Google Ads Copy for Lead Gen: RSAs, DKI & Pinning 2026

Google Ads Copy for Lead Gen — Elevarus

Share This Post

Good Google Ads copy for lead generation is not the ad that scores “Excellent” on Google’s meter. It is the ad that says something true, specific, and compliant to a buyer who is ready to call, and that hands the auction enough distinct assets to find the version that actually converts. That job looks different in HVAC, in solar, and in private health insurance, because each vertical hits a compliance wall before it hits a creative one. This guide covers the mechanics that carry across all three, then the copy rules specific to each, including the two things that changed in 2026: the solar federal tax credit went away, and the FTC put a price on deceptive short-term health marketing.

TL;DR

  • A responsive search ad gives you 3 to 15 headlines (30 characters), 2 to 4 descriptions (90 characters), and 2 path fields (15 characters). Cover the buckets; run 2 or more RSAs per ad group.
  • Do not chase “Excellent” Ad Strength. Build to “Good,” then let conversion data drive iteration. Ad Strength is not Quality Score.
  • Dynamic keyword insertion belongs in tightly themed home-services groups and location insertion only. Never in insurance, finance, or on broad match.
  • Pin only what the law requires (a license number, a disclaimer, TPMO or CMS language). Pin the minimum, prefer a description, and pin 2 to 3 variants per slot.
  • Vertical gates: HVAC urgency must be operationally true; the solar 30% federal tax credit is gone for owned systems; U65 health copy must never imply comprehensive coverage.

Quick answers:

We are writing this as an operator, not a product reviewer. Elevarus buys clicks and sells verified calls and leads, so we care about the same thing you should: whether the copy brings in a buyer who converts, and whether the account survives a policy review. Everything below is written to that standard, with sources labeled so you can tell a Google rule from an agency opinion.

How is a responsive search ad actually structured?

A responsive search ad (RSA) is the only search text format Google still builds new. You supply the parts and Google assembles them per auction. The parts are fixed: 3 to 15 headlines at 30 characters each, 2 to 4 descriptions at 90 characters each, and 2 optional path fields at 15 characters each (Google Ads Help, Official). Google’s own advice is to write as many genuinely distinct headlines as you can, to write the first three “as if they will appear together,” and to include at least one keyword.

Anatomy of a Responsive Search Ad
3 to 15
Headlines
30 characters each

2 to 4
Descriptions
90 characters each

2
Path fields
15 characters each

The seven headline buckets
Keyword / serviceBenefitsDifferentiatorsCall to actionOffer / urgencyTrust / proofBrand
Write the first three headlines as if they will show together, include at least one keyword, and run two or more RSAs per ad group so the auction has real variety to test.
RSA structure per Google Ads Help (Official). Character limits and the two-plus-RSA guidance are Google’s own; the seven buckets are practitioner consensus.

The practitioner move is to think in buckets, not slots. Cover keyword or service, benefit, differentiator, call to action, offer or urgency, trust or proof, and brand, then avoid repeating the same idea across headlines. Descriptions do a slightly different job: front-load the most important words in the first 15 to 20 characters, expand the offer, and end on a working call to action that will not get cut off. For lead gen that usually means a benefit, a proof point, and a clear “Get a free quote today.”

Operator Note: Run at least two RSAs per ad group. Google’s aggregate data shows going from one RSA to two lifts conversions, and a second ad gives you a real read on which angle the auction prefers before you commit budget. One well-built ad is not a test.

One format detail that quietly matters: Optmyzr found sentence case tends to outperform Title Case and ALL CAPS on the primary metrics (agency study, treat as directional). It costs nothing to standardize on it.

Does “Excellent” Ad Strength actually mean better performance?

This is where a lot of lead-gen budget gets misdirected. Ad Strength is the colored meter next to your ad, from Poor to Excellent. Google says moving from Poor to Excellent yields about 15% more conversions on average, and is clear that Ad Strength does not affect ad eligibility and does not measure landing-page quality (Google Ads Help, Official).

Does ‘Excellent’ Ad Strength Mean Better Results?
What Google says
Poor to Excellent = about 15% more conversions
Google’s own correlational internal data. Google also states Ad Strength does not affect ad eligibility and does not measure landing page quality.

What the largest study found
No clear correlation with performance
Optmyzr’s analysis of 1M+ ads found ‘Average’ Ad Strength often posted the best CPA and conversion rate and ‘Excellent’ the worst. Headline length and case predicted results more than the label. (Agency study, treat as directional.)

Operator takeaway: build to at least ‘Good’ for eligibility and visibility, then let conversion data drive iteration. Ad Strength is a creative diagnostic, not Quality Score, and not proof your copy converts.
Left: Google Ads Help, Ad Strength (Official). Right: Optmyzr Ad Strength study (agency; 2024 data, reaffirmed in later analyses) – labeled as an estimate.

Here is the caveat Google does not put on the meter. The largest independent analysis available, Optmyzr’s study of more than a million ads, found no clear correlation between Ad Strength and performance. “Average” Ad Strength ads often posted the best cost per acquisition and conversion rate, and “Excellent” the worst. Headline length and case predicted results more reliably than the label did. The study also found the meter rewards shorter, brand-agnostic copy, which means chasing “Excellent” can quietly push you toward generic messaging that converts worse for a differentiated offer. That is an agency dataset, so treat the exact figures as directional, but the direction is consistent enough to act on.

Key Concept: Ad Strength is not Quality Score. Quality Score (expected click-through rate, ad relevance, and landing-page experience) affects what you pay and where you rank. Ad Strength is a creative-completeness meter that affects neither. If you want the lever that moves cost, read our Google Ads Quality Score guide; Ad Strength is a diagnostic, not a scoreboard.

The practical rule: build every RSA to at least “Good” so it is eligible and fully assembled, then stop optimizing for the meter and start optimizing for conversions. The moment “improving Ad Strength” means deleting your sharpest differentiator, you are optimizing the wrong number.

Dynamic keyword insertion: when should you use it, and when should you never?

Dynamic keyword insertion (DKI) writes the triggering keyword into your ad automatically. The syntax is {KeyWord:default text}, and Google inserts the keyword that triggered the ad, not the user’s search term, falling back to your default text when the keyword will not fit. The compliance catch is explicit in Google’s docs: you are responsible for the ad complying with policy after the insertion happens, which means DKI can generate a policy-violating ad out of text you never wrote (Google Ads Help, Official).

Dynamic Keyword Insertion: Use It or Skip It
Syntax {KeyWord:default text}. It inserts the keyword that triggered the ad, not the user’s search term, and you own compliance after the insertion.
Green light
Reasonable to use
  • Tightly themed exact or phrase ad groups where the keywords share grammar
  • Location insertion {LOCATION(City):Your Area} for multi-city service businesses (the safer, more endorsed pattern)
Red flag
Do not use
  • Broad match keywords (nonsensical insertions)
  • Loosely themed or generic buckets
  • Misspellings and close variants
  • Competitor or trademarked terms (trademark risk)
  • Regulated verticals: health, finance, insurance

Mechanics from Google Ads Help, keyword insertion and location insertion (Official). When-not guidance is practitioner consensus; no insurance source endorses DKI.

There is a narrow band where DKI earns its place: tightly themed exact or phrase ad groups where the keywords share the same grammar, so the inserted term always reads cleanly. The more consistently endorsed pattern is location insertion, {LOCATION(City):Your Area}, for a multi-city service business. That is the version home-services agencies actually recommend, and it is far safer than keyword insertion because a city name rarely breaks a sentence.

Everywhere else, the 2025 to 2026 consensus is to skip it. On broad match, DKI can insert a keyword that has nothing to do with the search and read as nonsense. In loosely themed groups it amplifies whatever junk is in the group. It mangles misspellings and close variants. And in regulated verticals it is a live compliance hazard: inserting a competitor’s trademarked name is a trademark violation whether a human typed it or an algorithm did, and no insurance-specific source in our research endorsed DKI at all. Pull branded and competitor keywords into their own ad group with no insertion.

Operator Note: The 2026 direction is Google’s own AI doing the situational rewriting. AI Max for Search (generally available since April 2026) can customize text and assemble assets from your landing pages, and Google is auto-upgrading broad match and Automatically Created Assets (Google’s own “ACA,” which is Automatically Created Assets, not the Affordable Care Act) into AI Max in late 2026. The clean setup is to give the machine well-structured, unpinned RSA inputs and let text customization handle the situational version, treating DKI as a fallback if you opt out of automatically created assets. If you run broad match, walk through our AI Max steering controls guide before the auto-upgrade lands.

Pinning: when is it actually necessary?

Pinning locks an asset to a fixed position: headline 1, 2, or 3, or description 1 or 2. Assets pinned to H1, H2, or D1 always show. If you pin every position, only pinned assets can appear. Google’s stance is blunt: pinning is not recommended for most advertisers and can lower Ad Strength, because it strips the auction’s ability to test combinations (Google Ads Help, Official).

Pinning: Pin the Minimum the Law Requires
  1. Pin only what a law or policy mandates in every ad: a license number, a required disclaimer, TPMO or CMS language.
  2. Prefer a description or headline position 2 or 3 over position 1. Position 1 carries the most algorithmic weight, so a pinned line there costs the most testing.
  3. Pin two to three compliant variants to the same slot, not one static line, so the auction keeps some room to test.
  4. Leave every non-mandated asset unpinned. Pinning limits testing and can lower Ad Strength, so it is a compliance lever, not a performance one.
As Google’s AI takes over ad assembly, pinning is one of the few manual brand-safety and compliance levers a regulated advertiser has left. Use it deliberately, and only for the minimum.
Google advises against pinning for most advertisers (Official). The regulated-disclosure use case and the two-to-three-variants mitigation are Google’s own guidance plus practitioner consensus.

So why does anyone pin? Because sometimes a line legally has to appear in every single ad, and pinning is the only way to guarantee it. That is the entire legitimate use case: regulated disclosures. A license number, a required disclaimer, TPMO or CMS language. Agencies have quantified the cost of over-pinning (one estimate puts the testing loss around 75%, and a single case study reported an unpinned variant beating a pinned one by a wide margin), but those are agency figures, so treat them as directional. The consensus that survives the disagreement is simple: pin only what compliance mandates, and nothing else.

When you do pin, three rules keep the damage down. Prefer a description or position 2 or 3 over position 1, since position 1 carries the most weight and costs the most testing. Pin two or three compliant variants to the slot instead of one static line, which is Google’s own mitigation. And leave every other asset unpinned.

Key Concept: As Google’s AI takes over ad assembly, pinning becomes one of the few manual brand-safety levers a regulated advertiser has left. That makes it more valuable, not less, for insurance and licensed trades, precisely because it is the one place you can force a legally required line to show. Use it deliberately.

Which ad assets earn their place in a lead-gen campaign?

Assets (formerly extensions) are where a lead-gen ad does most of its persuading, and both of our call-heavy verticals underuse them. Add them deliberately.

The Lead-Gen Asset Stack
Sitelinks
25 characters. Add 6 or more with descriptions; it boosts Ad Strength.

Structured snippets
3+ values per header (4+ is better). An ‘Insurance coverage’ header exists.

Callouts
25 characters, up to 10 can show. Specific beats vague.

Call assets
Set ‘Calls from ads’ as a primary conversion. Schedule to staffed hours; use a minimum call length.

Lead forms
Up to 10 questions at 60 characters. Requires a privacy policy; pre-qualify to cut junk.

Price / promotion
Real, dated offers only. Promotions require start and end dates and self-retire.

Deadline: Google stopped accepting new call-only ads in February 2026, and existing ones stop serving in February 2027. Both lead-gen verticals are call-heavy, so migrate to an RSA plus call assets now.
Asset mechanics and the call-only sunset from Google Ads Help and Google Ads policy (Official).

Sitelinks are the highest-leverage add: 25 characters each, and Google explicitly says adding six or more with descriptions improves Ad Strength and unlocks better formats. Structure them by ad group intent, not generically. Structured snippets list values under a predefined header (there is even an “Insurance coverage” header); use three or more values, four is better, and match the header to the values or the snippet gets disapproved. Callouts are short trust and specificity signals, 25 characters, up to ten can show; “Licensed and Insured” and “500+ 5-Star Reviews” beat “Great Service.”

Call assets are the ones that matter most here, because in both verticals the money is in the phone. Set “Calls from ads” as a primary conversion so Smart Bidding optimizes toward calls, schedule the number so it only shows when someone can answer, and count a call as a lead only past a minimum duration (30 to 60 seconds) so misdials do not train your bidding on garbage. Lead form assets can pull a name and number straight from the SERP, but that frictionless quality cuts both ways: add a screening question and route the lead to your CRM fast. Price and promotion assets work only for real, dated offers; promotions require start and end dates and self-retire, so a year-round “Spring Special” just reads as a lie.

Key Stat: Call-only ads are being retired. Google stopped accepting new call-only ads in February 2026, and existing call-only ads stop serving in February 2027 (Google, Official). Migrate every call-only campaign to an RSA plus call assets now, before the deadline forces a rushed rebuild.

One more current-name detail: the trust badge to reference is Google Verified, which replaced Google Guaranteed and Google Screened on October 20, 2025 (the money-back guarantee was discontinued). Use the current name in copy and on the landing page.

HVAC and home services: how should you structure the copy?

The single biggest structural mistake in HVAC lead generation and home-services search is one blended campaign. Emergency, installation, and maintenance are three different buyers with three different timelines and values, and copy written for the average of them converts none of them well. Split ad groups (and often campaigns) by intent.

The Copy Matrix by Vertical
Vertical Intent split Compliant hero angle #1 compliance gate Key asset
HVAC / home services Emergency vs install vs maintenance Response time + local + trust: ‘Same-Day AC Repair in [City]’ Urgency claims must be operationally true (staffed hours) Call assets, scheduled to staffed hours
Solar Quote / consult vs research Potential savings, free estimate: ‘See what solar could save you’ No federal tax-credit hook for owned systems; no ‘free solar’ or guaranteed savings Lead form with a homeownership pre-qualifier
Private / U65 health Quote-ready vs comparing plans Generic and true: ‘Get a Free Quote’, ‘Licensed Independent Agent’ G2RS certification; never imply comprehensive or ACA-equivalent coverage Pinned disclosure description (‘not a substitute for major medical’)
Compiled from the vertical sections below. Compliance gates trace to Google policy, the FTC, and the IRS (Official); tactics are practitioner consensus.

Emergency copy leads with response time, urgency, and the phone: “Same-Day AC Repair in [City], Call Now.” Installation and replacement copy leads with differentiation, financing, a free in-home estimate, and social proof: “0% Financing,” “Free Estimate,” “500+ 5-Star Reviews.” Maintenance copy leads with scheduling and seasonal timing: “Book Your Furnace Tune-Up Before the First Freeze.” Then match the ad to the landing page headline exactly; a “24/7 Emergency AC Repair” ad that lands on a generic homepage kills the conversion.

Trust language is the backbone: licensed, bonded, and insured, years in business, review counts, manufacturer certifications, and the Google Verified badge. Put the city in headline 2 or description 2, not just the landing page. But two guardrails are non-negotiable. Urgency claims like “24/7” and “Same-Day” have to be operationally true, because Google’s False or Misleading Claims policy treats an unstaffed “24/7” as a violation, and it is also just a bad customer experience; only run those claims and call assets during hours you can genuinely staff. And tighten geo-targeting to “Presence,” not “Presence or interest,” since audits routinely find 20% to 30% of spend leaking to areas the business cannot even serve (agency estimate). For the account architecture behind this, see our guides on HVAC lead generation and running HVAC Google Ads yourself versus buying verified leads.

Solar: what can you legally say now that the tax credit is gone?

Start here, before any solar copy ships: the 30% federal residential solar tax credit is gone for owned systems. The Residential Clean Energy Credit under IRC Section 25D “is not available for any property placed in service after December 31, 2025,” per the IRS, following the One Big Beautiful Bill Act signed in July 2025. That means a headline like “Claim Your 30% Federal Tax Credit” for a cash or loan system is factually false in 2026 and a live substantiation risk under Google’s misrepresentation policy and the FTC. Retire that hook. (Leases and power purchase agreements may route a different business-side credit that the leasing company claims; do not assert those mechanics or a hard date in copy without a fresh check.)

So what is left to say in solar lead generation, and it is plenty, has to be substantiation-safe. Frame savings as potential, never guaranteed: “See What Solar Could Save You” or “Free Savings Estimate,” never “Eliminate Your Bill,” “Free Solar,” or “Guaranteed Savings.” Treat “$0 down” as what it is, financing, and disclose the loan, lease, or PPA rather than implying free ownership. Keep incentive claims state and utility specific, eligibility-framed, and confined to the states where they are actually true, and never imply government affiliation or endorsement. Roof suitability is a consult, not a promise. Solar is not a Google-certified vertical the way health is, but the Misrepresentation and Unrealistic Claims policies apply in full, the FTC actively enforces solar savings claims and treats “free solar” as deceptive, and TCPA class-action risk in solar is real.

Operator Note: Lead quality is a backend job, not an ad-copy hook. Solar carries heavy lead fraud (fake homeownership and address fraud; practitioner estimates run 25% to 35%), so the operator move is to verify before a lead is billed: OTP and consent verification, bot and click-farm detection, and homeownership and address validation. That is how Elevarus runs the campaign, not something you write into a headline. For the buyer-facing version, see why every unqualified solar click costs more now and our solar lead generation page.

Private and U65 health: what copy stays compliant?

Private, under-65 health means off-exchange alternatives: short-term medical, fixed or hospital indemnity, and limited-benefit plans. It is not Medicare, and it is not ACA on-exchange coverage, and it should not be framed around the ACA enrollment calendar. Keep those three separate in your account and your copy; they carry different rules. (For the on-exchange side, we cover why ACA and off-exchange are two different Google Ads accounts.)

Certification comes first: Google requires G2 Risk Solutions certification to advertise individual health, short-term or limited-duration, and fixed-indemnity products, and only licensed exchanges, first-party insurers, and licensed brokers qualify. A separate ACA-plan certificate is required only for on-exchange qualified health plans; pure short-term or indemnity does not need it. Verify your short-term and indemnity framing with G2RS at onboarding, because their public page does not itemize it.

Then there is the list of things to never write, and it now has a price tag. In August 2025 the FTC settled with Assurance IQ and MediaAlpha for $145 million total over deceptive short-term and indemnity marketing. Based on that action: never claim the plan is equivalent to comprehensive coverage, has no caps on benefits, or covers pre-existing conditions when it does not; never imply a partnership or network with a major carrier without a basis; and never use government-sounding names, “$1 a day,” or a fake “Health Insurance Give Back Program.” The FTC’s bar is “competent and reliable evidence” for any coverage claim, and it treats the whole funnel, the lead-gen creative and the landing page, as in scope. That same case anchors the enforcement pattern we break down in our Limited Ad Serving policy guide.

Disclosure is required, too: HHS and CMS require a “not a substitute for major medical coverage” notice on fixed-indemnity and short-term materials. It is technically the insurer’s obligation, but carry a plain-language version through the ad description and the landing page. One thing not to imply: the federal individual-mandate penalty is $0 (only a handful of states impose their own), so do not suggest a universal tax penalty.

That leaves a compliant, still-compelling toolkit. Headlines stay generic and true: “Get a Free Quote,” “Compare Private Health Plans,” “Licensed Independent Agent,” “Short-Term Medical Options.” Push license, NPN, and disclosure into descriptions and the landing page, using a pinned description where a disclaimer legally must show. You can bid on competitor keywords, but keep carrier and competitor brand names out of the ad text (trademark policy), and never insert a competitor name with DKI.

Operator Note: Medicare is its own regime, keep it distinct and brief. Every agent is a TPMO, the CMS TPMO disclaimer must appear on marketing materials, and Social Security Act Section 1140 bans government-affiliation language, the CMS logo, and titles like “Medicare Expert.” Naming a carrier in a comparison needs that carrier’s approval plus a CMS filing, so Medicare comparison copy stays fully generic (“Compare Medicare Options”). Do not let U65, ACA, and Medicare copy bleed into each other.

The operator’s pre-launch checklist

Before any lead-gen ad goes live, run it against the same short list we do:

  1. Two or more RSAs per ad group, each covering the buckets, with the first three headlines readable together and a keyword present.
  2. Ad Strength at “Good” or better, but the winner chosen by conversions, not the meter.
  3. DKI used only for location insertion or a tightly themed exact or phrase group, and never in health, finance, or on broad match.
  4. Pins limited to legally required disclosures, on a description or position 2 or 3, with two or three variants per slot.
  5. Six or more sitelinks with descriptions, structured snippets and callouts populated, and call assets set as a primary conversion with staffed-hours scheduling and a minimum call length.
  6. Every claim substantiated: HVAC urgency operationally true, no dead solar tax-credit hook, no comprehensive-coverage implication in U65 health.
  7. Geo set to “Presence,” message match from ad to landing page, and the required disclosures carried through to the page.

Copy is only half of a lead. The other half is what happens after the click: whether the call is a real buyer or a bot, whether the form is a homeowner or a fraud, whether the lead reaches you fast enough to close. That verification layer — ad fraud and OTP verification in lead generation — is the part of the funnel most SERP advice ignores, and it is the part Elevarus runs. See our lead generation model and what a media buying operator actually does for how the copy and the verification fit together.

Frequently Asked Questions

Should I fill all 15 headlines in a responsive search ad?

Not just to hit the number. Google lets you add up to 15 headlines and advises writing as many genuinely distinct ones as you can, but the largest independent analysis (Optmyzr, an agency study) found that distinct, well-crafted copy beats maxing out character count. Write enough unique headlines to cover the buckets that matter (keyword, benefit, differentiator, CTA, offer, trust, brand), then let conversion data trim the losers.

Is a higher Ad Strength score worth chasing?

Build to at least ‘Good’ so your ad is eligible and visible, but do not chase ‘Excellent’ at the cost of differentiated copy. Google says moving from Poor to Excellent yields about 15% more conversions on average, but that is Google’s own correlational data, and the largest independent study found no clear correlation, with ‘Average’ Ad Strength often posting the best CPA and conversion rate. Ad Strength is a creative diagnostic, not Quality Score. Let conversions decide.

Can I use dynamic keyword insertion in insurance or solar ads?

It is not recommended. No insurance-specific source endorses DKI, and the trademark and misleading-claim risk in regulated verticals outweighs the click-through benefit. DKI inserts the triggering keyword and leaves you responsible for compliance after the insertion, which is exactly the control a regulated advertiser cannot give up. Hand-write health and solar headlines; reserve insertion for location insertion in tightly themed home-services groups.

When is pinning actually necessary?

When a law or platform policy requires a specific disclosure to appear in every ad: a license number, a required disclaimer, or TPMO and CMS language. Pin the minimum, prefer a description or headline position 2 or 3 over position 1, and pin two or three compliant variants per slot so the auction keeps some room to test. Leave everything else unpinned, because pinning limits testing and can lower Ad Strength.

Can I still advertise the 30% federal solar tax credit in 2026?

No, not for owned (cash or loan) systems. The federal Residential Clean Energy Credit under IRC Section 25D is not available for property placed in service after December 31, 2025, per the IRS, following the One Big Beautiful Bill Act signed in July 2025. Advertising a 30% federal credit for an owned system now is a live substantiation risk under Google’s misrepresentation policy and the FTC. Frame savings as potential, disclose financing, and confine state or utility incentive claims to where they are actually true.

What can I never say in a short-term or indemnity health ad?

In August 2025 the FTC settled with Assurance IQ and MediaAlpha for $145 million total over deceptive short-term and indemnity marketing. Based on that action, never claim the plan is equivalent to comprehensive coverage, has no caps on benefits, or covers pre-existing conditions when it does not; never imply a partnership with a major carrier without a basis; and never use government-sounding names, ‘$1 a day’, or a fake ‘Health Insurance Give Back Program’. Keep headlines generic and true, and carry the required ‘not a substitute for major medical coverage’ notice through the description and landing page.

What happens to my call-only ads?

They are being retired. Google stopped accepting new call-only ads in February 2026, and existing call-only ads stop serving in February 2027. Migrate to a responsive search ad with call assets: set ‘Calls from ads’ as a primary conversion, schedule the number to hours you can genuinely staff, and count a call as a lead only past a minimum duration so ad spend chases real conversations, not misdials.

Sources



Work with Elevarus

Are You Ready to Grow With a Proven Lead Generation & Performance Marketing Agency?

Get a free, no-pressure strategy call with our lead-generation team. We'll map the fastest path to more qualified leads for your business.

Book a free call →

Ready to put this into action?

Picture of <a href="https://elevarus.com/shane-mcintyre/">SHANE MCINTYRE</a>

Founder and CEO of Elevarus, specializing in paid media, lead generation, pay-per-call, and customer acquisition.