Why Solar Projects Stall After the Sale, and Where the Time Actually Goes

Ask a residential PV solar company where its projects lose time and you will usually get an answer about crews, weather or the utility. Then look at a cohort of actual projects, stage by stage, and a different picture appears: the installation was one or two days, and everything else, often the great majority of the elapsed calendar, was paperwork, waiting on somebody's review, or waiting on somebody to notice that a project needed a decision.

This matters commercially, not just operationally. A project that sits is a customer losing confidence, a commission not yet earned, and, because milestone funding is normally tied to install and activation rather than signature, cash already committed and not yet collected. Cycle time is a cash-flow number wearing an operations costume.

Here is where the weeks actually go, and which of them are yours.

The install is not the long pole

Start by separating the two things that get conflated when people say a project is "in progress."

Construction time is the days a crew is physically on site. On a straightforward residential PV solar retrofit that is short, and it is the part of the process most companies already run well.

Everything else is design, structural and electrical engineering, the permit package and its review, the utility interconnection application and its review, funding documentation, material procurement and delivery, crew scheduling, the final inspection, and permission to operate.

The second category is where a project spends nearly all of its life. That is not a criticism of anyone's crews. It is a structural fact about residential solar, and it means efforts to improve cycle time by pushing installers harder are aimed at the smallest available target.

Stage one: an incomplete sale poisons everything downstream

The single cheapest place to fix a delay is the day the contract is signed, and the single most expensive place to discover it is after submission.

A project sold without a complete picture of the site carries that gap forward. The recurring examples are familiar to anyone who has run operations:

• Usage based on an estimate rather than the actual bill. The system gets designed against a number that turns out to be wrong, and the design, or the customer's expectation, has to change later.
• A roof whose age or condition was not properly assessed. This surfaces as a change order, sometimes as a cancelled project, and occasionally as a roofing job that should have been sold alongside the PV solar.
• An electrical panel that cannot accept the interconnection as designed. A main panel upgrade discovered at design time is a scheduling problem. Discovered at install, it is a truck roll, a redesign, a new permit and a new inspection.
• Account and address details that do not match the utility's records. Trivial to fix at the point of sale; a rejected interconnection application weeks later.
• Shading or setback conditions that were not captured. The array shrinks, the production estimate falls, and the financing that was approved against the original numbers may need revisiting.

None of these are exotic. What they share is that the cost of resolving them rises sharply with distance from the sale, and that all of them are decided by the quality of the information collected in the first conversation.

Why this is a systems problem, not a training problem

It is tempting to treat this as reps being careless. Occasionally it is. More often the rep is being asked to collect information the sales process does not actually require them to collect, on a call optimised for closing rather than for capturing a complete project. If a field is optional, under pressure it will be left blank, and the consequence will land on somebody else weeks later, which is precisely the structure that lets it keep happening.

The fix is to make the sale produce a submittable project, not merely a signature. Everything on the list above is confirmed or contradicted at the site survey, and a deal that fails it does not reach notice to proceed, which is the gate a pipeline actually stops at, whatever the closed-deal count says.

Stage two: design and the rework loop

Once a project reaches design, the number that governs your cycle time is not how fast a designer works. It is what proportion of your submissions come back with corrections.

This is worth being precise about, because the arithmetic is not intuitive. A correction notice does not cost you the hour it takes to fix the drawing. It costs you:

1. the time before anyone notices the notice arrived,
2. the time to route it to whoever can act on it,
3. the time to make the correction,
4. and then the full review queue again, from the back.

Step four is usually larger than the other three combined, and it is invisible on any internal timesheet. It is also why two companies submitting into the same jurisdiction, with identical crews and identical software, can have materially different cycle times. The difference is submitter quality, and it compounds.

Stage three: two approval queues, running on somebody else's clock

Every residential PV solar project needs approval from two separate organisations that do not coordinate with each other, ask different questions, and keep separate queues. One is the utility; the other is the authority having jurisdiction, and confusing the two sends people chasing the wrong party for days. If any of the vocabulary in this section is unfamiliar, the solar fulfillment glossary defines it.

Two practical consequences follow, and both are routinely missed.

They are not sequential, and treating them as sequential adds weeks. The interconnection application does not need to wait for the permit. Filing it late is one of the most common self-inflicted delays in the industry, because its review runs on the utility's calendar regardless of where your permit stands. We have written about this side of the process in detail in solar interconnection and permission to operate.

Passing one tells you nothing about the other. A final inspection signed off by the jurisdiction has no bearing on whether the utility has completed its review. Teams that carry "permitting" as a single line on a project schedule discover this after installation, when the remaining work belongs entirely to somebody else.

Stage four: the handoffs between everybody

Here is the failure mode that produces the longest stalls, and it is not a capability problem at all.

In a fragmented project, design sits with one vendor, engineering with another, permitting with a third, the installer is a separate company, and financing runs through a lender portal somebody in the office updates. Each party is competent. Each does its own work correctly. This is exactly the arrangement what 'turnkey' actually means is written to help you test before you rely on it.

Then a correction notice arrives, and lands in the inbox of whichever party submitted the package, which is frequently not the party that can resolve it. Nothing is technically blocked. But the project now needs a conversation between two companies before it can move, and until somebody initiates that conversation, it simply waits.

Multiply by the number of handoffs on a project and you have the real explanation for why a job with no problems can take months. Nobody dropped it. It was never anybody's in particular.

Stage five: scheduling against projects that are not ready

The last stall is self-inflicted in a different way. Under pressure to show a customer a date, a project gets an install slot before it is actually permitted and approved. The date then slips, sometimes more than once, and a paperwork delay the customer would have tolerated becomes a broken promise they will not.

Booking capacity against unready projects also corrupts your own numbers. Crew utilisation looks poor, because slots are held for jobs that cannot proceed, which produces pressure to sell more to fill them, which puts more unready projects into the schedule.

The discipline is unglamorous: a project becomes schedulable when its approvals are in hand, not when a customer asks for a date.

What to do about it, in order

If you want cycle time down, the sequence matters. These are ordered by return, not by effort.

1. Make the sale produce a complete project. Bill, roof, panel, address, shading, required, not optional. Every downstream stage inherits whatever the sale did not capture.
2. Measure first-pass approval rate. Per jurisdiction, per utility. You cannot improve a rework rate you have never counted, and rework is the largest single lever in the middle of the process.
3. File interconnection early and in parallel. Not after the permit. It runs on a queue you do not control, so start it as early as the utility will accept it.
4. Count your handoffs and reduce them. Every boundary between companies is a place a project can wait unowned. Consolidating design, permitting and interconnection under one owner removes several at once.
5. Only schedule approved projects. Protect the customer's date by not making it until you can keep it.
6. Track the gap between install and activation as its own number. It is where cash sits after it has been spent, see our M1/M2 milestone funding explainer for why that gap is a financing question as much as an operations one.

The bottom line

Projects rarely stall because the work is hard. They stall because a sale left something out, a submission came back for correction and lost its place in a queue, an application was filed later than it needed to be, or a correction notice landed with somebody who could not act on it.

All four of those are addressable, and none of them are about installing faster. They are about how much of the middle of the project one party actually owns.

That middle is the whole of what Seamless Home does. Design, engineering, permitting, interconnection, materials and milestone funding run through one accountable counterparty, so a correction is a task rather than a phone call between vendors. If your cycle time is drifting and you cannot see exactly where, get in touch. The diagnosis is usually more specific than people expect.