Should we use OKRs for everything?

One of the most common questions I get about Objectives and Key Results is, “Should we use OKRs for everything?” If you read the most popular books on this topic you get a fairly consistent point of view: No, you shouldn’t use OKRs for everything, They are for the most important things you are trying to optimize in the immediate future.   

This question comes into even sharper focus when you you compare the kind of big, innovative goals many companies put into their OKRs with what’s often called Business As Usual metrics or BAU (in a recent short chat, Christina Wodtke, author of Radical Focus, told me that she prefers the term “heartbeat metrics” to BAU as it gives a greater sense of value to what are often day-to-day measures of the operations of the business). 

In fact, Christina offers a clear description of how to differentiate between OKRs and heartbeat or BAU metrics in your planning process. She visualizes it in this four quadrant approach:

Source: Christina Wodtke, One Objective to Rule Them All

Despite the general opinion being to separate “big” goals from “daily” efforts I remain unconvinced.

Prioritizing goals

The first issue that comes to mind for me is deciding which goals a team should prioritize each day if they’re managing two sets of success metrics. For example, let’s take a customer acquisition and onboarding team at a B2C company. If they were asked to divide their goals into OKRs and BAU metrics it could look something like this:

Objective: Create the most seamless onboarding experience in the online job search industry by Q2 022

KR: 90% of new users complete onboarding on the first try

KR: Reduce landing page bounce rate by 45%

While their heartbeat metrics could look like this:

Marketing email deliverability at 95% or higher

Landing pages show most current offer 100% of the time

Which of these do they work on when they come to work each day? How do they choose? What is the relationship between these two sets of goals?

You might try to divide the team into two sub-teams — one focused on the OKRs and one focused on the BAU metrics. That would work for a while until the effort of one of those teams optimizes locally for their goals while negatively impacting the other team’s goals. In our example above this could happen as the operations team pushes to ensure email delivery is consistent even though the OKR team is noticing increased churn due to “too much email” from the company negatively impacting their own goals. 

The BAU team isn’t doing anything wrong. They’re optimizing for their goals. It’s just that their goals have stopped being in alignment with their colleagues. 

In my opinion, the BAU goals in this example are features of the onboarding process. They are ways for the entire team to achieve their OKR goals and should be treated as such to reduce confusion and inter-team competition as well as hyperlocal optimization. 

What if I don’t do “innovation” work?

There are many folks inside an organization who don’t do customer-facing innovation work. In fact, a company should be clear on each team’s mission. And some of those teams will have to continue to do operations, maintenance and customer service work. If that’s the world you work in, should you still work towards OKRs? 

Maintaining customer service metrics like call duration, number of calls serviced per day, first time call resolution, customer satisfaction with the CS team are all strong candidates for BAU metrics. Teams don’t necessarily work to “innovate” on these metrics but rather to maintain them at a stable high level. The same could be said for operations or security teams. Do OKR’s make sense in these situations? 

I take a general point of view that everyone makes a “product.” As long as there is a human consuming the thing you make at some point then you’re creating a product, you have “users” and your work impacts those users’ behavior. Expanding this perspective, it is then your responsibility to ensure, not just a consistent level of performance but one that is continuously improving. Those improvements might be small, incremental and impact your service slowly over time. 

Heartbeat metrics seem to make sense in these cases since big leaps in innovation are rare and the overall goal is stability rather than exponential growth. How might these teams then react to their goals being everyday “business as usual” goals rather than the innovation goals the company craves and celebrates?

OKRs in non-technical disciplines

Should HR, Finance and Legal have OKRs too? If you agree with my perspective above that everyone makes a “product” of some kind then the short answer here is yes. HR makes policies and programs. Finance makes budgets and plans. Legal also creates policies and guidelines. These are designed to elicit behavior from their target audience — company staff. The goal is not to ship the policy, program, guideline or budget but to change staff behavior. In these cases OKRs make clear sense as a goal-setting tool. 

Tracking multiple goal systems and the dependencies between them

This is where the rubber, as they say, hits the road. How do we keep track of different types of goals and the dependencies between them? In addition, how do we reconcile the cultural impact of some people working on “big innovation” initiatives and others on seemingly mundane day-to-day tasks? 

If a BAU metric conflicts with a team’s effort to achieve an OKR, how do we choose which one to optimize? 

The main argument for OKRs to be used solely for big things you’d like to focus on in the coming cycle is prioritization. I agree that OKRs are a powerful prioritization tool. However if we continue to maintain other ways to measure success and reward our staff we risk diluting that power and the impact of OKRs.

While the majority of OKR canon advocates maintaining a separation between these two types of efforts I remain unconvinced that this is somehow a clearer, less confusing way to work. 

What do you think? What’s been your experience?