When people hear that I submit a data return, they often imagine something fairly contained.
A dataset.
A deadline.
Some validation errors.
A sign-off.
And yes, all of that is there. But that description misses the part I have become most interested in over time: what actually happens in the system when I do this work.
Because when you are the only person in a small provider responsible for a statutory return, you are not just submitting data. You are sitting inside a web of relationships, incentives, assumptions, and feedback loops that extend well beyond your own organisation.
This is where the systems thinking question starts to matter:
What do I do when I do what I do?
On paper, I submit data
Formally, my role looks quite narrow.
I liaise with HESA.
I keep an eye on sector interpretation through communities like SROC.
I ask colleagues in other departments for clarification, corrections, or missing context.
The output is a compliant return.
But that description treats the work as linear. In reality, it is anything but.
In practice, I shape how the organisation understands itself
Every question I ask a colleague carries an implicit message.
Sometimes that message is:
• This matters
• This definition is important
• This needs evidence
Other times, unintentionally, it might be:
• This is a bureaucratic requirement
• This is something “the data person” worries about
• This only matters once a year
None of that is written down. But it is learned.
When I push back on an assumption, I am not just correcting a value. I am signalling what kind of accuracy is expected.
When I accept an approximation, I am not just meeting a deadline. I am teaching the system what is “good enough”.
Over time, those micro-decisions accumulate into organisational behaviour.
I also act as a translator between systems
In a small provider, the data return does not sit neatly within one function.
I am constantly moving between:
• External definitions and internal practice
• Sector-level expectations and local realities
• Formal rules and informal workarounds
What looks like a single submission is actually a series of translations.
Systems thinking helped me see that translation is not neutral.
Every time I simplify something to make it understandable, I also decide what not to carry across. Every time I contextualise a rule, I choose how flexible it becomes in practice.
The system adapts to those choices.
Feedback loops I didn’t see at first
For a long time, I thought my job ended with submission.
But the effects loop back.
• How colleagues respond next time I ask for data
• Whether issues are flagged earlier or later in the year
• How much confidence people place in numbers once published
• Whether data is seen as a shared responsibility or a specialist chore
None of these outcomes are explicitly designed. They emerge.
The return becomes less about compliance and more about conditioning. It teaches the organisation how to behave around data, effort, and accountability.
This is not about blame or burden
It would be easy to read this and feel the weight of responsibility grow uncomfortably large.
That is not the point.
Systems thinking is not about asking individuals to carry more. It is about noticing the influence we already have, often without realising it.
I did not create these systems. I work within them. But I do have leverage points, simply by virtue of being the person who connects them.
Once I noticed that, the question changed.
Not “How do I get this return done?”
But “What kind of system is being reinforced because this is how it gets done?”
The question I now sit with
When I submit a data return, I try to hold a quieter question alongside the technical ones.
If this process repeats year after year:
• What becomes easier?
• What remains fragile?
• Who learns, and who opts out?
• What kind of relationship to data is being normalised?
That question does not slow the work down. It sharpens it.
Because submitting a return is never just about the data. It is about the system that learns how to produce it.
And once you see that, it is hard to unsee.