Hacker News

When targets and metrics are bad for business(thehustle.co)

59 pointstbeutel posted 6 months ago6 Comments
Jtsummers said 6 months ago:

For more on this I recommend both The Goal and Out of the Crisis (the latter more, but the former also discusses this).

Targets are often useful because they're proxies for the unmeasurable. But, as discussed in the article, they can easily become goals themselves, which take you away from the actual intended business of the day.

In The Goal, the company has a target of 100% productivity of all equipment/work stations. The thought being, if it's not producing then that's lost time. However, it increases costs (overall) because you end up with excess inventory leading to the constraint or bottleneck of the system. That inventory has no value (it's not sold or sellable) and has costs (storage, at a minimum; possibly maintenance) and can become stale (becoming worthless, at which point with the carrying cost it has negative value). EDIT: Another problem with that target is that it penalizes those at the end of the production chain, who can't maintain high productivity (they're after the bottleneck). Their production rates are limited to whatever the bottleneck can produce. Ever been in the IV&V team? "We're late because of you...", no you're late because we received the product one month before it was supposed to ship.

Out of the Crisis also discusses another problem with targets: If it was just a matter of setting the target, why were you missing it in the first place? There's a larger system the employees fit in which constrains their performance. Address that system, or the target will be unattainable or unsustainable.

gav said 6 months ago:

> Targets are often useful because they're proxies for the unmeasurable

Not just the unmeasurable, there's a lot of KPIs used because they are easier than figuring out metrics that are meaningful.

Here's some examples I've seen:

Number of Visitors: a large manufacturer with no direct channel considered the number of visitors to their website as the most important thing and this was reported to the highest levels. Without some way of determining the quality of the visitors, or if their experience was valuable, the actual amount is useless.

Engagement: a bank was tracking how many visits and how minutes customers spent on their website every month, they considered more time a good thing. Personally I want to spend the minimum possible time dealing with my bank, preferably zero interaction, making processes more streamlined rather than having more content is key.

Bad metrics allow you to report up the chain that you are moving the needle in some way and justify budgets.

Jtsummers said 6 months ago:

I believe this was meant as a reply to my comment, and I agree with you entirely.

Often the easiest metrics are the least useful. I worked in an office that wanted "production hours" (the business was largely labor-oriented, where that was actually a semi-useful metric, but it's not a useful metric in software development). It was gamed, as long as people charged directly to a project they were "productive". The costs of training and all were hidden as much as possible so projects and teams could seem productive. This meant that customers were being billed hundreds of dollars an hour for training that may not have even related to their project (bordering on fraud here, certainly abuse of some kind). But it also meant that the training budget (organizational) was slowly being cut, because no one was using it.

Selecting your KPIs carefully may not prevent failure, but choosing them poorly can almost guarantee failure or mediocrity.

forgingahead said 6 months ago:

Scott Adams has written a fair amount on Systems instead of Goals:


rahimnathwani said 6 months ago:

If you're interested in setting good targets and metrics, I highly recommend these books on OKRs:

Measure what matters, Doerr: The newest book on OKRs, by the guy who introducted OKRs to Google

Move, Patty Azzarello: A really really great book on executing strategic change, that has a section about setting 'Control Points', which are roughly equivalent to what Intel/Google etc. refer to as 'Key Results'

Radical Focus: A light read about OKRs. Good as a first book on OKRs, but not as detailed as the two books above.

Introduction to OKRs: A very short book on OKRs. Good as a first book on OKRs if you have a short attention span.

Jtsummers said 6 months ago:

Measure What Matters: https://www.amazon.com/dp/B078FZ9SYB/ (also on Safari Books Online)

Move: https://www.amazon.com/dp/B01N9ZVTGG/

Radical Focus: https://www.amazon.com/Radical-Focus-Achieving-Important-Obj...

Introduction to OKRs: https://www.oreilly.com/business/free/files/introduction-to-... (I'm assuming this is the one, came up in Google and you didn't provide an author).

I'm not sure if Move or Radical Focus are on Safari Books Online, but they're not on my company's subscription to it. Thanks for the reads, I'm pulling up the 4th one to read because I know my management doesn't have the attention span to read anything longer than 37 pages.