Data Analysis for Business

Where this is used

Statistics is what turns the numbers you pulled out of Excel or SQL into an actual decision. Get comfortable with three tools — the mean, the standard deviation and the correlation coefficient — and you can spot when a KPI has gone off the rails, tell whether a marketing campaign really worked, and walk into a leadership meeting with evidence instead of opinions. Now that almost every company claims to be data-driven, this is the baseline literacy the job expects.

Start gently

Data analysis is a loop: collect → tidy → decide → check. Keeping to that order does more for you than any clever technique.

Key points

The Mean and Your KPI Baseline

Add up your n values and divide by n — the most basic summary there is. In practice you set the average monthly figure as a baseline and then watch for months that drift away from it. Its weakness is that one unusual value (a surprise bulk order, say) drags it off, so always read it alongside the median.

Standard Deviation (How Much Things Vary)

How far the individual values sit from the average. It is the foundation of "volatility" in investing and of process capability indices in quality control. The bigger the standard deviation, the less stable the numbers are. In Excel, the STDEV function gives it to you in one step.

Correlation (How Strongly Two Things Move Together)

A number from -1 to +1 that says how closely two variables move together — ad spend and sales, temperature and drink sales, and so on. Close to +1 means they rise and fall together; close to -1 means one goes up as the other goes down. Excel's CORREL function calculates it for you.

How Pivot Analysis Works

You summarise the data along two axes at once and compare across both. For example: rows = product category, columns = month, values = total sales. In practice this is Excel's pivot table or SQL's GROUP BY clause. In a single view you can see which segments are growing and which periods are slipping. It is also called cross-tabulation.

Linear Regression (The Basis of Prediction Models)

Fit a straight line through the relationship between two variables and use it to predict future values. The classic case is estimating sales (y) from ad spend (x). The slope b tells you how much sales go up for each extra ¥10,000 you spend on ads. Excel's LINEST function, or a trendline on a scatter plot, gets you there immediately.

See it drawn

How the drop-off rate moved before and after the change

It fell — but it could be seasonal. Do not call it a win until you have compared with the same period last year.

Jobs that use this

Data Analyst$110k

Marketing Analyst$95k

Management Consultant$150k

Data Analysis for Business

Start gently

Key points

See it drawn

Jobs that use this

Members-only from here

Comments