New Starter Onboarding Use Case

Fixing the New Joiner Onboarding Process with Robotic Process Automation

There’s one major business process where everyone involved is unanimously agreed, is generally very broken! And that is the new starter onboarding process (followed closely by the related movers and leavers processes).

Onboarding is one process that involves and relies on the highest number of different people, departments, users and systems…it affects your new starters directly, either setting a high note or a low tone as their first ciritical employee experience with your company.

The onboarding process, if successful, should ensure that all of your new starters details are captured accurately, timely and securely, then enters them all in the necessary systems…it should automatically trigger the purchase or allocation of equipment, licenses, tools, permissions and log-ins…and it should do this with minimal disruption and manual effort for EVERYBODY involved. And given that working arrangements are shifting to remote or hybrid and possibly even 4-day weeks and job sharing, the more flexible and accommodating the better.

Unfortunately, this is most often NOT the case.

Overview
New joiner onboarding is an essential process for bringing new employees into the company.

The nature of most onboarding processes lends itself perfectly to automation, freeing up vital resources and time.

Overview
New joiner onboarding is an essential process for bringing new employees into the company.

The nature of most onboarding processes lends itself perfectly to automation, freeing up vital resources and time.

 

Fixing the New Joiner Onboarding Process with Robotic Process Automation

There’s one major business process where everyone involved is unanimously agreed, is generally very broken! And that is the new starter onboarding process (followed closely by the related movers and leavers processes).

Onboarding is one process that involves and relies on the highest number of different people, departments, users and systems…it affects your new starters directly, either setting a high note or a low tone as their first ciritical employee experience with your company.

The onboarding process, if successful, should ensure that all of your new starters details are captured accurately, timely and securely, then enters them all in the necessary systems…it should automatically trigger the purchase or allocation of equipment, licenses, tools, permissions and log-ins…and it should do this with minimal disruption and manual effort for EVERYBODY involved. And given that working arrangements are shifting to remote or hybrid and possibly even 4-day weeks and job sharing, the more flexible and accommodating the better.

Unfortunately, this is most often NOT the case.

An Old & Familiar Foe

Let’s take our use case for instance, a real-life onboarding process that:

  • involves no fewer than ELEVEN different departments
  • includes no fewer than TEN different systems
  • takes an average of SIX weeks to complete (and don’t forget, most employees join 4 weeks from contract signature)

The process is dependent and relies on all the different departments and users doing their bit accurately and on time, which doesn’t always happen. The process also relies on manual tracking, spreadsheets, and follow-up / chasing to ensure the process happens at all, due to multiple data sources and complexities involved. The onboarding process is highly repetitive, overly-manual and open to errors, mistakes and delays. All of which reduces productivity and efficiency for everyone involved and risks employee dissatisfaction and turnover.

To ensure successful, robust automations, process discovery is THE most important step and begins with understanding the “happy path” of what (should) occur most frequently, but this is the easy part. It’s the system and business exceptions that will typically take up the lion-share of development effort.

Also important are the steps which a human operator will sometimes take for granted, for example, “the system is always much slower to respond on Monday mornings during the back-up”, as these have to be considered during the design of the automated solution too. And lastly, leading with the required outcomes will keep everyone laser-focused on the automation imperatives through testing and into go-live.

So what can be done to address this situation and what options do you have? On the upside, you do have several choices, but it depends largely on what’s really broken and what you want to fix. For example:

  1. Build integrations between the various systems using APIs (Application Programming Interface), to eliminate the manual data entry, duplication, errors and delays. However, this can be costly and take time to complete, and not all of your systems may have a suitable API available.
  2. Alternatively, a programmer could develop the necessary unique integrations needed, but again, this can be costly and time-consuming, and will need to be repeated when (not if) any of the various systems change.
  3. You could also explore what, if any, changes to the processes and procedures that might help – this could take time though, and may not result in the benefits needed. Your onboarding process could also cover multiple countries, employee, and tax jurisdictions, which adds complexity and expectations, not too mention the number of people who need to agree to the changes.
  4. You could try to identify anywhere to improve or simplify the collection of data and which users access or update the various systems involved, but again, this won’t always deliver the benefits needed.
  5. You could condsider implementing a brand new system that does all, or mostly all, of the end-to-end onboarding process, but this will take a generous amount of money and time.
  6. And lastly, you could consider outsourcing the entire end-to-end onboarding process to a single specialist services provider, but this will also be an expensive and time-consuming exercise.

You could also combine any number of these options to get your desired result, but the thought of all this may have many HR managers running for the hills. So, what else could you do?

How to Solve This Nightmare?

The Problem & Requirements

The company in this case, not surprisingly, didn’t fancy any of the above options.

They didn’t want to invest more time or money into multiple solutions that didn’t solve all their pains. They also didn’t want to consider other new technologies, as they’d spent several years selecting and becoming used to the ones they had. They also didn’t want to outsource control for the onboarding process either.

Most importantly, they wanted minimum disruption and change within the process and systems, as people were already change fatigued and didn’t have apetite for much more – essentially, they wanted to change everything, but change nothing – a very difficult brief indeed…but, one which we had a solution to…and that was Robotic Process Automation (aka Robots or Bots), which as a non-invasive technology that automates work at the user interface level, requires no integration and can be deployed quickly with minimal disruption or change for the users, ticked all of the boxes.

So, we outlined our solution, our approach, budget, timelines and what the expectations would be for them, in order for us to deliver what they needed. Our demo followed, in order to show and prove our proposed robot solution, and this was presented to a very wide team of decision makers, influencers and process owners.

The result was a resounding yes to proceed, and we began the detailed discovery work to understand and document the current processes and systems, and agree the scope of work and timelines. Our proposed solution was simple and exactly what they needed – “change everything, but change nothing”. We agreed a few critical principles:

  • No system changes – from a user perspective, no new or additional systems to adopt, use or learn
  • No data changes – no new data points or sources to be obtained or collected and nothing to leave the company’s control or environments
  • No major change management and minimal process change – everything had to appear exactly as it was now, only much better
  • Employee experience had to be significantly improved and the time to complete the onboarding process had to be at least halved, down to 3 weeks on average

We started by examining the available existing process and system documentation (which wasn’t much), and with a combination of process and task mining tools and working with key stakeholders, the existing processes were first mapped in detail (down to screenshot  and keystroke level). This not only confirmed the extent of the known issues, but identified some unknown ones too, all of which were filtered into the solution specifications and requirements.

With the “As-Is” view clear, the “To-Be” design followed, adhering strictly to the agreed principles. At this point, a number of decision-points and design principles emerged, which involved the following:

  1. Activity re-sequencing – a number of activies could be done in the same order as they were currently, but this would have been sub-optimal. Changing the sequence for optimal robotic performance was proposed and approved.
  2. Activity elimination – a number of activies existed due to human operation, but were unecessary and could be entirely eliminated with an automated process, this was also approved.
  3. Activity simplification – a number of activies were discovered that were overly-complex and complicated and added no real value, to either human or robotic operation, so simplifying these was approved.
  4. Some activities initially thought “unautomatable” were found to be entirely automatable, and therefore approved to be included in scope.
  5. A variety of non-standard process deviations had crept in over time, especially documents and spreadsheets that were being used. Standardising these documents and tasks was also proposed and approved.

Both the “As-Is” and “To-Be” were then reviewed and examined by more stakeholders to ensure we had an accurate view of what was happening and what would happen in future, and with those approved, we proceeded with the technical architecture design, which the robot developers would use to build and test the solution.

The Approach Taken

The Challenges Encountered

The current process initially presented some challenges in capturing all of the steps and the stakeholders involved, and the timelines to complete them varied significantly from case to case, but ultimately, the final process was defined as per figure 1. The diagram has been simplified slightly and anonymised for the purposes of presentation. What is clear from this process map is the sheer number of people and departments involved and the amount of manual tasks and dependencies

Figure 1: As-Is Onboarding Process

Working from the very start of the onboarding process, where recruitment becomes employment, the first priority was to make sure the offer letters and contracts were at least digital. This was a relatively simple step as the company already sent documents via email or DocSign, so both formats were built into the solution for the robots to handle.

Next up was to address the current lack of links between the multiple systems, so that data could pass freely and automatically between them:

  • The recruitment and background check systems are both external service provider web-based tools that didn’t have suitable API’s, but no problem for bots, which could simply access the tools the same way users already were – just user names and passswords for the robots.
  • The IDAM, HR, Service Ticket, and Payroll systems are all internal on premise or cloud tools, and so very easily accessed either by API or the UI. The data sources, fields and selectors were all clear from the process discovery and design work already done. The robot was programmed to access all systems automatically and in the correct sequence to ensure data was created and updated at the right time. Some “human-in-the-loop” sequences were also incorporated to facilitate exception handling and approvals where necessary. This ensured 100% of data entry was automated across all systems, in real-time eliminating delays and errors.
  • Lastly, the robots were programmed to check equipment inventory, raise requisition orders for hardware and software and create essential user account permissions and access, because these systems too were all internal and easily accessible.

A number of automatic notifications via the systems, emails and reports were developed to communicate and inform key stakeholders throughout the onboarding process, including hiring manager, HR, payroll, finance and IT.

How We Solved

The Outcome

Figure 2 shows a high-level overview of the new automated onboarding process, which you’ll see is significantly simplified, with 80% of the process now handled automatically by the robots, based on a few simple human triggers and oversight from output reporting and exception handling.

The biggest advantage of the automated solution has also been the biggest change to manage for the company, and that’s the speed of the new automation-enabled onboarding process. This soltuion has highlight and magnified other weaker or less efficient areas of their business, such as the external service provider relationships, but this is a small price to pay compared to the combined hundreds of hours in savings per year, the improved ability to scale and vastly better employee experience. They’ve also discovered that the related processes within IT and accounts will need to be improved or automated too, in order to keep pace with the speed and demands of future joiners, movers and leavers, which is where the focus is shifting to next.

Figure 2: To-Be Onboarding Process