Goodhart's Law: An obsession with the numbers can sink your strategy.

Goodhart’s Law: when a measure becomes a target, it ceases to be a good measure.

Our observations from having reviewed hundreds of "less successful" OKR implementations is that most common reason for weak OKR results is people's tendency to confuse the achieving the Key Results with achieving the Objectives. Instead of chasing the Objectives, they chase the Key Results.

That's what we call "surrogation" - confusing what's being measured (the 'O') with the metric being used ( the KRs).

We have seen surrogation in at least 70% of the failed OKR solutions that we have reviewed. The good news is that it is preventable with good OKR design, and a recoverable error if you did not avoid surrogation in your initial launch - so long as you catch it early enough (unlike my friend).

The most cited example of this law in effect is the case of nail factories in the Soviet Union. The goal of central planners was to measure performance of the factories, so factory operators were given targets around the number of nails produced. To meet and exceed the targets, factory operators produced millions of tiny, useless nails. When targets were switched to the total weight of nails produced, operators instead produced several enormous, heavy and useless nails.


Cases closed are both the measure of performance and the target of customer service organizations, and employees might choose to close cases without fully investigating and resolving them. When page views are both the measure and the goal of news sites and blogs, editors are tempted to post shocking and controversial content. Over the long run, of course, this behavior degrades the quality of the site, and the page views measure is no longer a useful indicator of an engaged user base.


Here are five simple pointers to avoid surrogation and combat Goodhart’s Law:

1. Create Objectives with Concrete Outcomes

Objectives should be stated as outcomes, not outputs. For example, we often see organizations setting an Objective around quality, such as "Manufacture within specifications" (or in the service sector "Close Severity 1 trouble tickets quickly"). This is an output, not an outcome. Surrogation happens with "output" Objectives because they are completely focused on the deliverable, not the benefits for the organization or customer. With an Objective like this, the first thing we do is look at the KR to determine the specific attributes of quality - thing like: "# parts out of specifications", "# Failures in the Field" etc.

To fix these 'output' Objectives, ask yourself "why" (five times) on why we want to "Deliver Quality Solutions" -

Why? "So that they have fewer problems with our product."

Why? " So that they get better value for their purchase"

Why? "So that they make repeat purchases and become advocates for our company."

Now you can write your Objective "Deliver products that drive repeat purchases and build customer advocates". This 'outcome' Objective would have significantly different KRs than the original "Deliver quality solutions". Selected KRs might be things like "% Revenue from Repeat Sales", "# Repeat Customers", "# Favorable Social Media Mentions", etc.

This 'outcome' approach does a few great things for your company, all at once:

Reduces the likelihood of surrogation by removing the focus on the numbers.

Creates the space for creativity on part of your sales team - there are many more ways to generate repeat sales and many other ways to encourage customer advocates.

2. Measure What Matters, Not What is Easy

Your KRs need to address the desired outcome, not things that are "close", but miss the target!

Whoever said "be careful what you ask for, you might just get it" was talking about KRs.

Find Key Results that are exactly what the Objective is striving for. When surrogation happens with KRs that are directionally correct, but not precisely correct, the organization quickly heads in the wrong direction.

3. Not All KRs Were Created Equal

Some KRs are far more important than others. For our above example, if leadership assigned an 80% weighting to "% Revenue from Repeat Sales", and 10% for each of "# Repeat Customers", and "# Favorable Social Media Mentions" they would encourage the Sales Team to increase repeating revenue (maybe from a few customers).

Whereas a weighting of 10% weighting to "% Revenue from Repeat Sales", and 80% for "# Repeat Customers" would drive a different result - gaining a broad selection of repeat customers.

This is also what we call "the fair deal". The fair deal occurs when management commits in advance to where one's priorities should be. (The un-fair deal is when management does not set priorities and then later criticizes people (with the benefit of hindsight) for not knowing where they should have been focusing.)

4. Focus on the Objectives, Not the Key Results

It is very important that the topic of every performance meeting is the Objective, not the Key Result. Not only does this align the teams towards the right outcomes, but it also releases a lot of creativity as people think about ways to achieve the outcome, not the KR.

5. Loosen the Link Between OKRs and Rewards

Notice I say "loosen", not "remove". In my mind, it would be crazy to set OKRs based on what people should be doing, but not include those OKRs in your reward conversations. The trick is to make the OKRs only a component of those conversations. I have written about the "Six Stages of OKR Meeting Magic" (Valuation, Navigation, Compensation, Calibration, Communication and Regulation). OKRs are a "navigational" tool - they help you navigate the business (do we turn left or right?). These KRs tend to be leading indicators.

Compensation metrics tend to be lagging (and therefore not useful as KRs), must be auditable (as people need to know they are rewarded based on 'hard' numbers), and have to fit the S.M.A.R.T. framework (which goes against the whole 'aspirational' / stretch goal idea).

In conclusion

Surrogation is just one of the top five reasons that OKR solutions fail - but they are an important one because many people do not see this problem occurring until it is too late.

The happy ending for our clients is that by including these points in your OKR design you can quickly get your team back on track.

Brett Knowles

Brett Knowles is a thought leader in the Strategy Execution space for high-tech organizations. His client work has been published in Harvard Business Review, Forbes, Fortune, and many other business publications.

Latest Posts


Looking at management "best practices" as connected lines


The post explains how OKRs help organizations adapt to changing business environments and align strategy, budget, and performance.