involved for fulfilling the purpose of a circle. By making metrics transparent you create clarity and allow others to help.
Basing decisions on good quality data
Basing decisions on (good quality) data also prevents decisions that are solely made on intuition by the HiPPO (highest paid person’s opinion). When difficult decisions need to be made but there’s no data or data analysis to determine the right course of action one way or another, a group will often defer to the judgement of the HiPPO.
Enough reasons that show how metrics and good quality data for basing those on are important in a Holacratic organization. The next question is: how do you define these metrics?
How to choose a good metric
In Holacracy every circle and every role within that circle has a purpose: a reason to exist. Metrics must indicate whether you are realizing your purpose. Besides this, metrics can also say something about the health of an organization and display data that help to make the right decisions. We can make a general distinction between Team metrics, Strategic metrics and One-time metrics.
How do you measure whether a circle is successful? As mentioned, every circle has a purpose. Circles must be able to determine the way in which they fulfil this purpose (plans, goals and strategy). Therefore, circles should choose their own metrics.
Good metrics are:
- simple (“as simple as possible, but not simpler”) i.e., easy to transfer, easy to calculate and easy to compare with others;
- standardized, via external or internal standards;
- accurate and precise;
- as direct as possible (they measure the underlying process).
Use metrics to show where there are deviations in the plans and use these as a trigger to find out why.
Setting metrics will be a continuous process: through the build-measure-learn cycle, teams will feel calm, clear and confident and adjust plans, goals and strategy on a frequent basis.
How do you measure whether the organization as a whole is successful? Strategic metrics are linked to the strategic goals of the organization.
These metrics should be:
- well defined (unambiguous);
By making these measurements openly available you will enable self-regulation, innovation, learning and control within the organization.
One-time metrics are data you only need once. Examples are “all email addresses that meet these criteria” or to “check once a year how the market share has shifted”.
Requirements for one-time metrics can be different depending on the purpose of gathering them. In general, requirements are less strict than for circle or strategic metrics, also since one-time metrics are generally meant for a smaller audience. Nevertheless accuracy, precision and unambiguity will still be important.
Choose your metrics carefully
Realize that metrics will determine people’s behavior. Therefore, choose metrics with a relative goal (for example, deflecting a trend) as opposed to achieving an absolute goal. Not only measure the result-oriented criteria, but also choose process-oriented metrics that say something about the input in order to get closer to that goal.
By choosing process-oriented metrics you support and stimulate improvements. By preferring relative over absolute goals, you prevent an unhealthy singular focus on achieving the target and getting only what’s measured rather than finding ways to improve in a wider sense.
Facilitating working with metrics
As a company, we want to continue to invest in the availability and applicability of data for metrics. Self-directed teams should be able to easily (and as much as possible themselves) find, use and analyze the right data to steer and improve. This way, teams become even more autonomous.
However, we do not expect data-driven behavior to self-emerge. That all circles will suddenly by themselves start to use data to make decisions, define follow-up steps and other plans, rather than on the basis of gut feelings. Nor do we expect that just telling teams to do so will magically work. There are a number of conditions that favor more data-based decision making over intuitive decision making. The key to success is to positively influence these conditions.
Foster the right culture
The first condition that you can influence is motivation. Holacracy actively makes every member explicitly responsible for being transparent, including the duty to track and share any metrics you are responsible for. This is both an external and internal motivator, because as a Holacratic organization you generally attract people that feel responsible as well.
By fostering a ‘proof-it’ culture, in which ideas are challenged until clear evidence is presented (e.g. by motivating people to set-up A/B tests or proof of concepts) and by fostering an open and transparent culture (making decisions and outcomes visible through presentations, documentation or dashboards), you can make data-based decisions socially acceptable. This triggers the need for people to belong, and thus creates a stark motivator.
Make it cheap and easy
Next, you can lower the barrier to make decisions based on data, by making sure that people are able to use data. This starts with making data easily available: it should take little time, physical and mental effort to obtain and work with reliable data. Training people (enhance their data literacy) is another way to increase data-based decision making over intuitive decision making.
Make sure you trigger setting metrics
Another factor that will help getting data-based decisions as ‘the norm’, is providing enough opportunity to start doing it. You can influence this factor to your preference by triggering people often enough to define a metric for measuring their success. At Spindle we use our cyclic purpose alignment process to our advantage for this.
When starting a new episode (every four months) or when starting a new project, we stimulate people to think about how to measure its success. Once you have determined this, you will have the means during the project or episode to see if your actions have the desired effect. With metrics you create a feedback loop and with that you create stability and the possibility for improvement in small steps.
Consequences of manual data collection
It is useful to make a role responsible for collecting, presenting and interpreting the metric. This collection can initially be done manually. If a metric turns out to be valuable, one should automate the process of collecting and presenting where possible. By first collecting the data manually, the usefulness of the metric is tested before starting automating, a process that can take a lot of time.
Within Spindle this went well for years. We used Google Spreadsheets for part of the data, had many links to third party systems, linked to (external) databases and did the visualization part in Klipfolio. But Spindle has been growing, and gradually more roles got involved. Circles changed and also new systems came in. After 10 years of data, we sometimes couldn’t see the forest for the trees, which resulted in a lot of double work. Also, the number of data questions grew along with the team, so it was time for a better solution.
Investing in people and IT tools
We decided to create a circle to improve the organization with relevant data. The Data & Process circle set out creating a strategy that supports making informed decisions for improvements. The circle provides a place in which validated data for e.g. creating metrics, benchmarking and signaling can be found. The data can be displayed and combined easily.
In our next blog we will explain how we realized this place and how data-warehouse automation helps us to improve our company effectiveness.
Super expensive, extraordinarily valuable
You could say “jeez, isn’t this overkill for an organization that has to provide 125 people with information?”.
We don’t think so. More data insights and good metrics will help the organization to continuously improve. Improvements to our products, the service we provide to our customers, the way we work, the way we measure and scale our social impact, the way we share our story with the world, and much, much more.
By building up history and making trends visible, you move away from targets sucked from your thumb towards realistic forecasts (as meant in the beyond budgeting system we use within our company) and you greatly increase the manageability of the organization.
By ensuring that everyone has data available for analysis, you promote transparency and ensure true self-management.
Every self-organizing team needs metrics to fulfill their purpose effectively. You can use several types of metrics to achieve this purpose, for example team metrics, strategic metrics, and one-time metrics. Getting an organization to actually use metrics can be fostered by:
- having an open and transparent culture in which people feel responsible and in which providing metrics is stimulated;
- making it cheap and easy to access data;
- and employing regular triggers to set up metrics.
This requires a decent investment (in training, people and infrastructure), but that’s worth it.
A transformation to a truly data-driven organization is no easy task. Issues such as cultural factors and education (data literacy) can arise. However, without data that is readily available, combinable and reliable, a data-driven organization remains a utopia.
No comments so far