Phoenix, frankly speaking: Part III

Back to our swamp mapping, so we can get to that mountain (mountains are nice and stable) of a stable, functioning as intended, pay system. Right indicators= good map building.

Business process mapping

One of the key reasons Phoenix is failing is because gathering of critical knowledge held by compensation advisors, as well as critical business process mapping — particularly the intersection of the HR processes to pay processes — wasn’t done. Do it and measure:

· Number and percentage overall of business processes mapped

· Number and percentage overall of knowledge transfer exchanges

· Number and percentage of mechanisms for knowledge transfer between expert and novice compensation advisors

· Number and type of capability building for community of compensation advisors

Roll these guys up into an index for overall current state mapping of processes and knowledge about those processes in the compensation advisor community.

Defect management

Defect management (actually, waste management, in the business processing world) is a huge headache — the burden of which is almost exclusively borne by public servants. The employer can get a sense of the scope of the issues by tracking the following indicators, and working to reduce or increase measures reported where necessary.

· Percentage of files requiring re-opening

· Percentage of complex files vis a vis ‘easy’ ones (you’ll need some criteria here)

· Percentage successful transfer from of defect ‘fixes’ from sandbox environments and pilots to live operations

· Number and type of new errors introduced into system

· Number and type of defects (including functionality defects; integration defects — we know there are at least were 69–19 major; 46 moderate; and 4)

· Time to resolve defects (Days? Minutes are better, but if it has to be days, fine, but aim for hours)

· Number and type of manual workarounds required (by type of transaction, e.g. retroactive transactions)

· Time to convert manual to automated

· Number and type of change orders for systems integration (with rationale and associated resource profiles)

You need to set timelines for completion and report on progress towards those completion deadlines. And yes, roll these up into a ‘Defect Management’ index if you’d like, just make sure you can disaggregate the data when you need to.

Comparator data

Some comparator indicators will provide some useful data to assess Phoenix stabilising efforts — you want it to perform at least as well as the outdated systems being used in the departments that have yet to migrate to the new HR/Pay system. At a minimum, you want to gather data and compare against what was committed to in the 2011–2012 Report on Plans and Priorities by PWGSC — now renamed Public Service and Procurement Canada — PSPC. (If you don’t have the data because you never resourced the data strategy to capture that data, just say so, but I find it challenging to believe no system performance data was tracked). Those indicators were as follows:

· Decreased time to train pay administration staff

· Decreased cost to deliver pay administration

· Decreased materiel usage (e.g. paper, mail)

· Decreased work effort for Compensation Advisors

· Increased pay administration staff retention (commonly known as ‘churn rate’ — nice, huh?)

· Increased number of accounts per Compensation Advisors

I can see that some of these measures may be a little off-putting in addition to creating perverse and unintended consequences, but balance that against the other lines of data you’ll soon have coming online.

Organisational characteristics of roll-outs

We know that many organisations have retained their own compensation advisors. We know that many organisations have not yet migrated to Phoenix (32%). We know that many have not migrated to Peoplesoft (MyGCHR) (only 44 have). But are we talking about the same organisations? With clear profiles of these different departments, and measurement of the following error rates, we could get some insight into sources of adverse system performance.

· Error-rates for organisations that have in-house vs Miramichi located compensation advisors (is distance a factor?)

· Pay error-rates of Phoenix vs legacy systems

· Pay error-rates of departments using Peoplesoft vs those that aren’t

· Churn rates for compensation advisors (disaggregated by Miramichi; in-house; re-hires by location)

· Percentage breakdown of compensation advisors retained (departures, relocations and new recruits)

· Percentage and type of training taken by compensation advisors

· Percentage and type of training taken by managers and employees

The indicators suggested are not exhaustive — for instance, you really should have some robust client satisfaction indicators (index, even). I recommend looking at the UK National Audit Office model — it’s used around the world, and works across public/private sectors. Good stuff (also: free).

Roles and Responsibilities (Governance)

Sound project management requires that roles and responsibilities not only be understood, but communicated. That is a central feature of the Results Agenda, and its supporting Policy on Results as well. It is now clear that this aspect of the file does not meet expectations, given the confusing chain of accountabilities and/or overlaps that have been identified. It is critical that this be addressed, supported with an implementation plan with defined accountabilities; activities; milestones; resource profiles etc.). A revised Treasury Board submission, horizontal in nature, with one lead (TBS), and partners having short, medium and long term expected results is what is needed.

Flowing from the Ministerial accountability model, and the publishing of Mandate Letters, it would be advisable to ensure that commitments for supporting the Phoenix transformation be included in the performance management agreements of deputy heads, with progress publicly reported — consistent practice with that of their Ministers. Performance bonuses should be withheld unless expectations are met on this file. (And no gameable indicators such as ‘held 6 meetings with fellow DHs’ in the performance agreement).

Joint employer-employee committees exist at the technical level, but committees by their nature are relatively passive instruments of information sharing — the needs of the employees dictate a more engaged response. There needs to be a joint task force at the strategic level with a mandate to act that has terms of reference shared publicly. This task team should report directly to the Clerk of TBS and bring together the key stakeholders: unions, employer, subject matter experts, and an external audit partner (acting in a capacity similar to a Departmental Audit member). Meetings should be publicly reported on, including position of persons attending (to nip the delegation issue in the bud).

Open and Transparent Operations

There are relationships to rebuild around this file, and trust. Adopting a position grounded in openness and transparency is a concrete stance from which to rebuild that trust. Every day, the same question is asked around this file: how did it get to this state? Some of the reasons for the current state were mapped out in the Lessons Learned Report, but there is more to tell that will provide necessary indicators about what’s gone wrong. For one thing, the employer can make easily findable the Project Complexity and Risk Assessment analysis that had to have been done for this project. This would provide a sense of where misjudgements and under-evaluation of risks occurred, to support corrective interventions going forward.

The employer can also provide evidence that Lessons Learned are being applied. For instance, when the Clerk received the Deputy Head responses on Nov. 16, these can be published as tangible evidence of concrete steps being taken in one organisation that might be usefully applied elsewhere while also provide demonstrable action on the part of the employer to manage the harm that has come to employees (e.g. they can provide their management action plans with associated accountabilities; resource profiles; milestones; completion rates on interventions, etc.).

The employer can commit to a review and update of the project business case for this transformation initiative (likely inseparable from associated TB submissions on this file) and select key performance indicators from associated Treasury Board submissions to report on publicly — in line with the Mandate Letter commitments for the PSPC Minister, reported on the Canada.ca/results site. In fact, many of the indicators proposed above, in the form of an index, would provide valuable performance information on progress towards expected results. If all else fails, there needs to be a Plan B. Work should be underway already to devise one, and when finalised, it should be communicated.

So I have outlined some suggestions and provided many concrete appropriate indicators that can be used to map the current state to get us to that future stable state. I’ve summed up the key recommendations here — to be taken in conjunction with the indicators outlined above:

· Leverage innovation in HR staffing processes to stand up cross-functional teams to tackle this file

· Commit senior management of departments to spend time on the shop floor, walking the line, working the files at the coal face

· Create cross-functional, joint employer-employee task team with a clear mandate to act in published terms of reference

· Create an anonymous early warning system to surface risks, hazards, and issues

· Engage in meaningful engagement with the compensation advisor community that starts with a public apology

· Publish a regularly updated Expected Results management plan with metrics related to areas such as the effectiveness of the business transformation as well as the efficiency and quality of operations

· Map the current state in order to plan, risk manage and project manage the solution process (whatever that solution state is must be informed by accurate system performance information — use mine, they are good)

· Publish the roles and responsibilities (horizontal as well as vertical — TBS is lead accountable)

· Publish Deputy Head performance management commitments on this file and have them monitored and reported on

· Release the PCRA on this project; as well, Mandate Letter indicators on this file going forward that are monitored and reported

· Releaser the Deputy Head responses of Nov. 16 to the Clerk, with any associated indicators and current performance

· Review and update the business case for this project and make public any associated Results Appendix

Phew. There you go. Not sexy, just concrete, time-tested approaches to get us out of the Phoenix fiasco.

Phoenix has stumbled badly. Responses now need to demonstrate agility, openness and responsiveness — key elements in any case for the culture change that needs to happen in the public service generally. It is possible to use Phoenix as a test bed of employer commitment to leading the culture change we need to see.

And at the end of the day, the employer must act in recognition that employees are being harmed daily by Phoenix — some severely. If this group of affected persons were not public servants, this situation would have panned out very differently.

I remind those who are reading that those employees are Canadians, citizens, first, and public servants second. They deserve results.

Part IV provides a final, missing element to move this disaster to a steady state.