full
Understanding the Difference Between OKRs and KPIs
Learn about OKRs, KPIs, how they are similar, and how they are different.
Transcript
I did write, like a page of stuff on the difference between
KJ:OKRs and KPIs yesterday. Okay, if I know that I don't know how
KJ:that would fit in with the customer lifecycle. Somewhere,
KJ:but I just, it's in my head. So if you need me to riff for 10
KJ:minutes on that sort of stuff, I can do that.
Stephen N.:The difference OKRs and KPIs.
KJ:Yeah. Or is there other ones that you're looking at?
Stephen N.:Well, the only thing I was thinking of that is like
Stephen N.:we can we can riff on that. I think the benefit would be. And
Stephen N.:we talked about incorporating this into key results, but the
Stephen N.:difference between committed goals, stretch goals, and
Stephen N.:moonshot goals, you know?
KJ:Yes. I think that, that lies within that topic, our race. I
KJ:mean, because I think I looked at what are the similarities
KJ:with OKRs, and KPIs, but the difference is primarily being
KJ:the intent of those measurable like, they're both quantifiable
KJ:measurements and frameworks to measure performance. But a KPI
KJ:is often it's taken commonly across businesses are quite
KJ:common, and they're quite, they're not as adaptable. You
KJ:know, and they're, they're always committed. And they come
KJ:from a place where the HR people set them to determine
KJ:compensation. So they have this emotional attachment because
KJ:they're connected with money, whether an OKR, as you say,
KJ:could be a moonshot where you're not connecting someone's comp,
KJ:to heading 100% of the moonshot, you're simply setting it to
KJ:encourage one's ambition.
Stephen N.:Right, which is where I think having that
Stephen N.:distinction you can still use, I believe you can still use KPIs
Stephen N.:within OKRs. There's a school of thought that sense to keep those
Stephen N.:separate separation of church and state, right. Yeah. But the
Stephen N.:people will always gravitate towards what's going to be
Stephen N.:incentive incentivize for them. Yeah. What will be your typical
Stephen N.:day to day KPIs? revenue generation, customer
Stephen N.:satisfaction. I think that's where the like a committed key
Stephen N.:result can be basically a KPI in disguise, versus a stretch. key
Stephen N.:result is something that pushes the limitations versus the moon
Stephen N.:shot. That's, that's my opinion of how those three should be
Stephen N.:categorized. But yeah, the KPI the that's, that's a big problem
Stephen N.:with a lot of these implementations, because they
Stephen N.:just wild Why would I buy a do I care about some stretch? It OKR
Stephen N.:I'm trying to hit my number.
KJ:Yeah, I dealt with it man. Like I lead the our team of
KJ:sales engineers, and we got given mandated a KPI from our
KJ:Chief Revenue Officer down. And then I wanted to try out OKRs
KJ:with the team, because there was six of us, and I thought I'd
KJ:give it a go. And people really found it difficult to say, I was
KJ:pitching them like OKR is this great framework where you can
KJ:prioritize what you need to focus on. It's like, it would
KJ:KJ, here's our KPIs. And here are OKRs. Now, which one do we
KJ:do? Like, do we do the one that gets us paid the bonus? Or do we
KJ:do the one that you're telling us to do? And ultimately, they
KJ:all chose the bonus. So they wanted to get their bounce?
KJ:Right? So it's really difficult to ask people and you undermine
KJ:your argument where OKRs will help you focus? Well, they
KJ:won't, if you have this other force called KPI is pulling you
KJ:in a different direction. So you have to have a reconciliation
KJ:between bolts, but what I was going to say about KPIs is that
KJ:I read this great good hearts law statement, which is any
KJ:observed statistical regularity will tend to collapse once
KJ:pressure is placed upon it for purposes of control. So in
KJ:essence, if you take a an observable measurement, like you
KJ:know, revenue or some sort of KPI and you place A lot of
KJ:pressure on it, for the purposes of micromanaging or controlling
KJ:someone underneath you, it immediately collapses. So that's
KJ:what KPIs I feel tend to be for the purposes of command and
KJ:control. Whether if you perceive OKR isn't that way, it's not
KJ:going to work, you have to use OKRs to empower people from that
KJ:should be autonomy of the measurement. And the
KJ:quantifiable measurement should be given to the person creating
KJ:the OKR. It shouldn't be your boss telling you to hit 100%,
KJ:that should be them deciding what they want to hit. It's a
KJ:different different sense, or a different intent. It's not for
KJ:commanding you and controlling you. It's for comparing.
Stephen N.:Yeah, yeah, no, that makes sense. And the other piece
Stephen N.:in terms of like, how do you reconcile the two is really, at
Stephen N.:the end of the day, like a KPI is a lagging indicator. Yeah.
Stephen N.:And key results. OKRs can be a powerful leading indicator, can
Stephen N.:we execute just on these simple things? Can we to use a
Stephen N.:marketing example? Can we improve our conversion rate from
Stephen N.:four to 8%? That's a, that's a leading indicator, because you
Stephen N.:can go off and you can execute any number of projects and
Stephen N.:initiatives to see if you are moving the needle on that. Now,
Stephen N.:your ultimate KPI might be the number of leads that you
Stephen N.:generated. Yeah, right. And that's that number is what it
Stephen N.:is. But before you get to that number, this is a great bridge
Stephen N.:to get there. And it's incremental, too. It's like,
Stephen N.:it's bite sized chunks if you're not trying to, like boil the
Stephen N.:ocean. Like, what are the you know, our ultimate objective
Stephen N.:here is we want to have a full pipeline. Okay, great. Are we
Stephen N.:want to hit these revenue numbers? That's fine. The to get
Stephen N.:there, it's gonna take a series of smaller steps to get to the
Stephen N.:bigger picture. Yeah, that's how those two are combined.
KJ:Yes. Yeah, you're right. And combined, you get a more
KJ:holistic view of performance, you know, to like, if you want
KJ:to really evaluate someone's performance at their job. Don't
KJ:just take the lagging indicator. That's what most people do.
KJ:They'll obsess over the KPI like, did this sales man hit
KJ:this KPI quota? And if you did not, that's, then you're out
KJ:whether you have to take a more holistic view of this person's
KJ:contribution, not just the KPI. But how did they contribute to
KJ:the OKRs, which, as you rightly said, facilitate and foster
KJ:communication about how we're going to get to that KPI. They
KJ:facilitate collaboration with other people how to get there,
KJ:they facilitate experimentation. Because you set an OKR. And you
KJ:experiment and you say, maybe our hypothesis is, if we put a
KJ:cart button here, we might drive behavioral change to increase
KJ:conversion rates on the cart. Let's try it out. Let's develop
KJ:it. Let's do it. Let's design let's put it in, did it work?
KJ:Okay, it didn't, okay, great. Well, we'll just try something
KJ:else. OKR is allow for that. Or the KPI is the command or going,
KJ:you have to hit 100. The OKR is the one that kind of allows you
KJ:to be agile to get there. Yeah.
Stephen N.:And the beauty of all this is when you look at it
Stephen N.:all through the lens of the logic model. If you are
Stephen N.:empowered to select your own OKR, select your key results,
Stephen N.:the things that you believe will move the needle, that's great.
Stephen N.:You still need to pick the right activities, and you can't do
Stephen N.:everything you get you have to be very selective. And it's it's
Stephen N.:a game of prior prioritization really, like what are the top
Stephen N.:three or four things that I can actually tangibly work on today,
Stephen N.:this week, this month, this quarter, whatever, to move the
Stephen N.:needle, but even going before that is that's all well and
Stephen N.:good. But you're only one individual or small team, do you
Stephen N.:have the resources that you need to hit that to do the activities
Stephen N.:that you desire to drive the impact on the key results to
Stephen N.:ultimately lead to your KPI? So if we, you know, as a little
Stephen N.:startup said, Hey, guys, we're gonna, our revenue goal for 2022
Stephen N.:is $10 million. We would say, well, we just don't have the
Stephen N.:resources to do that at all. It's a non starter. And that's
Stephen N.:where a lot of frustration happens with like, tech
Stephen N.:companies, marketing people, SAS salespeople, they just don't
Stephen N.:have the resources or they don't have the features and
Stephen N.:functionality or they don't have the manpower whatever it is. Do
Stephen N.:the things they want to do. And so there's a disconnect. And I
Stephen N.:think the logic model helps bridge that gap. And I think the
Stephen N.:key results leading into KPIs helps to bridge that gap too.
KJ:Great. That's good. I agree. But tell everyone what, what
KJ:that logic model is an illustrate to us, you know, at
KJ:the end example.
Stephen N.:Yeah, so every, every impactful outcome starts
Stephen N.:with resources and resources as people, technology, finances,
Stephen N.:any sort of asset, capital, yeah, capital. That's the
Stephen N.:resources, the things that you can use at your disposal to do
Stephen N.:the activities, the activities, the actions, we want to create,
Stephen N.:you know, this number of articles, or we want to build
Stephen N.:these types of features and functionality and that those
Stephen N.:activities lead to output. So the number of activities that
Stephen N.:you've done, and so we've created, you know, we've done 10
Stephen N.:integrations, that's all well and good. But that leads to the
Stephen N.:outcomes, which is the behavioral change that you're
Stephen N.:seeking, you can build 10 integrations to keep using that
Stephen N.:example, if nobody actually uses them, it's kind of a moot point
Stephen N.:doesn't matter. And then ultimately, the outcomes lead to
Stephen N.:impact. So those are more longer term. And those are related to
Stephen N.:economic impact, environmental impact, societal impact. So if
Stephen N.:you look at any sort of nonprofit company, they're
Stephen N.:really tied to this model, because they're trying to drive
Stephen N.:long term impact, like, how can we create a more sustainable
Stephen N.:planet? How can we, you know, lower, reduce climate change?
Stephen N.:How can we provide water and health care to you know, the
Stephen N.:globe, whatever? How can we put a man on a different planet?
Stephen N.:Like these are all long term, like three to 510 20 year
Stephen N.:impact? Things are trying to move the needle? That always
Stephen N.:comes back to the resources? Like do they have the money? Do
Stephen N.:they have the people? Are we working on the right things? Are
Stephen N.:we producing the right outcomes? Are we driving change? And if we
Stephen N.:do all those things, we will make an impact? Yeah, that's the
Stephen N.:logical model.
KJ:It's great. Yeah. Like it starts with sort of what assets
KJ:they have. Activities, how you deploy those resources and
KJ:assets. The outcome is the quantifiable measurement, we
KJ:deliver 10 integrations out the output is the quantifiable
KJ:performance 10 integrations, outcome is the quality, the
KJ:behavioral change, and the impact is closer to your is
KJ:closest, probably to a company's mission. Right? Yeah, impacting
KJ:society by you know, delivering better health care, whatever,
KJ:it's great. It's it really sums up everything in a very linear,
KJ:logical way.
Stephen N.:That's why it's called the logical model.
KJ:That's what it's called, the logic model.
Stephen N.:Maybe we can get into that one. Separately,
Stephen N.:because there's more there that we can unpacked, especially from
Stephen N.:a business standpoint, because every business leader wants the
Stephen N.:the impact of their business to be a higher valuation. Yeah. How
Stephen N.:can the value of my company be higher and higher and higher,
Stephen N.:but you got to put in the work beforehand?
KJ:Yeah, it'd be great. We could do a session on diagnosing
KJ:the most common pitfalls of the model error where people really
KJ:struggle with it, I have a sense that people really struggle with
KJ:resources and activities, like they want or maybe resources and
KJ:outputs, they want a certain output. They want 10
KJ:integrations, but they don't pile up the resources or plan
KJ:the resources necessary to deliver the output. You know,
KJ:there's the probably, maybe there's the chink the chain,
Stephen N.:there's more we can unpack there. But I think for
Stephen N.:now, it's pretty good. By the way, this day in history, Ford
Stephen N.:Motor Company, under Henry Ford in 1915. manufactures its 1
Stephen N.:million automobile at the river rogue plant in Detroit. So 1
Stephen N.:million automobiles.
KJ:That's output. Yeah. 1915 I think 15 Day at a million cars
KJ:built by 1915.
Stephen N.:Yeah, that's the model, probably the Model T. So
Stephen N.:that's your output. So your resources are your people and
Stephen N.:your assembly line and your factories and for factory
Stephen N.:workers. Yeah. And then you that's the number you create
Stephen N.:that many can. That's the output and then but the behavioral
Stephen N.:change the outcome is do people make the decision to buy these
Stephen N.:cars and yes, overwhelming? Yes. And why? is the impact of all
Stephen N.:that? Well, it's the Industrial Revolution. Yeah.
KJ:To say, just to say, Have you mobilized the earth's
KJ:population? Yeah.
Stephen N.:Good job, Henry.
KJ:Henry. That's pretty shitty impact, to be honest, like I
KJ:could do that.
Stephen N.:Yeah. I'd prefer to be running around on a horse.
