Selling Solar in
Washington DC: Small Roofs, Big Economics

Washington DC is not a normal solar market, and anyone who has actually sold there knows it inside of a week. The housing stock is row houses and townhomes with small, flat, obstructed roofs. Historic districts restrict what can go on a street-facing plane. Neighboring buildings and mature street trees throw shade that no windshield estimate can account for. Two houses that look identical from the sidewalk can differ by half the array size.

The Roof Problem

In a suburban market you can survey a roof from the street and be roughly right. In DC you cannot. Usable roof area is the whole deal, and a rep can burn an entire day on homes whose roofs were never going to carry a system worth selling.

Row house roofs are small to begin with. Take off the space lost to HVAC equipment, vents, stair bulkheads, and required fire setbacks, and what remains is often a fraction of what the building footprint suggests. Add a mature tree canopy and a taller building next door, and the difference between a good address and a dead one is invisible from a car.

Google Sunroof: Why the Roof Comes First

The single biggest lift in every one of these funnels came from putting Google Project Sunroof at the front of the form. Sunroof uses Google's aerial imagery and 3D roof modeling to calculate how much usable roof a home actually has, how many hours of usable sunlight it gets per year after shading from trees and neighboring structures, and how large an array that roof can physically carry.

That let the landing page answer the homeowner's real first question — will solar even work on my house? — before asking them for a phone number. Instead of a generic lead form, they saw their own roof outlined, a recommended kilowatt size, and an estimated production number in kilowatt-hours. The page did the sizing conversation the rep used to do on a porch with a clipboard.

It also did the disqualifying for us, which mattered just as much. Heavily shaded roofs, north-facing slopes with almost no usable plane, and homes with too little roof to justify a system got filtered out at the top of the funnel instead of consuming an appointment slot. Reps stopped driving to houses that were never going to pencil.

Why the Economics Still Worked

DC compensates for its difficult roofs with some of the strongest solar economics in the country. District SREC prices ran far above neighboring states, and every megawatt-hour a system produced generated a certificate the homeowner could sell. Add the federal investment tax credit and high local electricity rates, and a smaller DC array could out-earn a much larger system in a cheap-power market.

That reframed the entire pitch. In DC we were not selling square footage of panel — we were selling a production asset with a second revenue stream attached to it. The funnel had to communicate that, because a homeowner comparing a DC quote against a national average would otherwise see a small system at a high per-watt cost and walk away from the best deal available to them.

The Numbers We Were Selling Against

Residential systems in these markets averaged roughly $25,000 to $35,000 installed before incentives — somewhere around $3.50 per watt for a typical six-to-nine kilowatt system. The federal investment tax credit came off the back of that, and state and utility incentives moved the net number further depending on the market.

Because the ticket was that size, the funnel could not treat price as a footnote. We put the two ways to buy directly on the page, side by side, and let the homeowner self-select before the appointment was ever booked.

How we sold loans

Most homeowners bought on a solar loan. The pitch was a straight payment comparison: their current utility bill against a fixed monthly loan payment on a system they own, with zero down at signing. On a $30,000 system amortized over a long-term solar loan, that payment frequently landed at or below what they were already sending the utility every month — and the utility number climbs every year while the loan payment does not.

The funnel supported that pitch directly by pulling the homeowner's actual monthly bill into the estimate, then showing the projected loan payment beside it. Ownership was the other half of the argument: on a loan the homeowner owns the array, claims the tax credit themselves, and keeps the equipment and the production for the life of the system.

How we sold cash

A meaningful slice of buyers paid the full amount upfront in cash, and those were the cleanest deals in the pipeline — no lender approval, no credit stipulations, no funding delays, and the shortest path from signed contract to install. Cash buyers were sold on total cost of ownership rather than monthly payment: no finance charge at all, the full tax credit applied against the purchase, and the fastest payback period of any product we offered.

So the funnel qualified for it directly. The financing selector let the homeowner indicate a cash purchase before an appointment existed, which changed how the rep prepared and which numbers they walked in with. Sorting loan intent from cash intent at the form — rather than forty minutes into a kitchen-table appointment — is one of the highest-leverage things a solar funnel can do.

What DC Taught Us About Funnels

DC is the market that proved the roof-modeling layer was not a nice-to-have. In an easy suburban market you can get away with a generic lead form and sort it out in the field — you will waste appointments, but you will survive. In a market where a large share of addresses are structurally disqualified, the funnel either does that filtering or your reps do it with their windshield time.

We would rather it be the funnel. That principle carried directly into every lead engine PartPixel builds today: the form should know something the salesperson would otherwise have to drive across town to find out.