A card-spending panel covers about 3% of Kalyra Mart's sales. Panel spending is up 14% this quarter, and the panel skews to younger urban shoppers. How do you turn the panel into a revenue estimate, and what biases do you correct for?

Calibrate the panel on history before scaling it: on eight past quarters, 14% panel growth maps to about 9.8% reported growth, roughly Rs 1,098 crore, below the 12% consensus. The naive scale-up says Rs 1,140 crore and a beat. The panel overstates growth because its young urban users shop more at new stores and pay more often by card. The estimate carries a band of about Rs 1,079 to 1,118 crore.

A poll of voters taken only outside college campuses is still data, but it is data about students. A card panel measures the shoppers who are in the panel, paying the way the panel sees, so its growth equals the company's growth only if those shoppers behave like all of Kalyra's customers. Three biases push this panel above the truth: its users are young and urban, where Kalyra's new stores are opening; more shoppers each year switch from cash to cards, which looks like growth in card data; and the panel's own membership changes as users join and leave.

Let history tell you how the panel translates. Put panel growth against reported growth for the past eight quarters and fit a line. Here reported growth has been about 1.8 points plus 0.58 times panel growth, so the panel overstates growth, and a 14% panel reading implies about 9.8%. The scatter of past quarters around the line, about 1.0 points, is your error, which gives a range of about two errors either side. Where possible, go further: reweight the panel by city and age to match Kalyra's store footprint, and hold the panel to users present in both years so that panel churn does not masquerade as growth.

Lead with the number and its range, then the call. The panel points to growth of about 10%, and the whole two-error range sits below the 12% consensus, so the data leans towards a miss rather than a beat. Then the caveats, briefly: eight quarters is a short history to fit a line on, a new store programme can change the relationship, and the alternative dataData from outside company filings, such as card spending, web traffic or satellite images, used to estimate results before they are reported. vendor's panel may itself have changed. A good analyst also checks whether consensus already moved on the same data, because if every fund buys this panel, the edge is in the correction, not the raw number.