Understand Business Metrics for Data Analysts (Most Get This Wrong)

0:00In this video, we're gonna be talking

0:01about the Pyramid Framework.

0:02We're just gonna help you understand business metrics

0:04and how they relate to KPIs and OKRs,

0:07as well as what you actually need to know about them

0:09in order to land a job as a data analyst.

0:12So if you've ever wondered,

0:13what is a business metric really?

0:15How does it relate to KPIs and OKRs?

0:17How many of them do I actually need to know?

0:19And how do I speak about them

0:20so it sounds like I actually know what I'm talking about?

0:23Then this video is for you.

0:24We're gonna be talking about

0:25one of the simplest mental models

0:27for understanding the landscape of metrics.

0:29Then I'm going to share a trick that enables you to apply domain knowledge of business metrics

0:33across industries. And then at the end, I'm going to share some practical next steps

0:37for improving your understanding of business metrics so that you can land a job as a data

0:41analyst. If you're new here, I'm Christine. I'm a former data director and a hiring manager who

0:45now helps people break into their first job in data by learning how to think, speak, and operate

0:50like the top 1% of analysts out there. Let's dive in. Now, these days, companies don't just want

0:58data analysts who know how to build reports and dashboards. They want data analysts who actually

1:03speak their language. And what that means is understanding their numbers, their goals,

1:07their stakeholders, and how this all ties back to what's available in the data, which is why

1:11many data analyst job descriptions these days say things like help measure and report key metrics,

1:17or own and establish KPIs. And you might be wondering, isn't this the company's job? Don't

1:21they already know what numbers they actually want to track and how those numbers are doing? The

1:28to make sense of what's going on in these values,

1:30help pull the most important numbers to the surface,

1:33and help see whether or not those numbers

1:34are performing well or badly.

1:36By the way, if you want to shortcut

1:38your understanding of business metrics

1:39and save about 20 hours of Googling,

1:41I have made a business metrics cheat sheet for you guys,

1:44which is going to show you the top metrics

1:46for popular data analyst industries,

1:48how to speak about them,

1:49what stakeholders actually care about them,

1:51and show you how to integrate these metrics

1:52into your projects.

1:53So make sure you download that below.

1:54And if your goal is to become a standout data analyst

1:57in six months or less,

1:58then you might wanna check out my mentorship program,

2:00the Analytics Accelerator.

2:02I've just opened spots for the next 30-person cohort

2:05where we give you real-time feedback

2:07and tailored mentorship and guidance

2:09so that you can learn the business intuition,

2:10communication skills, and standout technical skills

2:13to land that data job this year.

2:14In order to do this,

2:15you need to understand the concept of the metrics pyramid.

2:18Now, metrics are everything that we can track.

2:20KPIs are the metrics that actually matter,

2:22and the OKRs are the change that we're trying to drive.

2:25Let's break this down a little bit.

2:26So metrics are literally anything that you can track as long as it's a quantitative value and it's a metric

2:31So for example, just the other day

2:33I was doing some analysis with my marketing lead on the performance of

2:36this YouTube channel and the metrics that we could track are things like

2:40Views subscribers number of likes number of comments the video length average view duration

2:45Impressions the number of people who watch me on two times speed because I'm talking too slowly

2:49You get the idea so at a company metrics are the quantitative values that live in the data

2:54And at this point, we're not yet deciding whether or not these metrics are actually

2:58relevant to the company performance or whether or not they're actually helpful to track.

3:02Some metrics are important and others aren't. And that actually brings us to the next layer.

3:07Now, the KPIs layer is where we start moving on to the metrics that we actually care about.

3:11It's where we start asking ourselves, OK, out of the bajillion metrics that we could track,

3:15what are the three to five that actually tell us if we're doing well or badly?

3:18So for example, going back to my YouTube situation,

3:21if I asked my marketing lead,

3:23can you do some analysis that helps us understand

3:24the overall performance of our videos?

3:26Then he defined the KPIs to be average view duration

3:29because this is a really big factor

3:31for how YouTube actually ranks your video.

3:33The click-through rate,

3:34because this tells us whether or not our thumbnails

3:36and titles are effective.

3:38And the number of new subscribers per video,

3:40because this of course tells us

3:41if whether or not you guys are interested

3:43in the videos that we're making.

3:44And just like that,

3:45the hundred metrics that we could have tracked

3:47then become just three.

3:48the three that we actually care about.

3:50Most teams at work will actually really care

3:52about five to 10 KPIs maximum

3:54because these are the metrics

3:55that are used to track performance.

3:57They're also often shared

3:58with the higher level leadership team.

4:00And oftentimes the leadership team has to sign off

4:02on each team's KPIs and agree,

4:04yes, this is what we're going to use

4:06to evaluate your team's performance.

4:08So a good rule of thumb is that

4:09if it seems like a leadership team

4:10really wouldn't care about this metric at all,

4:12then it's probably just a metric and not a KPI.

4:15So at the top of the pyramid,

4:16we have something called OKRs

4:17or objective key results.

4:19And this is where leadership

4:20is usually setting the direction of the company

4:22and OKRs are what makes that direction concrete.

4:25It's basically saying,

4:26here is where we want to be by this date.

4:28It's more big directional stuff.

4:30So OKRs for this channel might be something like,

4:32we want to hit 90K subscribers by December,

4:35or we want to increase course revenue by 30%

4:37or decrease our turn by 5% by the end of this year.

4:41These aren't just numbers.

4:42They're actually signals about

4:44what matters the most right now.

4:46And usually at a company, there's maybe three to five main OKRs that get circulated across the

4:51entire company because they're the big top line priorities that leadership wants everyone to

4:56align themselves on. While metrics are the quantified values that live in the available data,

5:01the KPIs are the metrics that get surfaced in dashboards because they're the metrics that

5:05teams have decided to report their performance on. And OKRs usually live in town halls or all hands

5:10or leadership updates because of the big priorities that everyone should be aware about. So they're

5:14all metrics at the end of the day. KPIs and OKRs technically are all quantified values, but they're

5:20all at different levels of altitude and different levels of importance. So how does this actually

5:25impact how you should communicate on the job and in interviews? I typically find that more junior or

5:30new data analysts will live in the metrics layer. When they share their insights, they'll say things

5:34like the total page views was 47,000 or the average open rate across the quarter was about 20% or that

5:41bounce rate was 2.3 percent and they'll just kind of stop at the actual metrics whereas someone who

5:45is more experienced and more advanced can tie the metrics to the kpis and the okrs so they might say

5:52things like our two main kpis this quarter of conversion rate and aov were both up about 18

5:58this actually contributes most directly to our overall okr of increasing revenue by 30 this

6:04quarter so they'll kind of travel through the different layers of the pyramid and this is

6:11on the actual job.

6:14Okay, so now you're probably thinking,

6:16cool, I get the pyramid,

6:17but how do I know what metrics I should actually focus on?

6:20Because every company seems so different.

6:22Now here's a cheat code

6:22that is going to make your life so much easier.

6:24It's the concept that most metric pyramids

6:27are similar within departments across industries.

6:29So what I mean by that is most companies,

6:31whether they're selling software, solar panels, or socks,

6:34usually will have their version of a marketing team,

6:37a finance team, a product team, an operations team.

6:40Even if their name's slightly different

6:42across different industries,

6:43the function that they serve is pretty much the same,

6:45which means that if you can understand the metrics pyramid

6:47within the marketing team in one industry,

6:49you've also largely understood marketing metrics

6:51in another industry.

6:52So with marketing, for example,

6:54if you're working at a tech startup or a SaaS business

6:57or a supply chain or logistics company,

6:59then the stuff that the marketing team cares about

7:02largely stays the same.

7:03For example, the metrics layer might have something

7:06like cost per click, the total number of page views,

7:09total traffic, session length, engagement, reach, impressions, whereas the KPIs might end up being

7:16something like customer acquisition cost or conversion rate or the total page views or

7:22traffic. And then the final OKRs would sound something like decrease CAC by 15% or increase

7:28the number of qualified leads by 20%. So you can see that the pyramid stays the same even if you're

7:33swapping out the product. Now, I used to feel pretty overwhelmed thinking about how do we as

7:38analysts help choose which metrics should become KPIs. And this is actually also an interview

7:43question that I once failed for an interview with Facebook when they asked me out of all these

7:47metrics, which one do you think is most important to evaluate the health of the newsfeed? And over

7:52time, I realized that most companies are picking their KPIs on two main factors. Number one, can we

7:58actually influence this metric? And number two, does this metric genuinely reflect performance?

8:03So for example, for a marketing team, one of the metrics that you might have available to you is

8:06how long someone spends on the actual website.

8:10But if someone spends 20 minutes versus just one minute

8:13and still buys the product,

8:14does the time that that person spent on the website

8:16really matter?

8:17So ultimately, most teams will pick around five to 10 metrics

8:20to become their main KPIs.

8:22And yeah, KPIs can overlap between teams.

8:25So for example, the finance team and the sales team

8:27will probably both care about a metric

8:29like monthly recurring revenue.

8:30It actually makes a lot of sense

8:32because it means that you don't have your teams

8:33optimizing on different things.

8:35So once you realize that marketing metrics in fintech look very similar to marketing metrics

8:39in e-commerce, you stop feeling like you need to learn an entirely new language just to switch

8:44industries. You just need to learn the department pyramids once. And once you understand that,

8:48you can step into any marketing team and say, oh, what's the cat currently? And what's the

8:52forecast for this quarter? And that immediately makes you sound like you've already worked there

8:56for months. So how do you start implementing metrics on your data journey so that you sound

9:00like an experienced analyst? Well, the first step is to know the top 10 metrics in your industry.

9:04And when I say know the metrics, what I mean by that is you should understand how the metric is

9:09calculated from a technical standpoint. So what is the actual formula for that metric? Then how do

9:14you speak about that metric in plain English or whatever language so that you can explain it to

9:19a non-technical stakeholder? And lastly, how does that metric actually relate to a different

9:23stakeholder team? Once you have a grasp of this, you're immediately going to sound like someone

9:27who's already worked in that industry compared to someone who's learning it for the first time.

9:31The second step is to choose data sets that actually map to real teams.

9:35So when you're picking data for your portfolio projects,

9:37always think about the metrics in this data set.

9:39What teams do they actually relate to?

9:41Is it a finance team?

9:42Is it a sales team, a product team, a marketing team, an operations team,

9:45or a team that's going to be specific to the industry that you're interested in working in?

9:49If you can't answer that question,

9:50then it points to the fact that it might not be relevant to an actual company.

9:54The best kind of data sets usually feature metric categories that relate across domains.

9:59so revenue metrics marketing metrics customer metrics usage metrics and engagement metrics if

10:05you want to learn more about how to find these data sets besides on kaggle then check out my

10:09portfolio playbook playlist here now the third step is to define your kpis up front so in your

10:13portfolio projects and in speaking about metrics on the actual job it's usually helpful to make sure

10:18that you're being really clear about what metrics you're focusing on so you might actually say you

10:22know for this analysis the kpis we decided to focus on are total sales refunds and average order

10:28value because it's tied to the overall team goal of raising revenue by 10%. This helps make sure

10:33that everyone is on the same page about what numbers matter, whether it's hiring managers or

10:37your stakeholders on the job. If you want to sound more experienced, then start studying things like

10:41investor decks, earnings reports, and earnings calls, where you can literally see how leaders in

10:45the industry talk about metrics, KPIs, and OKRs. So if this intro to business metrics was helpful,

10:51let me know in the comments below as well as what industry you're targeting, because

10:55if enough people comment, then I will turn this business metrics idea into a metrics series where

11:00I'm going to introduce the top 10 metrics across industries and give you even more deep dive

11:05frameworks for how to apply it in your job hunt and on the actual job. And if your goal is to become

11:09a standout data analyst in six months or less, then you might want to check out my mentorship program,

11:14the analytics accelerator. I've just opened spots for the next 30 person cohort where we give you

11:19real-time feedback and tailored mentorship and guidance so that you can learn the business

11:23intuition, communication skills, and standout technical skills to land that data job this year.

11:28That's it for this video. I will see you guys in the next one.