Writing

Three questions I now ask before making any sales chart

Data Visualisation and Analytics is teaching me that a sales dashboard is only as good as the questions behind it.

· 2 min read · Charlie Granier

This term I’m taking Data Visualisation and Analytics, and one idea from it has stuck with me: you can draw two honest charts from exactly the same sales data and have one show a business growing steadily while the other shows it in trouble. Nothing is faked. The difference is the time period, the axis and what gets grouped together.

If the same numbers can tell opposite stories, the chart isn’t the analysis. The questions are.

Here are the three I now try to answer before I build anything.

1. Who is going to act on this, and what decision will they make?

A sales director deciding where to put next quarter’s effort needs something different from a finance team checking the forecast or a customer success manager wondering which accounts are at risk.

If I can’t name the person and the decision, I’m probably making a chart that looks good and changes nothing. A lot of dashboards seem to show everything the data can show, rather than the one thing someone needs to know.

2. What is the right comparison?

A number on its own means very little. £2 million in sales this month is good or bad depending on what you compare it with:

  • the same month last year (seasonality)
  • the target or forecast
  • other regions or product lines
  • the number of customers or sales staff (is growth coming from more customers or bigger orders?)

Picking the comparison is where the business understanding comes in. Get it wrong and the analysis can be technically correct and completely misleading.

3. What would change my mind?

This is the one I find hardest. Before I look at the chart, I try to write down what I expect to see and what result would surprise me. If revenue is up, I’ll check whether margin is up too. If one region is booming, I’ll look at whether it’s one big customer.

It’s a small habit, but it stops me finding whatever I hoped to find.

Where AI comes in

AI tools are getting very good at turning a spreadsheet into a dashboard. You can ask in plain English for “sales by region over time” and get a clean chart in seconds. That’s useful, and I use it.

But it makes these three questions more important, not less. When charts are free, the risk isn’t that nobody makes them. It’s that everyone makes lots of them, quickly, without asking whether they mean anything.

Why I care about this

The role I want sits between a business and its technology. The tools will keep getting better. The thing I can offer is helping a team ask the right question, pick the fair comparison and explain what the numbers really say. It’s where my accounting, statistics and international business modules all meet.

My next step is to take a public sales dataset, build a small dashboard, and write up three things it tells me and one thing it can’t.