What it means
Analytics is usually described in four levels. Descriptive analytics reports what happened, diagnostic analytics explains why, predictive analytics estimates what is likely to happen next, and prescriptive analytics recommends what to do about it.
Most organisations spend far more time on the first level than they admit. A monthly management pack full of tables is descriptive analytics, and it only becomes valuable when someone digs into why a margin moved rather than simply noting that it did.
The finance angle matters because analytics projects are investments like any other, with a cost, a benefit and a payback period. Software licences, data engineering time and analyst salaries are real expenditure, so the discipline is to attach a specific decision and a specific number to each initiative.
The most reliable wins tend to be unglamorous: pricing consistency, discount leakage, stock levels, collection timing and identifying unprofitable customers or products. These produce measurable margin improvements because they change a decision that is repeated thousands of times a year.
The main nuance is that analytics is only as trustworthy as the underlying data. If the same customer exists three times in the system under slightly different names, no amount of sophisticated analysis will produce a number worth acting on.
Scale matters less than people assume when choosing tools. A small business with clean accounting data can answer most of its important questions in a spreadsheet, and the case for a warehouse and dedicated platform usually arrives only when data sits in several systems that will not talk to each other.
In practice
Real-world examples.
Example
A wholesale food business analyses order data by delivery route and finds that one third of its routes generate negative contribution once vehicle time is allocated properly. It reshapes the delivery schedule and raises minimum order values on the worst routes.
Example
A subscription media company builds a simple model linking early usage to cancellation. Customers who fail to use a core feature in their first two weeks are flagged for a targeted onboarding email, which measurably reduces first quarter cancellations.
Example
A manufacturer compares quoted prices against invoiced prices across 18 months and finds an average leakage of 2.3% from unrecorded discounts and freight concessions. Tightening quote approval recovers most of that margin without changing list prices at all, and the sales team keeps its win rate because the leakage had been concentrated in deals that were never competitive on price in the first place.
Formula
Calculation
Return on an analytics investment = (annual benefit - annual cost) / annual cost, with payback period = annual cost / annual benefit expressed in years
A specialist retailer with $30,000,000 of annual revenue invests in analytics to improve markdown and discount decisions.
Analytics platform and data warehouse: $60,000
Two analysts, fully loaded: 2 x $60,000 = $120,000
Total annual cost: $60,000 + $120,000 = $180,000
Gross margin improves from 38.0% to 39.2% because discounting becomes more disciplined.
Gross profit before: $30,000,000 x 0.380 = $11,400,000
Gross profit after: $30,000,000 x 0.392 = $11,760,000
Annual benefit: $11,760,000 - $11,400,000 = $360,000
Return on investment: ($360,000 - $180,000) / $180,000 = 100%
Payback period: $180,000 / $360,000 = 0.5 years, or six monthsCase study
Seen in the real world.
Larkfield Components is an illustrative and clearly fictional maker of industrial fittings with revenue of $22,000,000 and 4,100 stock items. Management believed the product range was broadly profitable because the overall gross margin sat at a comfortable 34%.
An analytics review allocated freight, handling and warehouse space to each item. In this fictional case it showed that 1,300 slow moving items contributed just $180,000 of gross profit while absorbing an estimated $310,000 of handling and storage cost, making them loss making once fully costed.
Larkfield discontinued 900 of those items and raised prices on the remainder. Working capital tied up in stock fell by around $1,100,000 and reported gross margin rose to 36.5% within a year, which was worth roughly $550,000 of additional gross profit on unchanged revenue.
Watch out
Common mistakes.
- Starting with the tool rather than the decision. Buying a platform before identifying the decisions it should improve produces attractive dashboards nobody acts on.
- Trusting analysis built on dirty data. Duplicate records, inconsistent product codes and missing dates will quietly invalidate a result no matter how good the technique.
- Confusing correlation with cause. Two numbers moving together is a prompt to investigate, not proof that one drives the other.
Questions
People also ask.
What is the difference between analytics and reporting?
Reporting states what happened, while analytics goes further to explain why it happened and what should be done next.
Do small businesses need analytics tools?
Often not at first, because a well structured spreadsheet built on clean accounting data answers most early questions at effectively no cost.
How should the benefit of an analytics project be measured?
Tie it to a specific financial outcome such as margin points recovered, stock reduced or collection days saved, rather than to usage of the dashboard itself.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%