Senior Design Systems Designer โ€ข Manchester, UK ๐Ÿ

OKRs are a goal-setting system used by some of the world's most successful tech giants like Google, Twitter and Amazon. Intel attributes its market dominance to it, if you read the history. The UK government has set out five key tests that have to be met in order to have confidence in lifting the coronavirus lockdown that are eerily compatible with the OKR framework. I believe we can take it one step further and frame the five conditions as true OKRs, to demonstrate that OKRs are highly flexible and actionable for any business.

OKR stands for Objectives and Key Results. If you don't know much about OKRs, I recommend reading Wikipedia or buying the book Measure What Matters.

The Tests

First, we need to start with the original tests. I pulled these from an article published on the Telegraph:

  1. The NHS must be able to cope with the healthcare demand
  2. A consistent fall in death rates
  3. Infections decrease to a manageable level
  4. Enough PPE exists for future demand
  5. Confidence that any changes do not incur a second peak

Setting the Objective

A great set of OKRs starts with a well defined objective. It can be fluffy, but it must be inspirational, encourage ownership, and set a clear direction. This is often the hardest part as many people aren't required to set clear goals or objectives until they're in senior positions in their career. Even then, they never get the right feedback to craft exceptional ones. Here we're tasked with lifting lockdown in essence, but I think we can reword that to something a bit more dreamy.

Contain the coronavirus so we can safely reconnect with our friends, family and loved ones in person

That's what we all want right now. Or an economy-centric one might be:

Give businesses the power to safely return employees, amid coronavirus, to work to provide for their families

Both don't clearly define exactly how we'll do it, but they describe the end state that we want. In reality either of these are perfectly fine ways to describe the same goal, but they're aimed at resonating with different people. I'll leave it to you to decide what resonates with you more, but for this article I'll be using the former.

Next up, let's create some key results that can indicate whether we're getting closer to this end state or not.

Turning the tests into Key Results

A well defined key result is one which is measurable and can clearly be seen as achieved or not achieved. If I say I want to "get better at dancing" it's not clear what better is, and doesn't have a time frame. Instead I might suggest "place in the top 20 of the October 2020 regional salsa dance heats". Suddenly, I have a time frame and a measure against which I can clearly say whether I did or did not achieve it. That's the not-so-secret secret.

Here's how the five key measures could be reframed as clear key results.

The NHS must be able to cope with the healthcare demand

I'm slightly cheating here, since the inspiration to write this blog piece is from a news article I read where we can measure the ability of the NHS to cope by the number of vacant beds suitable for coronavirus patients. I'm making an assumption that that's an adequate measure since there's many variables at play. I'll take 85% utilisation as an appropriate number of beds to illustrate, but I'm sure an expert could weigh in for me the real number required.

If the recovery rate for those needing hospitalisation is up to 42 days, and the incubation period is 2 weeks, then we can take a rolling average and aim for the influx of cases to not exceed our pre-determined maximum capacity of 85% of hospital beds for the duration of their stay. If there are 100 beds and 70 people are admitted, then over the next 42 days we need to have 15 or fewer new cases before those original 70 can be discharged. I'm oversimplifying, but it presents a working model for this test, but leave 15% extra capacity to account for risk.

Therefore, we could phrase a key result as such:

Hospital bed capacity must not go above 85% capacity consistently for a month, using a two-week rolling average.

We may exceed that capacity as rules are relaxed one by one, but here we clearly have a metric to aim for and set as a benchmark or baseline.

A consistent fall in death rates

Okay, this is much closer to a well-defined key result but it lacks concreteness. If we go from 20,000, to 19,999 the next day, does that constitute a consistent fall? I'm sure many of you would disagree, but we need to be explicit what we mean in order to decide whether we've achieved this or not. With a little bit of tweaking and a clear exit condition we can more like:

Death rates must consistently fall by 5% week-to-week for at least three consecutive weeks.

This accounts for any variable number of deaths. Taking the original 20,000 death as a weekly death toll example if it becomes roughly 17,000 weekly deaths after three weeks then we have successfully achieved this key result.

Infections decrease to a manageable level

One of the initiatives that the government seem to be pushing for in recent days comes down to our capacity to test for coronavirus. This touches upon one or two other key results, but I'd like to use it here to reverse engineer into a key result. It infers that we're not at a manageable level too, which is misleading. It seems that at the time of writing the general consensus is we've passed the peak of infection rate, based on the NHS's ability to cope. As long as the rate of infection is below a certain threshold then we can calculate what our testing capacity needs to be.

I'll make a few assumptions using incorrect numbers, but hopefully it's enough to illustrate. If we want to re-mobilise 100,000 people back into the public per day, and account for false negatives on tests, we could phrase a key result as:

Achieve the capability to test 110,000 people per day at 99% accuracy

Even though 1% of 100,000 people would only be 1,000 extra tests needed, it gives us some spare capacity for retesting people, emergency tests if required and incoming people to the U.K. who may be returning to the place they call home. With this as a basis, you can calculate the infection rate whilst accounting for test error and it simply becomes business as usual to monitor and adjust the guidelines as necessary.

Enough PPE exists for future demand

Great stuff, this is a little easier to articulate as it's a case of defining what the future demand will be and plotting that against our ability to manufacture or obtain PPE. For the uninitiated, that's Personal Protective Equipmentโ€”face masks and gloves as common examples. Supplying the NHS is the first number to identify, together with the number of people who would be out in public using PPE. A staged lifting of lockdown would help identify the rate at which we need to acquire and use or store PPE in order to account for any risks or second peaks. In a nutshell we have a logic condition:

PPE acquisition rate > PPE usage rate

Add in some sensible numbers and we might get something like:

Our rate of acquiring PPE is 5% greater than the rate of PPE usage, week on week for three weeks.

With this on the go, you'd be able to more confidently lift lockdown measures bit by bit in order for there to be enough PPE equipment to accumulate for each "lift stage". During each stage you'd need to assess the rate of PPE usage and adjust accordingly to have enough in storage. Three weeks is an arbitrary number to have some confidence and a clear exit condition.

Confidence that any changes do not incur a second peak

This might be the toughest one to turn into a key result, since you're almost asking to predict the future. In fact, I'd probably turn this into a separate goal but in the interests of simplicity we'll do our best to reframe as a key result. What you can do to achieve this effect is to assess the risk of any change, keep it below a certain threshold, and enforce a rollback condition should infections rise too quickly.

We can determine with statistical certainty the risk factor of introducing any new changes to lockdown procedures, and determine a rollback procedure for each change.

You might set an initiative to assemble a team of experts that could do this, but the key result is that we have the capability to determine risk factors. Taking this to an extreme, if we had a crystal ball that could tell us this then we've achieve this key result. With this ability, we can have the confidence to navigate forward with a reduced level of risk to trigger a second peak of coronavirus infections. We may not be able to proceed with haste, but that's probably one for a future set of key results once these have been achieved.

The new UK Coronavirus Key Results

Now this is more like it. When we put it all back together we then have a set of fresh OKRs that look something like this.

Objective: Contain the coronavirus so we can safely reconnect with our friends, family and loved ones in person.

Key results

  1. Hospital bed capacity must not go above 85% capacity consistently for a month, using a two-week rolling average
  2. Death rates must consistently fall by 5% week-to-week for at least three consecutive weeks
  3. Achieve the capability to test 110,000 people per day at 99% accuracy
  4. Our rate of acquiring PPE is 5% greater than the rate of PPE usage, week on week for three weeks
  5. We can determine with statistical certainty the risk factor of introducing any new changes to lockdown procedures, and determine a rollback procedure for each change

What's important to note is that although we can achieve many of these key results, it can just as easily bounce back. The intention is that by striving for and working towards these milestones we put the necessary pieces and processes in place for us to sufficiently manage coronavirus and get one step closer to the objective of reconnecting with our friends, family and loved ones.

Next steps

There are two key steps involved to take these forward. Firstly is that you use the notion of initiatives to drive forward your key results. For example, an initiative might be "acquire three PPE production companies". Not the greatest example, but by doing so we are putting the pieces in place to achieve key result number 4 of acquiring PPE faster than we use it.

Once you've taken on a number of initiatives and worked towards these key results for a fixed amount of time or achieved them, whichever comes first, then you need to review and revise your OKRs. Either you smashed them and didn't set yourself hard enough goals, or you achieved none of them and weren't realistic. Or worst case scenario you achieved most of them but they didn't help the objective. Whatever the outcome, it's important to be honest, transparent, and iterate.

If you want to see your business thrive in these tough times, framing relevant OKRs may be right for you.

Help yourself

I'm a big fan of self-development and learning, so I do encourage you to pick up the book and give it a go.

Disclaimer: I may make a small commission if you go ahead and purchase the book using this link. It helps support me giving you more free knowledge.


Liked this? Forward to a friend, subscribe to my newsletter, Medium, YouTube, or become a Medium member to support my writing (and other writers, too!).
Youโ€™ve successfully subscribed to Chuck Rice
Welcome back! Youโ€™ve successfully signed in.
Great! Youโ€™ve successfully signed up.
Your link has expired
Success! Check your email for magic link to sign-in.