What causes a strategy to change?

Hi again,

Continuing our work on automation strategy, we need the community’s help one again :smiley:

This time, we want to hear your thoughts on the following question:

Why would you need to update a strategy? What changes in a context would cause you to make updates in a strategy? How would you mitigate these changes?

We want to demonstrate that there are many reasons why we might need to update a strategy. Perhaps the product vision has changed, we’ve learnt new information about our users, etc. We’d like to share examples with learners to help them better understand what to look out for when maintaining a strategy.

We look forward to the responses and really do appreciate your contribution :robot:

1 Like

The big headings that come to mind would be a change in:

  • Explicit needs and desires for the product or our knowledge or understanding of them
  • Tacit needs and desires for the product or our knowledge or understanding of them
  • The product itself
  • Our understanding of risk
  • The resources available to us
2 Likes

changes to a “test automation strategy”, I reckon from context. We have strategies on many levels…

A strategy (ie a decision to leverage resources) could be to “automate everything that can be automated” (@angryweasel) might be detailed so that those user stories with seemingly closed (as compared to open) articulation are automated. Closed example: given the user clicks the button, the dashboard is shown. Open example: The dashboard gives a good overview of the case. We learn that there is a grey zone, that given unlimited resources we could probably make an algorithm to determine what a good overview is. Thus these (few) things are probably better tested by a person.

Another automation strategy could be more based on the classical test pyramid (Fowler 2012). It might turn out that the business domain we are tackling has so many rules, edge cases, and implicit know-how that it’s best resolved by heavy testing by subject matter experts.

Often, practically speaking, a strategy “changes” by being introduced mid-process, or by being re-written or evolving because someone new has joined the team and they have a different way of thinking. Sometimes the strategy is updated as the team get a new understanding of the importance of something and want to see it at strategic level (e.g. accessibility testing having more of a legal underpinning) knowledge which can come from attending a conference, training, meetup or simply a chat between coworkers.

My initial thoughts on adjusting strategy:

I see the strategy, whether at the team level or at the corporate level, as consisting of three layers:

  • goal: what are you doing it for?
  • vision: what needs to be in place to achieve the goal?
  • plan: what actions will you perform to implement the vision and in what order?

The goal should be in business terms, because:

  1. You can then align with the business so that you know how it can offer the most value.
  2. You can make decisions that will target that value effectively.

The goal can change because:

  • The old goal has been achieved
  • Business decided on a new course and it calls for another goal
  • The scope of the automation effort changes (other systems, interfaces, teams, test stages, …)
  • New insight made it clear that another goal would offer more value

The vision can change because:

  • The vision has been achieved
  • The goal changed
  • New insight made it clear that some aspect should be changed:
    • It is not necessary after all
    • It is necessary after all
    • It is not (technically / financially / …) possible or not allowed after all
    • It does not combine with another aspect
    • It was overlooked earlier
    • Its importance was overestimated / underestimated earlier
    • Staffing changes (people added, removed, or switched out)

The plan can change for all the usual reasons, including availability of all kinds of resources, changing priorities (here or elsewhere), etc.

Of course, there are all sorts of examples for all of these , but adding these would make this rather long …