Pensions auto-enrolment has thrown a harsh spotlight on payroll and HR data systems, finds Robert Gray


Plenty of attention has been devoted to the take-up rates and timetables associated with auto-enrolment. Less talked about, but just as important, is the knock-on effect pensions legislation will have on HR data – everything from data protection, through to accuracy and ensuring there is sufficient processing capacity to meet demand. The size of the problem will depend on corporate culture, complexity of the organisation and the kind of HR and benefits systems already in place. But it is never too early to start. As they say, forewarned is forearmed. Here we outline four key problems and their solutions:

1 Changing staff circumstances

The problem

Superficially auto-enrolment eligibility criteria seem clear. All workers must be offered the option to enrol if they are aged between 22 and the State Pension Age, and as long as they earn at least £8,105 a year (the same level as the personal allowance for income tax). However, the real world is not as straightforward as these bald facts suggest.

As well as ensuring systems are in place to identify when younger employees reach 22, employers will need to factor in any pay rises – overtime, bonuses or commission – that suddenly take previously ineligible employees above the minimum earnings threshold and kickstart a process that asks them whether they want to opt-in or not. Conversely, bosses must also bear in mind the impact of offering salary sacrifice benefits, such as childcare vouchers, which have the impact of pushing borderline employees below the income threshold.

“There are some companies where HR and payroll are intrinsically linked, for instance if they are offering salary sacrifice,” says Matt West, chief sales & marketing officer at employee benefits company Benefex.

“But for other companies, where there is no history of payroll having access to personal data, plugging it all together is creating many issues they need to tackle.”

What to do

  • Water-tight processes will need to anticipate or promptly deal with age changes, income changes, any opt-in requests and opt-out notices.
  • Processes will also need to sweep people back in during any re-enrolment exercise, and pick up people who may not currently be eligible, but who perhaps become so in the future. Every three years employers must give eligible jobholders another chance to re-enrol.
  • Systems will need to be re-programmed to make them easily updatable because companies will also have to keep a close eye on government policy. Although the income threshold currently sits at £8,105, the coalition is ideologically committed to raising this, in successive Budget statements, to £10,000.
  • Employers will have to keep their eye on the auto-enrolment ball during the complete life cycle of each and every worker.

2 Integrating disparate systems

The problem

The previous problem presents the next: moving data from payroll and HR systems into pension systems – in bulk and much more quickly than before. The main issue with many organisations is that HR departments operate legacy-type systems and therefore bespoke tailoring of existing applications is necessary to meet the new data requirements. But the challenge is much bigger for larger companies, for whom collecting data is more time consuming. With greater scale comes more changes to incorporate; for instance HR and payroll systems may not be from the same supplier. There could be discussion about whose responsibility it is.

“Out-of-date systems in HR will not be able to meet the criteria for the new demands of auto-enrolment data legislation,” argues Capita Hartshead client strategy director Paul Sturgess. “In addition, different operational areas within a business do not often notify HR of changes relating to staff. Local managers who don’t have the time to update records can then be the source of bad data.” 

What to do

  • One option is a ‘middleware’ solution, usually referred to as a data hub, that will interface with an employer’s payroll, HR and pension systems to carry out the detailed checks required on a per-pay-period basis.
  • Not all hubs are created equal. Some hubs may simply be provided alongside existing HR or payroll systems and only take data in certain formats without managing exceptions. In these instances, it will be up to the employer’s in-house team to ensure the data from the hub meets auto-enrolment requirements. Other hubs may carry out a full validation check and, either through software or human support, fix the problems.
  • Ideally, a good data validation tool will provide a detailed report on all data quality issues, not only categorising the problems but also assessing the skill levels needed to resolve the issues and the likely costs for the data cleanse project.

“A good auto-enrolment solution should run largely unattended and manage the end-to-end process including categorisation, production and issue of electronic communications, self service functions to capture employee decisions, automated scheduling to sync with payroll cycles and management information to prove compliance,” says Wayne Morris, senior consultant at Xafinity Claybrook.

3 Information overload

The problem

Payroll software may hold the answer to accurately assessing employees’ earnings and making pension contributions, but it is also the quantity of data being managed that needs considering. “The sheer volume of data means all IT processes need to be highly automated,” says Geraldine Brassett, client relationship manager at Aon Hewitt, and member of the benefits administration team. “But everything has to be done in bulk too.”

What to do

  • Employers must seek confirmation from their payroll provider, or software provider if payroll is in-house, about the extent to which their existing software will enable them to comply with their obligations.
  • For technology to successfully handle this challenge there needs to be a smooth implementation process in advance of the employer’s staging date. During this time employers should ensure the data the technology is using is of good quality.
  • If hubs are being used, they should be designed to be able to identify missing or incomplete data which needs to be updated.
  • Employers will also need to allow time for making updates in advance of their staging date. Friends Life head of corporate benefits marketing MartinuPalmer recommends starting the data cleaning process around six months in advance of going live.
  • Implemention processes are needed to ensure that new hire data is captured and kept up to date.
  • Data challenges will arise no matter which pension provider the employer selects.
  • There will still be the need to maintain accurate records of enrolments and contributions.

Morris of Xafinity Claybrook, recommends drawing up an action plan for resolving data issues complete with milestones and check points. “Data is a living, breathing commodity,” he says. “As such it is important to constantly re-validate all data during the cleansing project to ensure previously good data has not, in the meantime, been unintentionally ‘corrupted’. Your data tool should manage this validation for you and provide graphical progress reports.”

4 Audit trails

The problem

Without auditing and verifying information already held in personnel and payroll systems, there will be a risk of errors being made and some workers slipping through the net. The Pensions Regulator requires an audit trail showing details of how opt-outs have been handled, and numbers of opt-outs. These details need to be kept for six years, with the exception of opt-out information which needs to be stored for four years.

Any shortcomings in the rest of a company’s technology solution will become apparent in the audit trail, and could potentially lead to fines. For example, if a system merely identifies eligible job holders, but takes no action on that information, it will require further work from the employer to become compliant. “It’s not just the scheme, it’s the workflow that goes around it which is important,” says Jonathan Reynolds of RSM Tenon.

What to do

  • One of the ways the Regulator will identify possible offenders is by looking at employers with suspiciously high opt-out rates. It is therefore vital employers keep accurate records for staff who opt-out, confirming not just their name and the date of their decision, but also the reasons why, so underlying trends can be easily explained to the Regulator.
  • These records can be requested by the Regulator at any time. In particular employers will need to demonstrate that opt-out requests were not the result of any inducement action by the employer.
  • Brassett of Aon Hewitt, says many employers are finding the idea of outsourcing this area attractive, as they do not want to struggle with the details of compliance: “There are a lot of rules. It comes down to whether you can satisfy someone else, after the event, that what you did was compliant.”
  • Employers may wish to benchmark their statistics against sector averages, to highlight any industry-wide problems outside the employer’s control.