My Business Metrics Framework (Think Like a Senior Analyst)

0:00In the age of AI, a differentiating

0:01factor between analysts who get hired

0:03and promoted on the job and those who

0:05get stuck in the job market or in their

0:07current careers is the ability to tie

0:08metrics back to real business impact.

0:11And the ability to do that comes down to

0:13one main thing, which is understanding

0:15what actually drives the business and

0:17what metrics you need to track to prove

0:19that. So, if you are a working

0:20professional or an analyst trying to

0:22transition into data and you've been

0:24wondering, "How do I know what metrics

0:25actually matter for a business and what

0:27metrics do I need to show in my resume,

0:28my portfolio to land interviews and

0:31offers?" Then this video is for you. We

0:33are going to do a crash course on

0:34business metrics so you understand how

0:36metrics fit into the day-to-day job, how

0:38to actually decide which ones matter,

0:40and also how to connect these metrics

0:42back to real business impact. If you're

0:44new here, I'm Christine. I'm a former

0:45data director and a hiring manager who

0:47now helps working professionals and

0:48analysts stand out in the job market and

0:51on the actual job.

0:55So, how do you know what metrics

0:56actually matter? We are going to use one

0:58simple idea to make this a lot less

1:00overwhelming. You can think of a

1:01company's metrics like a layer cake

1:03where there's three different layers,

1:05each one holding up the one above it.

1:07So, at the very top you have the icing

1:09on the cake. This is basically the part

1:10that everyone outside the company

1:12actually sees and it's often also what

1:14creates the fastest impression of how

1:16the company is actually doing. So, in

1:18terms of who actually tracks these

1:20numbers, that would be the exact anyone

1:22in leadership and people on the board,

1:24for example, if it's a public company.

1:25These are metrics that are focused on

1:27the highest level business performance

1:28and they are also often shared with the

1:30public if it is a public company. Let's

1:32go with an example. Pretend you are a

1:33product analyst. Then a company-level

1:35metric on the product side would be

1:37something like active users. This is the

1:39total number of people who are actually

1:40using the product and it captures things

1:42like product engagement and adoption.

1:44The next layer down is the actual body

1:46of the cake. So, team or department

1:48metrics. This is right under the actual

1:50icing and this is where leadership

1:52translates that top icing layer number

1:55into a larger set of metrics that

1:56capture what the team is focused on in

1:58the day-to-day job. So, in terms of who

2:00tracks these metrics, that would be

2:01people like directors, team leaders, and

2:04managers. And these metrics directly

2:06contribute to the icing layer metric and

2:08represent team priorities.

2:10>> [music]

2:10>> So, for example, again, if I was a

2:12product analyst and I was working with a

2:13product team, the team level metrics

2:15would be things like retention rate or

2:17adoption rate. And retention rate is the

2:19percent of people who stick around month

2:21to month or year over year given their

2:23satisfaction with the product. The

2:25bottom layer is individual metrics. This

2:27is where you have the base that's

2:28holding up the entire cake. And this is

2:30often the most detailed or most granular

2:32layer metrics. Think of all of the tiny

2:34grains that make up the crust of the

2:35cake. So, without it, you don't really

2:37have a cake at all. These metrics are

2:39usually tracked by employees or

2:41managers. And these are actually metrics

2:42that have more to do with productivity,

2:44capacity, and accountability. So, for

2:47example, again, if I'm a product analyst

2:49and I wanted to track the productivity

2:51of a product team or the progress that

2:52we're making as a team, then I might

2:54look at something like the numbers of

2:55features shipped. And this is the actual

2:57changes that an individual product

2:59manager or an analyst help implement in

3:01the product. When we're thinking about

3:02metrics, we want to actually look at

3:04this cake all together, not as three

3:05separate slices. Because each of these

3:07layers actually directly feed into the

3:10layer above. In our example, the number

3:12of features shipped actually directly

3:14impacts the adoption and retention rate.

3:17And adoption and retention rate also

3:18directly impact the total number of

3:20active users. So, if we put it all

3:22together, it looks something like this.

3:23We've got our top layer company metrics,

3:25department and team metrics, and then

3:27individual level metrics. And this is

3:29what it looks like for one vertical

3:30across product. So, when it comes to the

3:32question of what metrics actually

3:33matter, the answer is it depends on what

3:36layer of the cake you're working with.

3:37So, if you're talking to someone in

3:39leadership, let's say, a CEO or someone

3:41in exec, then that usually be focused on

3:44company level metrics. And individual

3:46level metrics are going to be a little

3:47bit less relevant to them. But if you're

3:49talking to someone like a team lead or a

3:51manager, then individual-level metrics

3:53might be exactly what they're focused on

3:55in the day-to-day. As you can see, being

3:58a good communicator is half the job of

4:00being a good data analyst, which is why

4:02we need to keep in mind what layer of

4:03the cake our audience actually lives in.

4:05By the way, if you want to download my

4:07business metrics guide and shortcut your

4:08understanding of business metrics, so

4:10you can sound like an experienced

4:12analyst in interviews and on the job,

4:14then you can download that down below.

4:18Okay, so now let's connect this

4:20hierarchy to what your actual day-to-day

4:22looks like as a data analyst. So, across

4:24every layer of this cake, you have

4:25different versions of teams. So, it's

4:27usually some version of marketing,

4:29sales, product, operations, customer

4:31success, or finance. And the list is

4:33going to change a little bit depending

4:34on what company you're working at. So,

4:36consulting is going to be different than

4:37health insurance, but there's usually a

4:39significant overlap. So, if you

4:41understand team metrics and how they

4:42overlap with the layer cake, you can

4:44already shortcut your way to sounding

4:46like an experienced analyst who has real

4:48knowledge about the company and

4:50industry. So, let me show you how,

4:51starting with the three teams that are

4:53most common, which is marketing, sales,

4:55and product. So, for marketing, their

4:56main question that they're asking is how

4:58is demand for our product? So, an

5:00individual-level metric, let's say

5:02someone who's actually working on the

5:03marketing team and just tracking their

5:04progress for the month, they may look at

5:06something like posts published, right?

5:08Then the department or team metric that

5:10they're trying to directly influence is

5:11cost acquisition or cost per lead,

5:14whereas a company-level metric is going

5:16to look at the total revenue that came

5:18directly from these marketing channels.

5:20Now, sales is going to be asking

5:21question like how many new customers are

5:23we actually closing? So, their

5:24individual metric for a salesperson

5:26would be something like the total number

5:27of calls made, whereas for a department

5:29or a team, that might be something like

5:31sales cycle length or win rate. So, how

5:33many or what percent of calls they

5:35actually bring on board as a customer.

5:38And then the high-level company-level

5:40metric would be something like revenue,

5:42the revenue that is going to directly

5:44from a sales team or total revenue or

5:46something like ARR, uh annual recurring

5:48revenue. And then, if we add the last

5:50layer here, so product, which we already

5:52spoke about, the main question that

5:53they're asking is, are people getting

5:55real value from this? An individual

5:57metric might be tracking something like

5:59total number of features shipped, where

6:01engineers and designers are all

6:02collaborating on making these updates to

6:04the product. And then, a department or a

6:06team is looking at something like

6:07adoption rate, retention rate,

6:09activation rate. And the top-layer

6:11company metric is something like active

6:13users. So, this is what that table looks

6:14like all together.

6:18So, once you got this pattern down from

6:20marketing, sales, product, you can start

6:22to read this table in a similar way for

6:24any other team. Now, depending on the

6:25role that you're targeting, you want to

6:27make sure that you start to study the

6:28metrics for that specific team. That's

6:30going to help you sound like you've

6:31already worked there before you've had

6:32the actual job. This is actually the

6:34exact instinct that helped my student

6:35Nimrat, who was a student from my most

6:37recent cohort, who just landed a job as

6:39an inventory analyst at Sonoco, which is

6:41a huge fuel and logistics company. So,

6:44when she walked into that interview, she

6:46wasn't learning this kind of language

6:47for the first time. She was already

6:49translating skills that she had from

6:51another context to this version of that

6:53company.

6:57So, a North Star metric actually

6:58measures the core value a product

6:59delivers, while also tracking long-term

7:02business growth. So, you can think of

7:04these metrics as an actionable compass.

7:06I'm going to give you an example in a

7:07second. And then, a vanity metric looks

7:09impressive on a slide, but it doesn't

7:11actually translate to meaningful

7:13business results. So, you can think of

7:15hype metrics here. And these are easily

7:17inflatable and rarely actually guide

7:19business decisions. I remember when I

7:21was working as a data analyst at Vimeo,

7:23and during the company's IPO, I was the

7:25lead data analyst who was working

7:26directly with the investor relations

7:28team to calculate and design the metrics

7:30that we were going to share with the

7:31public. So, we had an investor page that

7:34basically bragged about video usage. So,

7:36the total number of minutes watched, the

7:38total number of users who had ever

7:39logged in, the total number of videos

7:41ever uploaded. These are really big,

7:44exciting numbers. But those numbers

7:45don't actually tell you how the person

7:47engaged with the product in the

7:48day-to-day. So, that's what vanity

7:50metrics are for. Here's another example.

7:52So, 50 plus improvements and 100 plus

7:54bug fixes. Without more context on what

7:57these bugs and actual improvements are,

8:00these are vanity metrics that create

8:02hype more than capture real value. So,

8:04let's go back to our product example.

8:05Some North Star metrics on the product

8:07side for Vimeo would be something like

8:09active users or average number of videos

8:12uploaded per active user, right? So,

8:14these have directly to do with how happy

8:16and engaged people are with the product.

8:18Whereas a vanity metric might be

8:20something like the total number of bugs

8:21fixed or the total minutes watched

8:23across time. This could be something in

8:25the millions or the billions. I don't

8:27actually remember, but you can see that

8:28these two metrics can sound like they're

8:30really, really big without actually

8:32changing or impacting the direct

8:34customer value. That's why when you're

8:35working on portfolio projects or talking

8:37about examples in interviews, you want

8:40to make sure that you're focusing on

8:41North Star metrics instead of vanity

8:42metrics. So, if you want to run a test

8:44for if something is a North Star metric

8:46for your portfolios and thing about

8:47interviews, remember that a North Star

8:49metric can only go up when the business

8:52is genuinely getting better. I think

8:54most of the time North Star metrics are

8:55those that you want to go up. There

8:56might be some edge cases where a metric

8:58is actually something that you want to

8:59decrease. But most of the time, you want

9:01to raise that number. So, if someone can

9:03game that number or it could rise while

9:05the underlying business is actually

9:07getting worse, then it's likely a vanity

9:09metric.

9:13So, once you have these three

9:14frameworks, right? We have our metric

9:15cake, we have our North Star versus

9:17vanity metrics, and we also understand

9:19how this maps onto team metrics, we want

9:21to put this together and use it in

9:24portfolio projects, in our interviews,

9:26in our resumes to sound like an

9:28experienced analyst. So, here are three

9:30ways to put that into practice

9:31immediately. For portfolio projects, you

9:32want to gear your project towards a

9:35specific industry or domain knowledge

9:37instead of some generic data set. So, do

9:39not use Kaggle data sets here. You want

9:41to identify two or three North Star

9:43metrics for your target company or

9:45industry and also center your project

9:47around investigating the trends and

9:49fluctuations in those metrics, not just

9:52describing the data. So, here's an

9:53example project where we actually define

9:55what these North Star metrics are up

9:57front and we can see that these North

9:58Star metrics are really relevant to an

10:01e-commerce company or to any kind of

10:03company selling a physical product. So,

10:05this project will stand out for those

10:07kinds of roles. For dashboards, so

10:08whether you're building for a project or

10:11on the actual job, you want to make sure

10:12that you have a really clear audience in

10:14mind and you're building for their layer

10:16of the cake. So, if you're designing for

10:17an exec, you want to prioritize top

10:19layer metrics like revenue and active

10:21users. If you're designing for a sales

10:23team to track their team performance,

10:25you want to use metrics like total calls

10:27made and win rate. Now, on the job, this

10:29is something that we would actually do

10:30in requirements gathering with

10:31stakeholders, but that's like a good

10:33benchmark to start with. So, here's an

10:34example dashboard that tracks the number

10:36of applications to a program and you can

10:38see at the top we're focusing more on

10:40team and department metrics that we

10:41would want to operationalize in the

10:43day-to-day job. And then under that, we

10:45have metrics that get a little bit more

10:46granular that we can then use to

10:48understand what's driving those top

10:50layer metrics. So, for your resume, when

10:52you show metrics that overlap with the

10:54industry and the team that you're

10:55applying to, it really helps a hiring

10:57manager see the relevance of your work.

10:59This is not about listing every single

11:01metric that you know, but more so about

11:03showing at least two or three metrics

11:05that show that you have experience in

11:07the same day-to-day data that you'd be

11:09working with on the job. So, here's a

11:10resume by a student who's focusing on

11:12metrics like total sales, total revenue,

11:14material and volume savings, the size of

11:17the data set, processing time, and also

11:19the size of the sales team. If you found

11:20this video helpful, make sure to

11:22subscribe and check out my other one

11:23here about metrics, KPIs, and OKRs. See

11:26you there.