How many times have you been asked ‘We have a new system being implemented next month, can you sort out the data?’. Okay this might be a bit extreme, but the proposition holds. Data is rarely considered in the same way as the other asset classes – finance, staff and estates.

This needs to change. Data is the fluid over which processes flow. If the impact is ignored or quality is assumed, this fluid can quickly turn into a grit. However, in many organisations, a lack of change governance is incompatible with a rigorous impact analysis approach. Leaving a combination pragmatism and education as the right approach.

Pragmastism is asking the same simple questions every time.  Education is finding a way to present these findings to show the value of treating data as an university wide asset. Visualisation is key to signposting the impact of any change. The HE capability map could be deployed successfully here. I showed how this works in a previous post.

The output of any assessment is to understand how much governance is required. It is a quick and efficient triage activity to categorise and rate the opportunities and threats to the data asset- and then to respond appropriately.

The template below summarises my approach, based on working with many institutions at various levels of data maturity.

1: What’s the value of the change?

In a perfect world, this would be aligned to university objectives demonstrating why this has priority and  focus. Mostly though value is not even qualitatively calculated. It needs to be for two reasons 1) so it can be judged against competing priorities and 2) so it can be assessed against the cost of the change

2: Who benefits?

How big and important is my ‘benefit cohort’? For example, is this a narrow but time consuming change for a few senior staff, or is it an initiative to solve operational problems for many people? The answer should be linked to 1: and it gives the idea of the breadth of that value. The sweet spot of course is at the intersection of ‘importance‘ and ‘breadth‘.

3: What’s the cost?

This is linked to the next two questions but I like to ask it first. It’s an assessment of the change – positive or negative – to the data asset. How many domains are involved? Does it change the quality rules?  Does it add and/or devalue any metrics? Does it affect overall quality for rolled up entities? Is it architecturally significant?  is it legal? Does it break any GDPR/FOI/External reporting/etc? Does it change definitions? Does it modify/add/delete any business intelligence or related reporting artefacts? Does it require mastering/change of interfaces/etc? I save this question until after 1 and 2, as it has often not been considered, so suddenly this  ‘minor change‘ looks rather more substantive.

4: What are the options

Are there any? For a small change no may be an appropriate answer. For bigger initiatives, the work needs to have been done to consider if a short term fix might be the right solution. Or can it be solved by an existing project on a slower timeline.  If the answer is ‘we don’t know‘, this suggests a lack of rigour in the analysis.

5: What else does it affect?

What’s the bigger picture? Has this solution been properly bottomed out? What are the consequences for technology, training, etc. This is not a specific data impact question, but it is another lens on the analysis of the problem.  With proper change/architectural governance, data should just be part of this wider assessment. Such an approach is not that common in HE.

6: What happens if we do nothing

Always a good one to finish on. This is essentially the ‘we can’t do everything, but we can do something‘ maxim. The questions needs to be asked if this the right something?

This is not an exhaustive list. Every organisation is different, but any impact assesment must consider any change through a number of lenses. We’re trying to derive a sense of the impact now, not later on in the project after  significant delivery costs are baked in.

This is not easy, but it is an essential tenet of professionaling our approach to data.  If you’d like to share your own questions, or would like to learn more about how this works in the real world, please do get in touch.