Why do eyewear stores open opposite a rival, and why did a dentist’s proximity turn out to predict ice cream sales? Rajit Bhattacharya has the data to answer both.
In this Water Cooler Conversation, the founder and CEO of DataSutram explains how location intelligence works, how a college hackathon became a B2B SaaS company, and how to pick your first customers wisely.
What is location intelligence? Rajit Bhattacharya’s core thesis
Location intelligence means collecting information about a place, from footfall to demographics, and using it to make business decisions. Rajit built DataSutram after seeing that large companies often decided on intuition, not because they wanted to, but because no usable external data existed.
His broader lesson for B2B founders is to resist building the full platform too early. Win one customer, deliver the solution end to end, and invest in technology once you have market fit.
- Data beats gut feel. Location data helps with store openings, lending decisions, marketing and demand planning.
- Do one proper pilot. Run it from start to finish for six months to a year before rushing to customer two.
- Pick growth-hungry customers. Look for companies in the middle of digital change, and say no to work that turns you into a services shop.
- Expect data pushback. Clients may resist numbers that challenge them, so trust in data is a people problem.
How did DataSutram start?
The idea began at Jadavpur University, where Rajit worked on a government of West Bengal project that used satellite data to suggest which crops farmers should grow. He saw that external data could guide decisions inside an organisation.
The company itself grew out of a retail hackathon run by the Future Group. His team used demographic and footfall data to explain sales at closed Easy Day stores and won, which showed that external data had real predictive value.
What does location intelligence do for businesses?
Rajit says location data comes from several sources: on-street collection, satellites and drones, mobile phones and public wi-fi. DataSutram works with about 250 data partners and combines them into a data layer, with a SaaS platform on top.
Because it is a B2B data layer, clients can look at an area, down to pin code or a 100 metre grid, and act on what they see.
- Retail: where to open stores, what products and inventory to stock, and where to target customers.
- Banking and finance: scoring locations for lending risk and flagging fraud.
- E-commerce: better address matching, which he called a major unsolved problem in India.
- Pharma and EV: demand mapping, and charging station placement.
How should B2B founders land their first customers?
Rajit warns that your first customer can shape your whole company, and can even turn you into a services business. His team learned this by rushing from first to second to third customer and later correcting the approach.
His rule for a new sector is to run one pilot from start to finish, then wait six months to a year to see output and feedback before moving on.
- Listen to the problems early customers describe, then solve those.
- Choose a growth-hungry client that values data, not necessarily the biggest one.
- Time your approach for when a company is already going through digital transformation.
- Learn from the pilot, then look for similar customers.
- Say no when the work pushes you into a service route.
Can a customer shape your product?
Yes. When COVID ended the need to open stores for one client, DataSutram helped them understand residential societies for medicine delivery. That learning led to the company’s hyper-local marketing tool.
Does a deep tech startup need heavy funding from day one?
Rajit says many founders focus too much on technology. In his view, you should build only what is needed to deliver the solution until you have a first customer.
He describes a chicken-and-egg tension: you need technology to win customers, but there is little point building it without one. Funds matter once you know what the market needs.
How does DataSutram keep its data accurate and compliant?
His answer is to cross-check multiple sources and validate samples on the ground. Zomato and Swiggy data can tell different stories, he noted, so no single source is absolute truth. Accuracy also drops in rural areas and smaller towns.
On privacy, he says founders should respect it even where laws are flexible. DataSutram says it checks that partners comply with laws such as GDPR and passes device IDs for targeting through platforms rather than contacting users itself.
- Combine several data partnerships to spot outliers.
- Use on-street checks to measure accuracy.
- Store historic data, because trends in a location matter.
Should students start up in their final year?
Rajit and his co-founders spent their final year doing service projects for companies rather than sitting for placements. Paying clients gave them confidence that the need was real.
He believes it is easier to experiment as a student because there is less pressure to sustain a lifestyle, and encourages college students to start. He also explained why he keeps the company headquartered in Kolkata despite most clients and investors being elsewhere.
Further reading and sources
Frequently asked questions
What is location intelligence used for?
It is used to turn location data into decisions. Examples from the episode include choosing store sites, assessing lending risk, improving delivery addresses, planning demand for pharma products and targeting customers in specific neighbourhoods.
How do I choose my first B2B customer?
Rajit suggests choosing a growth-hungry company that understands the value of technology and data and is already going through digital change. Run one full pilot with them and be ready to say no if they pull you toward custom services.
How accurate is location data?
Accuracy depends on the area and the sources. Rajit says combining multiple data partners and validating samples with on-street checks improves it, while accuracy drops in rural areas and small towns.
Conclusion
Rajit Bhattacharya’s story shows how a project at college can become a data business, but also how carefully founders must handle early customers. One well-run pilot teaches more than three rushed ones.
Your next step is to pick one prospective customer, define the end-to-end pilot, and agree on what success looks like before you build more.
Want more conversations like this? Join The Builders Club, our community of 35,000+ founders and CXOs, and explore more episodes of the Startup Founders Podcast.





