A 20% Claim Without a Sample: A Lesson in Reading Marketing Numbers

Headline and Context

The Statesman published on 12 May 2025 a report, datelined New Delhi and credited to SNS, saying MSMEs in tier-II and tier-III cities registered an average 20% revenue increase after integrating with ONDC. The data were released by Easy Pay, a fintech firm that helps MSMEs onboard onto the network, and its founder and managing director, Nilay Patel, commented.

What the article said

The growth was attributed to better digital visibility and national market access for businesses that had previously operated only locally. Rajasthan, Gujarat and Maharashtra had the most businesses joining. A jewellery store in Pune reported that orders and income had doubled within months, and its owner said daily order volume "has doubled in just three months." A chips seller in Rajasthan sold 1,200 packets within an hour during a local ONDC campaign.

“Before you believe a number, ask who counted and what they sell.”

What the article did not say

It did not give the number of MSMEs surveyed, the method, the period over which the data were collected or how the average was calculated. It gave no state-wise counts. The only aggregate figure was 20%. The other numbers were single cases. And the source is a company that earns money by onboarding businesses onto ONDC, so it has a commercial interest in showing good results. That does not make the claim false, but it means the claim cannot be treated as independent evidence.

Questions to ask of any statistic: our checklist

1. Who produced it, and what do they sell? A vendor's data are marketing until corroborated.

2. What is the sample? Twenty businesses and twenty thousand give very different confidence.

3. Which businesses are in it? If only firms that stayed on the platform are counted, those that left are missing, which flatters the average.

4. What is the comparison? Revenue would have changed anyway in a festive or growth period. A fair comparison needs similar businesses that did not join.

5. What is "average"? A few big gainers can lift a mean while most firms see little change. A median is more informative.

6. Over what period? Three months and three years tell different stories.

7. Revenue or profit? Higher revenue can come with higher costs from fees, discounts and logistics.

How a small seller can test the claim on their own business

You do not need an external study to find out. Pick a representative group of products and list them on the network for a fixed period, say eight weeks. Record orders, revenue, fees, shipping, returns and time spent. Compare against your normal channels over the same weeks. If net contribution after all costs is higher, extend; if not, change the product mix or stop.

Why this matters beyond ONDC

Small firms are constantly offered numbers: an advertising platform that promises a multiple on spend, a software firm that cites average gains, a consultant who quotes a case study. A habit of asking the questions above protects scarce money. It does not require suspicion of everyone, only a request for the basis of the claim.

What the example does tell us

Even with its limits, the report points to something plausible: that businesses with local reach can gain from a larger market, and that early gains often come from a few products or a campaign. It also shows that visibility does not guarantee volume; the chips seller's sales were tied to a campaign. Promotion and presence work together.

For media and advisers

Journalists and consultants who repeat figures like this should ask for the sample and method before publishing. We were unable to obtain either, which is why we label the claim as a vendor's and do not use it elsewhere in this series.

A one-page template for your own test

Draw a table with these columns: product, listing date, views or enquiries, orders, revenue, platform fee, shipping cost, returns and net contribution. Fill it weekly. At the end of eight weeks, add up net contribution and divide by the hours you spent. If the hourly return is lower than what you earn in your normal channels, the new channel needs improvement before expansion. If it is higher, ask whether the result came from one product or many, because a single hit is less reliable than steady spread.

Respecting the businesses in the report

None of this is a suggestion that the jewellery store or the chips seller did not gain. Their experiences may be accurate and instructive. The issue is only that one or two cases cannot tell another seller what to expect.