Switching payroll software

I conducted a comprehension heuristic evaluation setting up payroll and experienced multiple disjointed pathways often leading to dead ends, and a critical lack of explanatory guidance. Talking directly to customers I learnt about the errors they knew to avoid, their pathways and short cuts, and the long cumbersome waiting periods to get the information they need outside of our product.

All of this painted a picture of an experience that was both intimidating, error prone, and confusing for our customers. There were three teams assigned to work on the problem that I needed to work and collaborate with. One looking at onboarding patterns globally, the other setting up Payroll in all regions, and my team: Helping customers with payroll switch.

We initially had a short deadline to deliver, so I worked with my team to quickly mock up the smallest viable solutions by building upon design work from other designers. I suggested we should tackle the most technically challenging component first, the employee upload, ensuring that even if the scope changed, we’d still be able to deliver a valuable, and functional improvement.

While the engineering team was highly supportive of an iterative, agile build, the product team, new to the role, desired a high fidelity end-to-end vision upfront. I consistently advocated for iterative development demonstrating its benefits for de-risking the project, but was unsuccessful influencing a more agile mindset.

To enable the project to move, I committed to designing both a comprehension end state, and the immediate needs of the engineers. Having strong support from the engineering team was crucial, empowering me to continuously refine designs as new insights emerged and the end destination changed.

The core of our solution was an intuitive wizard workflow designed to significantly streamline the complex process of setting up payroll when switching from a competitor. We deliberately chose a sequential wizard format to provide clear, step-by-step guidance, eliminating the need for users to navigate scattered sections of the product. This approach reduced cognitive load and provided a reassuring sense of progress.

The wizard’s central innovation was allowing users to upload their last Full Payment Submission (FPS) file sent to HMRC. From this file, we intelligently extracted as much organisational and employee information as possible, drastically reducing manual data entry for over 80 potential fields. The workflow then focused only on guiding the user through the remaining, essential information needed for a successful setup.

As the project progressed we learnt more about what data fields were needed or not needed. For gathering the crucial missing employee information, I used a large editable table for its familiarity and flexible addition or removal of data as I hypothesised there would likely be changes.

The familiarity of spreadsheets was particularly beneficial for payroll business who could quickly migrate a client’s employee list. To try not overwhelm small business owners, I also introduced tabs to reduce content with mixed results. The headings matched the different sections of an employee record so they would be more familiar with the product.

Using a table was not without challenges. Some columns changed based on prior selections, other columns were uneditable, and some cells needed to retain what was previously written before the change.