Horizon 2 · Descriptive Analysis

Descriptive analysis

The answer is usually already in your history. The work is reading it properly: a lot of data, and someone who has worked with data long enough to pull a decision out of it.

A needle worth finding

Most businesses are sitting on years of their own data, and the answer to the question they are asking is usually inside it. The catch is that it does not announce itself. It takes someone who has worked with this kind of data to look at the haystack and know where the needle is, and what to do once they have it.

That is descriptive analysis: not a prettier chart, but the pattern that turns into an action you can defend, because the evidence is sitting right there in your own numbers. Here are two we have built, of the hundreds these become.

Example one

The beat plan: who to visit, and when.

A rep left to their own judgement visits fifteen outlets and eleven are a waste of the day. We take the month's target, break it to each outlet on its own history, and plan each day around one high-value cluster, so the travel is tight and the time goes to the names that move the number. Pick a day to plot that day's route on the map.

Sales beat planBengaluru territory · sample data
Interactive · live map
Visit · high-value, on the route Optional · smaller, only if there is time Avoid · not worth the trip
The map is live (OpenStreetMap, CARTO). Outlet names and figures are illustrative sample data; the method, target to outlet to one clustered day, is real.
Example two

The payment pattern hiding in your ledger.

A customer is on 15-day terms with a 1% early-pay discount. The ledger says he has never once paid before day 31. The discount is money given away for a promise he does not keep, and the same history tells you exactly when to call.

Receivables pattern · Sunrise FoodsLast 6 invoices · sample data
From the ledger
What the history says
Agreed terms
15 days
Actual average
34 days
Days to pay, last 6 invoices
promised 15
fair 30
The move
  • Reset him to 30-day terms, with his own ledger as the proof. He is already taking the time.
    drop the 1% he never earned
  • Reclaim the discount. 1% on roughly ₹1.2 Cr a year was leaving on a promise he does not keep.
    ₹1.2L back a year
  • Call on day 28, not day 40. He clears payables on the last Friday of every month, so the reminder lands just before his payment run, not a fortnight after it.
Names and numbers are illustrative. The pattern, promised vs actual, fair terms, and the timed call, is the real method.
Two of hundreds

The one that matters for you is probably not on this list. It is in your data.

Talk to us →