Monthly sales forecasts for brand finance teams, checked against search data
TL;DR Quarterglass helps finance teams at mid-size Indian consumer brands forecast next quarter's sales by adding public attention signals, like search interest and marketplace rankings, to their own sales history. The CFO pays a monthly subscription. It earns its keep where budgets, inventory buys and lender talks all hang on one forecast.
Indian consumer brands now sell on their own website, on marketplaces and on quick-commerce apps. E-retail reached about US$60 billion (roughly ₹5.8 lakh Cr) in 2024, and quick commerce already takes over two-thirds of online grocery orders[1]. Every channel reports on its own schedule, so the finance team builds its forecast from last year's numbers plus the founder's gut.
When demand turns, the forecast catches up a month or two late. By then the brand has ordered stock it can't sell or promised a lender numbers it will miss. I couldn't find a single public figure on forecast error at Indian brands, which tells you something - most of them do not measure it.
Quarterglass plugs into the brand's accounting and sales data and then adds outside signals that move before sales do - search interest for the brand and its category, marketplace ranks, ratings counts and app-store activity. A model learns which of these led sales in the brand's own history, and by how far. Every month the CFO gets a forecast with a range, plus a short note on which signals moved it.
Search data needs a careful hand. The search engine's own guidance says its trends data is a sample, scaled 0–100 with noise added for small queries, and should be one input among several[2]. Quarterglass treats it exactly that way.
The seed got a rewrite. It pitched attention data for financial forecasting, which in India means selling signals on listed companies to investors. That can count as research under SEBI's Research Analysts Regulations[3], which need registration. Selling to the brand's own finance team sidesteps that and serves a buyer with a plainer need.
The CFO or finance head pays, and the founder uses the output in board and lender meetings. Brands selling ₹50–500 Cr a year are the sweet spot. My assumed price is ₹30,000–75,000 a month. Big enterprise planning suites serve larger firms and don't describe a built-in feed of outside signals[4].
Quick commerce has added a fast-moving channel with its own demand swings[1], and the cost of joining many data feeds to a forecasting model has dropped. With 270 million people shopping online[1], search and marketplace activity now covers enough buyers to carry a real signal in many categories.
Nobody publishes a count of Indian consumer brands with a finance team that forecasts monthly. The solid number is the channel's size - about US$60 billion of e-retail and 270 million online shoppers in 2024[1]. One US alternative-data firm's 650+ customers show that people do pay for outside demand signals[5].
One US alternative-data firm sells spending and panel data to investors and corporate strategy teams in the US[5]. Planning suites handle planning but only know the data you feed them[4]. Indian retail-analytics tools stick to stock and shelf data. So the gap is a forecast service for Indian brand finance teams that brings the outside signals in for them.
| Difficulty | Moderate (34/100) |
| Starting capital | ₹50–90 L |
| Time to launch | 6–9 months |
The modelling is ordinary. The hard bit is cleaning each brand's messy sales data across channels and then proving that the outside signals actually add accuracy for that brand. For some categories they won't, and you have to tell the client so plainly.
No sector licence is needed as long as you sell forecasts to a brand about its own business. The line to watch is investors.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't trust AI forecasts to set budgets without a human check, and don't present a signal as useful if it didn't cut error in back-tests.
Want to work out the AI layer for Quarterglass? Talk to Yash →
I'd back a founder with FP&A or demand-planning experience at a consumer company. If you want to sell trading signals then this is the wrong idea, because that road needs SEBI registration. Prove one thing first - for five brands, outside signals cut forecast error on held-back months by at least a fifth.
Plan for roughly ₹50–90 L, mostly on a small data team and the work of cleaning each brand's sales data.
Not if you sell to brands about their own business. Selling views on listed companies to investors can need SEBI research analyst registration, so get legal advice first.
Yes, consumer brands everywhere struggle with forecasting. Check the local securities rules if you go near investors, and the data protection law for customer data.
The free Quarterglass validation kit has the line-by-line capital breakdown, the effort scoring, the data and SEBI checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for brand CFOs, and 5 assumptions to test. Download the kit →
Thinking about building Quarterglass or something close to it? The make-or-break work is showing, brand by brand, that outside signals beat the CFO's own spreadsheet. If you need a hand, you know where to find me. Talk to Yash →
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.