Data process automation - DPA, if you prefer the acronym - is one of those terms that sounds more technical than it is. It conjures images of enterprise IT projects, development teams, and lengthy implementation programmes. For most finance and tax professionals, it feels like something that happens to other organisations with larger budgets and more specialist resource.
That picture is wrong. And the gap between what DPA actually is and how it is perceived is one of the reasons so many teams continue doing manually what could be done automatically.
What is data process automation?
At its simplest, DPA is the use of software to perform data tasks that would otherwise be done by a person.
Not complex tasks that require expertise and judgement. Repetitive, rule-based tasks: extracting data from one system, reformatting it according to defined rules, validating it against expected values, and loading it into another system. These are the tasks that consume significant time in finance and tax teams - and they are, by definition, the tasks most amenable to automation.
The tools used for DPA range from relatively simple (scheduled scripts, Excel macros, robotic process automation) to more sophisticated (workflow platforms like Alteryx, API-based integrations, ETL pipelines). The right tool depends on the complexity of the task, the volume of data, and how often the process runs.
Five benefits
1. Efficiency The most obvious benefit is time. A process that takes an analyst three hours to run manually takes a well-designed workflow three minutes. That time can be reinvested in review, analysis, and judgement - the work that actually requires a qualified professional.
2. Accuracy Manual data processes are a significant source of error in financial reporting. Transcription mistakes, formula errors, copy-paste failures, version control problems - all of these disappear when a defined, tested workflow handles the process instead. The same inputs always produce the same outputs.
3. Productivity Automation frees people from repetitive tasks that are low-value relative to their skills. A tax manager who spends two days per quarter reformatting trial balance data is a tax manager who is not reviewing computations, managing relationships, or thinking about technical issues. Automation restores that capacity.
4. Scalability A manual process scales linearly with the volume of work - more entities, more time. An automated process handles additional entities without proportional additional effort. For groups adding entities through acquisition or expansion, this matters.
5. Competitive advantage Teams that automate their data processes have more capacity for the work that differentiates them: analysis, judgement, commercial insight. Those that do not are progressively outpaced by those that do. The advantage compounds over time - see Moving Fast by Starting Small for the maths behind why.
Where to start
The most common mistake is trying to start too ambitiously. A comprehensive automation programme for the entire compliance cycle is a large project with a long lead time and significant implementation risk.
A first automation for the most painful single task in your current process is achievable in weeks, not months. It produces demonstrable value. And it builds the team’s confidence and capability for the next step.
The criteria for a good first candidate:
- The process is repetitive (it runs at least quarterly, ideally monthly)
- The inputs are predictable and consistent
- The transformation is rule-based (the same rules apply each time)
- The output is well-defined (you know exactly what a correct output looks like)
- The process currently takes disproportionate manual effort
Most finance and tax teams have three to five processes that meet all of these criteria. Pick the most painful one.
Real examples
Data collection and normalisation. Trial balance data from multiple ERPs arrives in different formats, with different account structures. An Alteryx workflow collects the files, applies a standardisation process, and produces a single clean output in the format required by the tax platform - in minutes.
Data cleansing. A raw TB extract contains coding errors, missing accounts, and intercompany mismatches. Rather than fixing these manually each period, a workflow applies validation rules, flags exceptions, and produces a cleaner dataset. Recurring errors get fixed at source; edge cases get flagged for human review.
Trial balance loading. The clean TB output is loaded directly into ONESOURCE or another tax platform via API, rather than being uploaded manually. The platform receives data in exactly the right format, every time.
Reporting. Formatted reports for specific entities or periods are produced automatically from structured data, rather than being assembled manually in Excel. Prior-year comparatives are included without additional effort.
A practical mindset
The goal of DPA is not to eliminate people from finance and tax processes. It is to eliminate the parts of those processes that do not require their expertise.
The best outcomes come from teams that approach automation with a clear question: “Is a qualified professional the right person to be doing this step?” When the answer is no - when the step is mechanical, repetitive, and rule-based - that is the right candidate for automation.
Starting with that question, applied honestly to the current process, usually produces a short list of high-value targets for a first automation project. The rest follows from there.
If you would like to discuss what a first automation project might look like for your team, get in touch.
