Numbers, Numbers Everywhere…
Samuel Taylor Coleridge related his ‘Rime of the Ancient Mariner’ in which the grey-beard loon narrates his story of torture at sea after all his shipmates went down. Now you know how service managers came to be as they are.
Why is Reporting Such a Problem?
I have seen, and indeed found myself in the middle of, a number of service environments where nobody was in the least happy with the reporting provided. All data should be provided to support a purpose. Anecdotal evidence suggests that where the purpose and role is clear and stable, the reporting tends to be good too. The question of “how to ensure reporting is effective?” is thus addressed first by addressing clarity of purpose. In this article I shall pass over “informatics”. Good design most definitely enhances the speed of assimilation and impact of information.
In a small service or organisation, extensive work to generate elegant written and graphical reports will not be justified, and you will work with what your suppliers have readily to hand. In larger services, greater effort will be warranted. One of the best examples I have seen was from my days as an accountant. The CEO required a two-page letter on the state of the business each month. Although led by finance, it addressed aspects including HR and change programme progress. This brief presentation of the essential facts was supported by extensive data as raw material and was delivered through an interactive discussion, the written version later being widely circulated.
Problem 1: Clarity of Purpose
There are many ways in which a customer may choose to cut the different purposes of governance between a number of bodies. Most end up using a broadly common model, whilst some find that because of the size, complexity or otherwise of their service it is appropriate to combine or separate these functions. However you choose to divide functions between various boards and meetings, each convening on its own schedule, each must have its terms of reference defined. Good terms are collectively exhaustive and mutually exclusive, so it is clear which matters are addressed by which bodies. The terms should also contain a pro-forma agenda. So if one is to review performance against the SLA, the reporting should contain current data to permit this.
To achieve brevity, the authors of the report need to ask themselves “what is the board really interested in?” and filter the data accordingly. Thus, it is generally safe to start with the assumption that if all is going swimmingly, the headline statement is a large green blob, supported by the key data and a clear statement “all is well!” The implication of this is that there is someone who considers the data before it is presented to the board and has a good look at it. That person needs to understand what is important to the board, how they like to work, what is really going on and how to present a summary of the data that tells this as a story. The topics to be addressed follow from the services within scope and the Board’s terms of reference. This is a skill that can easily be assessed, but cannot be bought off the shelf as COTS software. It is also different every report, although the shape of the story and the leading characters are likely to be consistent. If the board is irrationally inconsistent, or the appointed person is too dense or inexperienced to understand their purpose, you need to change something! Have you resourced this role appropriately? In a multi-supplier model, this work will fall to the SIAM.
Problem 02: Mental Meandering
Many reports wander through a series of points without ever reaching a conclusion or clarifying an issue. At the height of the Second World War, Churchill wrote a masterly memorandum to his War Cabinet:
To write clearly (or present pertinent statistics), the challenge is that one must think clearly. In this, Churchill provided a fine example. Wonderful to observe; difficult to emulate. For lesser mortals needing help, Barbara Minto’s Pyramid Principle provides an excellent guide (see www.barbaraminto.com).
For a service report, the critical questions are frequently…
1. Is the service performing as well as we would hope and expect?
a. Identify good areas for congratulation (SLA and KPIs actual vs. target and trends over time). Understand why favourable exceptions have occurred and seek to deliver even more value.
b. Identify poor areas and drill down to uncover the causes and recommend corrective action. Debate as necessary to ensure that effective action is identified and taken.
2. Are the relationships functioning effectively?
3. How much is the service costing?
a. Analyse any variance from budget (volume, price, mix)
b. Identify any commitments requiring authorisation
4. Is the service delivering the value expected?
a. Are we getting the business benefits expected?
b. Are the obligations being delivered? What is in exception?
c. Does the contract need to be amended? Identify the implications of change and authorise where appropriate.
A governance body that asks no questions is stupid: it does not learn. A body that takes no decisions or action is useless: it makes no difference. Dr Bharat Vagadia’s recent article in Outsource (see ‘Only Strong Leadership And Governance Can Deliver High Performance’, Outsource #33), is a fine review of the essentials of effective governance. (For more from Dr Bharat Vagadia, see his columns for Outsource online, indexed now at www.outsourcemag.com/authour/bharat-vagadia/)
Problem 3: Measure Appropriately
IT Service Management suites can report vast sets of data. You need this to be able to drill down into an issue. Other than for this purpose, desist! Take Churchill’s advice if your service is small: report verbally and refer to the backup/appendix only by exception. In this case the report to governance is verbal and the appendix is printed. Make the presentation clear, concise, comprehensive, informative of performance and the issues. If there was concern last month about supplier X’s incident resolution record, think ahead and have the facts to hand for this month. State performance in numbers and plain English. Drill down to understand and communicate the contributory factors, and expect to be quizzed on this as the governance board validates your conclusions. Recommend congratulations or therapy as the data dictates.
Measures should be understandable. Look for the simplest numbers that do the job required. If you tell someone “this is the problem I am trying to solve, and this is the number I am tracking” they will normally understand you and your actions, thus are able to work with you. Look for a measure that changes your behaviour. If the indicator goes up, do you know what you should be doing? Many measures are introduced to support the diagnosis and correction of a problem. Once the problem has been solved and the data has been seen to be stable, it need no longer be reported as it no longer drives behaviour.
Governance is all about decisions. Decisions should be based on timely and appropriate facts, reliably assessed. The question of “what facts should I measure to present?” follows from the decision to be taken, which in turn is related to the terms of reference of the body concerned. There is a temptation to regurgitate half-chewed data in the manner of a cat presenting on a doorstep. Both promote revulsion and a violent reaction. ITIL does not help here. For each process, it suggests a list of possible measures to cover every eventuality. Whilst all are occasionally useful, particularly for those directly involved in doing the work of delivery, few should regularly be presented. Two to three measures per process per service is a reasonable rule of thumb for inclusion in a contractual SLA. If something is working well, 15 pages confirming the fact is overkill. If however something is not right, the access to meaningful data to isolate the cause and recommend action for decision is vital. This is where ITIL’s list is useful in recommending performance indicators.
Some vendors and customers seem determined to operate by guesswork and intuition alone. If the only evidence given confirms the shortfalls of the presenter, it is down to the governing body to throw out the report and the reporter, demanding they return when they have properly considered and appraised the situation. If Churchill can do it to a general, you can to a service manager or supplier. There can be no assurance of consistently good performance without the data to confirm it is on track.
Customers get the service they deserve. If the appointed governance body pays no attention to the facts or does not hold the supplier effectively to account, they are careless of their duty and should not be surprised if the supplier reacts by cutting costs and performance levels. This is especially so where procurement has driven prices down. This is an explanation, not an excuse.
Problem 4: Think Ahead
Kaplan and Norton’s seminal “The Balanced Scorecard” was hugely influential in redressing the way companies govern their performance. Too often, this had looked only at financial outcomes, to the neglect of indicators of input that could be used to steer future performance. Similarly in the Lean and Six Sigma areas of process control that grew out of the Quality movement (ISO 9000, ISO 20000 and the like), there were concepts of leading and lagging indicators. A lagging indicator reports an outcome of action e.g. £ spent, % availability for a service. A leading indicator is useful when it is predictive of performance. Action on leading indicators, taken in time, is used to keep a process; a service; a project under control. The statistically minded may use control charts to assist.
The implications are communicated through a story, leading the governance body to the conclusion that action is required and indicating what that may appropriately be. “Project X has missed the last three milestones by two, three, four weeks respectively. We looked into it and suspect that supplier Y is short of skilled developers and has put only half the expected numbers to work for us, and they are the B team. The options are ….”
A good report balances information on outcomes achieved with the prospects for continued success; informed by data on lagging and leading indicators. The governance body should not take what they are told on trust alone. It is their role to probe the thinking to ensure that the diagnosis is accurate and the recommendation is optimal, before committing the resources and actions to follow through. It then needs to track the actions and outcomes to conclusion to follow-through on its decisions.
Good reporting depends on clarity of purpose, understood and applied in advance of review to prepare a story of performance and what needs to be done to sustain it.
The story needs to be informed by fact, not drown in it. Coleridge’s Ancient Mariner was cursed to report the tale of his crime and the doom of his shipmates. Your governance review may well lack the pathos of that great poem, but the story is at the heart of moving people to action and delivering results.
This article originally appeared in Outsource magazine Issue #34 Winter 2013.