The growth is in Nagpur and Varanasi, not Mumbai
India added 6.6 GW of rooftop solar in the first half of 2026 and the districts leading it are not the metros. What that does to an EPC's sales geography.
India added 6.6 GW of rooftop solar in the first half of 2026, taking cumulative rooftop capacity past 30 GW by the end of May. Residential systems made up 82 percent of new installations.
The number that should change how you plan next quarter is not any of those. It is the district list. The markets driving adoption in early 2026 were Lucknow, Nagpur, Surat, Varanasi and Ernakulam. By state, Maharashtra led with about 17 percent of first quarter installations, Uttar Pradesh 16 percent and Gujarat 15 percent.
Smaller cities, not metros.
Why the metros are not the market
It is not mysterious once you look at what a rooftop sale requires.
It requires a roof somebody owns outright, a bill large enough to make the arithmetic work, and a household that can decide without a committee. Metros are full of apartments, which means no individual roof, a society to convince and common area load rather than household load. Tier two and tier three cities are full of independent houses with clear title, rising air conditioning load and an owner who can say yes in one conversation.
Add the tariff. Add that awareness of PM Surya Ghar is now near universal after crossing 50 lakh installations. The result is that the addressable market has moved outward from the places where most EPCs have their offices.
What this costs you to serve
Here is where EPCs get hurt, and it is an arithmetic problem rather than a strategy problem.
Serving a city 90 kilometres away is not the same business as serving your own. Run the numbers for your own operation rather than taking anyone's word for it, but the shape is this:
- A site visit in your home city costs you perhaps an hour of a salesperson's time plus fuel.
- The same visit 90 kilometres away costs half a day, and half a day again if a second visit is needed to deliver the proposal.
- A warranty callout at that distance costs a technician's entire day for a loose connection.
Two site visits and one callout in a distant city can consume more margin than the job carried. And because the jobs there are typically 2 kW and 3 kW residential systems, there is not much margin to begin with.
That is why the answer is rarely "expand into that city" and usually one of three narrower things.
Three ways in that actually work
1. Never leave without the proposal. The single biggest lever on distant selling is cutting two visits to one. If your salesperson can capture the roof, size the system against the bill and hand over a complete proposal before leaving, the economics of a distant enquiry change immediately. This is the argument in closing on the first site visit, and it matters twice as much at ninety kilometres as it does at nine.
2. Cluster, do not scatter. Take enquiries in a new city only until you have enough for a week of work in one area, then do them together. One trip, one crew deployment, one materials movement. An EPC doing one job a fortnight in four different towns is running the worst version of this business.
3. Recruit locally before you sell locally. A local electrician or a small installer who knows the DISCOM office and can attend a callout in twenty minutes turns a loss making service radius into a workable one. Pay them properly and standardise what they install so that quality does not become a lottery.
What to standardise before you spread
Distance punishes improvisation. Before working a new district, fix three things:
- A design standard. Two or three standard configurations for the roof types you see most, so a crew does not invent a mounting approach on site.
- A documentation pack. The DISCOM application, subsidy paperwork and handover documents assembled the same way every time, because you will not be there to fix an incomplete file.
- A service commitment you can keep. Publish a response time you can honour at that distance, then honour it. Fewer promises kept beats more promises made.
The strategic read
Rooftop solar in India is turning into a distributed, high volume, standardised business in places that are not where the industry's offices are. The EPCs who do well in the next two years will be the ones whose cost to acquire and serve a customer eighty kilometres away is close to their cost at home.
That is a process and tooling problem, not a marketing one. It is won by producing proposals on site, by clustering work, by standard designs, and by local hands. Where the remaining demand sits nationally is covered in 50 lakh homes done.
The short version
- India added 6.6 GW of rooftop solar in H1 2026 and passed 30 GW cumulative by end May. Residential was 82 percent of new installations.
- The leading districts were Lucknow, Nagpur, Surat, Varanasi and Ernakulam. By state, Maharashtra, Uttar Pradesh and Gujarat led the first quarter.
- Independent houses, clear roof title and a single decision maker are why smaller cities are outpacing metros.
- Distance is the margin killer. Two site visits and one callout can exceed the job's margin.
- Close on the first visit, cluster jobs geographically, and hire local hands before selling into a new district.
Sources
Keep reading
The collateral free loan most customers never hear about
Residential solar loans up to Rs 2 lakh need no collateral, run ten years and carry a six month repayment holiday. Most EPCs still lead with the system price.
Delhi will fund the whole 3 kW system. Then what?
Delhi approved zero upfront rooftop solar for homes under 400 units a month. What it does to a Delhi sales pipeline, and which customers it quietly leaves to you.
50 lakh homes done. The next 50 lakh are a harder sale
PM Surya Ghar crossed 50 lakh installations in July 2026. The households left over consume less, own less roof and need finance. What that changes for an EPC.