GQM’s approach is to ask wide-ranging questions about a Goal. This makes it more holistic and complete than OKRs or at least, a way to select higher impact Key Results.
On a recent product requirements document, I laid out measures of success using an old favorite: Goal-Question-Metric (GQM). The result was a very clear definition of success for the product and also a roadmap to measure progress.
GQM goes beyond even this, to offer either a comprehensive alternative to Objectives and Key Results (OKRs) or at least a starting point.
An OKR’s Objective describes what you want to achieve, while its Key Results describe how you’ll achieve it. The What Matters site, as the source for OKR understanding, lists three gut checks for a good Objective.
- Is it meaningful? Is the Objective a top priority? Does it articulate a clear direction?
- Is it audacious? Is the outcome taken for granted or does it take what you do every day to the next level? Does it represent a significant change from where we are today?
- Is it inspiring? Is the Objective easy to remember? Does it empower your teams?
Along with common characteristics of good Key Results:
- Are they specific and time-bound? Is it plainly stated what needs to occur and by when?
- Are they aggressive, yet realistic? Are they aspirational, yet not so outrageous that they could never be accomplished?
- Are they measurable and verifiable? Is it clear when the criteria for success will be met?
This “Minnesota Innovation” example is from What Matters, with Key Results presumably derived from relevant knowledge on satisfying the Objective.
And here is another example, “Improve company website”:
The OKR Objective is GQM’s Goal. The “meaningful, audacious, and inspiring” criteria of an OKR Objective can also be applied to a GQM Goal. From here, the two methods diverge — while OKR proceeds to Key Results, GQM has you ask questions.
Why is asking questions so worthwhile? Look at the examples above and notice how their Key Results are basically arbitrary. Presumably some research was done to justify them, but it isn’t obvious. By contrast, GQM questions are intended to be a fairly complete inquiry into the Goal. Because of their completeness, GQM’s questions help you stay focused (pun intended) on what matters.
Consider the “Minnesota Innovation” example from above, now rephrased using GQM, with questions replacing key results*. We try to ask every relevant question about the Goal that we can.
And here is a GQM version of “Improve company website”*:
Each of these questions should be constructed to accommodate a numeric or Yes/No value. As with OKR’s Key Results, GQM’s Questions must be converted into feasible targets. This can be done by taking baseline measurements for each question, which you will then seek to improve.
Let’s first summarize both methods.
Define an objective and choose a few key results that if satisfied, indicate you got there.
The Key Results of OKRs aren’t all the results that ever could be achieved for a given objective, but instead some key results in support of that objective. An objective could remain over multiple periods, with its key results either changed or their associated target metrics updated.
Here is the OKR for “Improve Company website”, shown graphically.
Define goals and measure how well you meet them.
GQM Questions aim to comprehensively measure the Goal. Your task is to initiate actions that improve metrics associated with specific questions or if you’re lucky, certain actions might improve numerous metrics.
Unlike OKRs, having a broad and multifaceted set of questions encourages you to find actions that have maximum impact; those actions can then be the basis for OKRs. For example, with Minnesota Innovation, these two questions…
- What is the number of startups that are founded or relocated to Minnesota per year, and how does it compare to other states?
- What is the amount of venture capital funding raised by Minnesota startups per year, and how does it compare to other states?
…could be folded into an OKR Key Result such as “Attract venture capitalists to Minnesota startups” because achieving this key result would probably impact metrics associated with both questions.
Below is the GQM for “Improve Company website”, shown graphically. The shaded areas represent progress over time towards questions. GQM’s multiplicity of questions for any Goal permit a more fine-grained understanding of progress towards it.
Let’s revisit the Key Results of the “Minnesota Innovation” OKR example. Why is it essential to place Minnesota on the State Technology and Science index? Why do we aim to increase the number of innovative businesses, etc. by 25%? Why do we want to double technology training for job seekers? OKR’s Key Results omit many crucial questions that could have been asked, perhaps even critical aspects of the objective.
The contrast between OKRs and GQM is even more pronounced when applied to the “Improve Company Website” example. By considering all feasible questions rather than selecting a few key results, we generate a roadmap rather than a shorter-term approach.
In attempting to fully characterize the Goal via questions, GQM pursues a more holistic approach than OKRs. But they’re also not incompatible as GQM can be used to establish a long term roadmap that is then delivered through OKRs:
- Find the Questions/Metrics associated with a Goal.
- Determine actions that most effectively improve the metrics.
- Establish these actions as your OKR Key Results; the Goal can be used as the Objective.
If after reading this essay, you still aren’t planning to even bootstrap OKRs with GQM, consider this alternative: GQM excels where OKRs aren’t specifically intended — in comprehensively assessing the metrics required to measure the operational success of a product.