For solar installers
The real split is not cheap versus expensive. It is channels somebody else controls and can switch off, against channels you control and cannot switch on quickly. You need one of each, and which one you start with depends on how much runway you have.
We sell a rented channel. You should read the owned column first anyway.
You own it
Referrals, content and your own canvassing team. Nobody can raise your price or cut your volume, and the cost per signed job is a fraction of anything you can buy. What you give up is responsiveness: none of these produce a customer next week because you decided they should.
Past-customer referrals — $385 per job
Slow to start, permanent once it does.
Your own content and SEO — $750 per job
Six to eighteen months before it is worth anything, then it compounds.
Canvassing and door-to-door — $1,857 per job
Immediate, and it stops the day you stop paying for it.
Home shows and events — $2,444 per job
Bursty — nothing, then a weekend's worth at once.
Somebody else owns it
Bought leads and paid media. Spend maps onto opportunities within days, volume moves with a dial, and you can stop entirely on a Monday. What you give up is control: the price, the quality and the existence of the channel are all somebody else's decisions.
Exclusive leads — $2,125 per jobOurs
Bounded by real demand, which means it cannot be scaled to fill a sales team on its own.
Paid social — $2,444 per job
High on volume, low on quality — the constraint is how many poor-fit leads your team can absorb.
Shared or resold leads — $3,000 per job
Effectively unlimited supply, which is itself the warning.
Paid search — $3,455 per job
Capped by search volume in your territory, and the cap arrives sooner than most people expect in a mid-sized metro.
Cost per signed job is the midpoint cost per lead divided by the midpoint close rate for each channel, on benchmark ranges from the US residential solar market. Owned channels are costed at steady state and therefore flattered — see the setup cost below.
The part spreadsheets miss
Every cost-per-lead comparison you will read compares channels at steady state. That is the wrong comparison for anyone deciding what to do this quarter, because the owned channels spend months producing nothing first, and that period is paid for out of the same account.
| Channel | Time to first customer | At steady state | What actually decides it |
|---|---|---|---|
| Past-customer referralsOwned | Slow to start, permanent once it does. | $385 | Whether your install and service experience is good enough that people volunteer your name. |
| Your own content and SEOOwned | Six to eighteen months before it is worth anything, then it compounds. | $750 | Whether you will still be publishing in a year. Almost nobody is, which is why the channel stays available. |
| Canvassing and door-to-doorOwned | Immediate, and it stops the day you stop paying for it. | $1,857 | Recruiting and retention. Rep churn, not lead cost, is what makes this channel expensive. |
| Home shows and eventsOwned | Bursty — nothing, then a weekend's worth at once. | $2,444 | Follow-up discipline in the week after the event, which is where most of these leads are lost. |
| Exclusive leadsRented | Immediate, and volume is capped by search demand in your territory rather than by your budget. | $2,125 | Speed to first call, and whether the lead arrived with enough context to make the first call worth having. |
| Paid socialRented | Immediate, and volume is easy to increase. | $2,444 | Speed to first call, more than anything else. Interest-based leads go cold fast because they were not looking for you. |
| Shared or resold leadsRented | Immediate, and available in volume in most territories. | $3,000 | Whether you can call within seconds. In a shared lead the first caller takes most of the value, and there is no second prize. |
| Paid searchRented | Immediate, and it stops the day you stop paying for it. | $3,455 | Whether you can run the account well. The spread between a good and a bad solar account is several times the gap between channels. |
A workable sequence
Until you can state leads-to-signed-contract, every price comparison on this site and everywhere else is unusable. Buy the smallest tranche of leads anyone will sell you, call every one within minutes, and count. At a 1.5% close rate a $85 lead breaks even; you need to know which side of that you are on.
Set the cap at what your team can genuinely answer within minutes, not at what you can afford. The most common reason installers conclude bought leads do not work is buying more than they can call, then reading the resulting close rate as a quality problem.
Referrals need an ask built into your handover process, not a hope. Content needs a publishing rhythm you will still be keeping in a year. Both are nearly free per customer and neither will exist in eighteen months unless somebody begins this month.
As referrals and content mature, buy fewer leads — but do not switch the bought channel off. Its value is that it is there in a quiet quarter, and a channel you have not run for a year is not a channel you can turn on in a week: the account is cold, the close rate is unmeasured, and the vendor has given your territory to someone else.
Where we fit in that sequence
Exclusive leads at $85 from homeowners who ran a full estimate and asked for quotes — capped at what you can answer, with no contract and nothing to cancel. The first 5 are free precisely so you can measure your close rate on real leads before spending anything, which is step one.
We would rather be a channel you keep running at a modest cap for years than one you scale hard for two quarters and resent. On $5,800 of gross profit per job, that modest version is the one that keeps working.
Neither, on its own. Self-generated channels are cheaper per customer and cannot be turned up when you need them; bought channels are more expensive and are the only ones that respond to a decision you make today. A company running only bought leads has no floor if the vendor changes price or quality. A company running only self-generated leads cannot fill a quiet quarter. The working answer for most installers is to cover current capacity with a bought channel while building referrals and content underneath it, and to shift the mix towards owned as the owned channels mature.
On a fully loaded basis, usually yes — and the gap is smaller than it looks on a spreadsheet. Past-customer referrals works out around $385 per signed job against roughly $2,125 for the cheapest bought channel, but that figure excludes the months of salary before the channel produced anything. Cost the setup period honestly and the first year rarely beats buying; the second and third years do, which is exactly why it is worth starting and why so few installers finish.
Canvassing produces conversations the first week and takes three to six months to produce a team that stays. Content and SEO take six to eighteen months to produce anything and then compound. Referrals take as long as your first cohort of installs takes to be happy, which means a year at minimum and forever after that. None of the three can be accelerated by spending more this month, which is the whole argument for having a bought channel running alongside them.
Compare fully loaded cost per signed job, not cost per knock. A canvassing rep is a salary, a manager, a training pipeline and a churn rate — most residential solar canvassing programmes lose reps faster than they produce reliable numbers, and the programmes that work are run by people who have run one before. If you have not, buying a capped number of leads is a far cheaper way to find out what your close rate actually is, and you will need that number before you can evaluate a canvassing programme anyway.
Yes, and the mechanism is a cap rather than a budget. Set the bought channel to a volume your team can call within minutes, and treat it as the variable that absorbs the gap between your owned channels and your capacity. When the owned channels grow, buy less. When a quarter is quiet, buy more. The failure mode is buying more leads than you can answer, which produces a low close rate that then gets blamed on lead quality.
Higher than bought, for reasons that are mostly about the conversation rather than the person. A referral closes at 20–45% because someone vouched for you; a content lead closes at 4–12% because they chose you; a bought exclusive lead closes at 2–6% because they asked for quotes generally and you are one of the answers. Do not compare those numbers directly to judge lead quality — compare cost per signed contract, which is the only figure that accounts for the difference.
Every way a residential solar company gets customers, what each one costs per signed job, and which to start with.
Put in your close rate, ticket and margin. Get cost per install, the close rate you break even at, and the most a lead is worth to you.
What each type of solar lead sells for, why the cheap ones are usually the expensive ones, and what you should be willing to pay.
Every state ranked on the two things published data can actually show: how easily solar pays back for the homeowner, and how much gross profit a job carries.