[MUSIC]

It's normal to do a bit of data discovery when you get a new data set.

But usually, the data analysis process starts in management.

Leaders of the company have goals for where they like the company to be.

And to make sure they hit those goals, leaders keep track of various variables.

For example, among other things, a manufacturing company would probably keep

track of expenses, defect rates and employee hours.

And they probably track these things over time.

That way it's easy to see how things are changing and measure performance.

When you use a variable to measure something, it turns into a metric.

So if the company has a goal to cut costs,

expenses might be a good metric to keep track of.

The metrics are always so generic.

And they differ from company to company.

After this video, maybe take a minute to look up some metrics for your industry.

It's always interesting to see what's being tracked.

Another thing to keep in mind is how you choose the metrics.

To help you pick the right metrics, here's four things you'll want to keep in mind.

One, measurability.

If you can't measure it, you can't use it.

Let's look at the example of a movie theater.

Knowing how frequently people watch movies at home

could be really useful information.

But that's just not something we can measure.

At least not reliably.

Two, actionability.

We should be able to impact the metric.

If we find that people are more likely to see movies when it's cold outside,

it's certainly interesting.

There's not a lot we can do to change the weather.

Three, easy to understand.

Our metrics need to be fairly simple.

If you wanted to use the ratio of daily popcorn sales to employee hours worked

as a metric, you'd better have a really good reason.

Four, goal-oriented.

When you are choosing metrics, keeping in mind the goal you are trying to achieve.

If the movie theater's are trying to raise popcorn sales,

it doesn't make a lot of sense to track how someone paid for their ticket.

Once you have chosen your metric, it's time to start calculating it.

Sometimes this is easy, but other times it can be tricky.

Let's look at an example where we own two movie theaters and

we're tracking what percentage of our revenue is made up by popcorn sales.

Let's say theater one's revenue is 20% popcorn and theater two's is 60% popcorn.

One way we could combine these would be to take the average.

So overall, 40% of our revenue is made up by popcorn sales.

But wait a minute, theater one brings in four times as much revenue as theater two.

So if we add up the raw revenue numbers,

it looks like popcorn sales makes up 28% of our revenue, not 40%.

When you're working with calculated rates, you wanna make sure that you're not

skewing the result due to different sample sizes.

Also, when you're comparing your data, you wanna make sure that all points are valid.

If theater one's popcorn machine was broken all last month,

then it probably doesn't make sense to include last month in our calculations.

Once you've calculated your metric, you'll want to check to see how it's performing.

Here, the key is to establish a good baseline.

If we're trying to increase monthly revenues,

a good baseline might be the average revenue from last year.

Though you'll want to be careful of seasonality.

Metrics tend to vary based on the time of the year.

For example, toy stores have a lot more revenue in December than they do in May.

Another term you'll need to be aware of is key performance indicator or KPI.

Key performance indicators are metrics used to make sure the company is executing

its plan.

A good example of a key performance indicator,

would be something like attendance figures for a movie theater.

If we had a goal of raising revenues, tracking attendance

figures would be a good way to let us know if we're on the right track.

Put another way,

attendance figures are key to knowing how the theater is performing.

There's a lot that goes in to choosing and calculating a good metric.

But by following these simple tips, you should be able to nail it every time.