Google Ads for Solar Companies

Solar paid search rewards a discipline that feels wrong to most advertisers, which is deliberately making your ads less appealing to most of the people who see them. Reported click costs for solar terms run high, with published figures placing home improvement among the most expensive categories and competitive solar markets well above that. Meanwhile vendor estimates suggest half or more of cheap, broadly targeted solar contacts are renters, unsuitable roofs, or people who will not buy for years. Those two facts together mean the highest-return work in a solar account is not persuasion. It is exclusion: naming the homeowner requirement in the ad, asking a bill amount before anything else, and telling the bidding algorithm that a raw form fill is not a conversion.

1Qualify in the Ad, Not the Landing Page

• Name the homeowner requirement in the copy. A renter who reads that the offer is for homeowners does not click, which saves you the click rather than the sales call.
• State the service area explicitly. Geographic targeting is imperfect and naming the region reinforces it at no cost.
• Reference a bill threshold where it fits. Signalling that the economics suit larger bills filters people the proposal would disappoint anyway.
• Say what the appointment involves. A consultation with a proposal, rather than a vague free quote, sets expectations and reduces no-shows.
• Do not chase click-through rate. A lower rate with better qualification is the correct outcome and it will look like underperformance on a default report.
• Put the hard questions on the form. Ownership, roof age, and bill amount, asked before submission rather than discovered later.
• Accept fewer leads deliberately. This is the strategy, and it needs agreeing with the client or the first month reads as failure.

2Stop Optimizing Toward Raw Leads

This single configuration decision does more damage in solar accounts than anything else, because the algorithm does exactly what it is told.

• A form fill is not a conversion in solar. Reported no-show rates on solar appointments run thirty to forty percent even when the appointment was purchased.
• Tell the algorithm what a good outcome is. A held appointment, or better, a signed contract, imported as an offline conversion.
• Otherwise it optimizes toward cheap. Which in solar means renters, small bills, and browsers, because those people fill in forms readily.
• Use a staged conversion model. Form fill as a secondary observation, qualified appointment as primary, signed contract with value where possible.
• Accept a slower learning period. Down-funnel conversions are less frequent, so the account needs patience and probably manual control early.
• Assign values by ownership model. A cash sale and a lease produce different economics and bidding should reflect that.
• This requires the CRM to cooperate. Which is usually the real project and the most valuable thing you will fix.

3Separating Research From Ready to Buy

• Research traffic is real demand, badly timed. Route it to content and nurture rather than to a sales calendar.
• Battery and resiliency interest behaves differently. It is frequently driven by outage experience and converts on different messaging.
• Commercial deserves complete separation. Different buyer, different economics, different sales cycle entirely.

4Keyword Strategy After the Credit

• Remove expired federal incentive terms. Bidding on searches for a credit that no longer applies buys traffic you cannot satisfy and a conversation that starts with disappointment.
• Shift weight to state and utility programmes. Where incentives still exist, those terms now carry the demand the federal credit used to.
• Lease and power purchase language matters more. As the market shifts toward third-party ownership, buyers search differently.
• Utility rate terms are underused. Homeowners searching about rate increases or time-of-use changes are motivated and rarely targeted.
• Net metering terms are high intent locally. Especially where rules changed recently.
• Battery and backup terms are growing. Frequently the entry point for a full system conversation.
• Keep terms current with policy. This keyword list needs review on a schedule, not at launch.

5Negatives That Save the Account

The free and grant category deserves particular attention in solar. Advertising in this space has historically leaned on programme language that attracts people expecting no-cost installation, and that traffic converts at close to zero while costing full price.

6Local Service Ads for Solar

• Reported costs are materially better. Published 2026 figures put solar Local Service Ads cost per lead at roughly half of blended Google Ads, with a substantially lower cost per closed customer.
• The badge matters more here than anywhere. A verified screening signal carries real weight for a high-ticket purchase in an industry with a trust problem.
• Verification excludes some competitors. Which is part of the value.
• Reviews drive placement. Connecting this channel directly to your review programme.
• Dispute unqualified leads. Wrong service and outside area are creditable and worth pursuing at these values.
• Volume is capped. Which is why search still runs above it rather than instead of it.
• Fund it first. Given the reported cost differential, this is where the first budget dollar belongs.

7Ad Copy and Savings Claims

• Savings and payback claims attract scrutiny. Advertising representations about financial outcomes in solar have drawn regulatory attention, and this is a matter for your own counsel before publication.
• Avoid specific savings figures in ad copy. Individual results depend on consumption, rates, roof, and system size, none of which are knowable at the ad.
• Do not describe systems as free or no cost. Even where a financing structure produces no upfront payment, that framing is a well-known source of complaints and regulatory interest.
• Never reference an incentive that has expired. Including in older ads and extensions nobody has reviewed.
• Lead with credibility instead. Years installing, local presence, licensing, and warranty support are factual and differentiating.
• Use qualification as a positioning asset. Copy that says you assess whether solar makes sense reads as honest and filters simultaneously.
• Review all copy on a policy schedule. Because incentive language ages badly and quickly.

8Tracking to Installs

• Solar has more stages than any comparable purchase. Lead, contact, held appointment, proposal, signed contract, permit, install, and permission to operate.
• Track drop-off at every stage. Because a channel that produces appointments and no contracts is a different problem from one producing no appointments.
• Import held appointments and signed contracts. As offline conversions, which is what makes bidding useful rather than harmful.
• Record ownership model on every deal. Cash, loan, lease, and power purchase produce different revenue and should carry different values.
• Use long attribution windows. Solar decisions run weeks to months and a default window attributes almost nothing correctly.
• Separate cancellation from close. Solar has meaningful post-signature cancellation, and counting signatures as revenue overstates performance.
• Reconcile with the CRM monthly. Platform conversions and installed systems diverge substantially in this vertical.

9Measuring Solar Google Ads

• Cost per closed install. The only number that matters, with reported industry figures spanning a wide range depending on lead source.
• Cost per held appointment. The working metric, since installs lag by months.
• Appointment hold rate by campaign. Where lead quality actually reveals itself.
• Disqualification rate and reason. Renters, roofs, bills, and geography, tracked so targeting can respond.
• Proposal to signature rate by source. Which separates media problems from sales problems.
• Ownership model mix by campaign. Since revenue and margin differ substantially.
• Cancellation rate by source. Some channels produce signatures that do not survive.
• Search terms weekly. Given how much adjacent and expired-incentive traffic exists in this category.