Call ads deprecated: migrate to RSA + call assets without losing phone demand

Direct answer — what should you do if call ads are deprecated?

Migrate phone-first Search demand to Responsive Search Ads (RSA) with call assets — and keep (or fix) call conversion tracking with a duration threshold. Do not wait for a hard sunset to invent a new account structure. Call-only / call ads were a creative format optimized for one tap-to-call; the modern stack is RSA (or Performance Max surfaces) + call assets + website call tracking so you can still win phone leads while keeping message testing, sitelinks, and landing-page journeys.

This insight owns the platform-change migration: what to turn off, what to build, how to preserve Smart Bidding signal, and how not to double-count. Full phone measurement OS: Google Ads call tracking for service businesses. Coverage across forms/calls/WhatsApp: multi-path lead tracking. Broader rebuild: conversion tracking setup.

Working rule: migrate creative format and measurement in the same sprint. Swapping to RSA without call assets (or with “every dial = conversion”) is how phone CPA looks “broken” for a month.

What “good” looks like after two weeks: call-only campaigns paused or removed; RSA live with call assets and business hours; qualified call conversions still flowing; form path still instrumented; leadership sees phone demand held or recovered — not a mystery traffic cliff.

If you only remember three moves: RSA + call assets, keep duration-filtered call conversions primary where phone closes jobs, validate test calls before raising budgets.

What “call ads deprecated” actually means

Google has been pushing advertisers off dedicated call-only / call ads creatives toward standard Search ads (RSA) with call assets (formerly call extensions) and other phone CTAs. Exact UI labels and sunset dates shift by account and product surface — treat vendor emails and in-product banners as authoritative for your MCC, but treat the operating destination as stable:

| Old pattern | Modern pattern | |---|---| | Call-only ad → phone dial | RSA headlines/descriptions → call asset and/or lander click-to-call | | Campaign exists mainly to get calls | Campaign exists to get qualified demand (calls + forms + later CRM) | | Conversion = any call from call ad | Conversion = answered call over threshold (and/or OCI stage) |

Deprecation is a creative and campaign-type change, not a reason to stop measuring phone. Phone demand did not disappear — the ad type that only existed to dial did.

Related entities: call ads, call-only ads, Responsive Search Ads (RSA), call assets / call extensions, business hours, Google forwarding numbers, website call conversions, dynamic number insertion (DNI), Smart Bidding, Performance Max, primary conversion actions, learning phase.

Out of scope here: full CallRail/DNI build (call tracking guide); WhatsApp CTAs (WhatsApp guide); deep RSA copy systems (separate ad-copy insight when shipped).

Why SMBs clung to call-only (and why that is risky now)

Call-only felt clean: mobile user, one button, phone rings, job books. For US home services and UK trades that was often true.

Risks that accumulated:

1. No landing page = no message match tests, no form backup, no proof stack. When users want a price band or license proof before dialing, call-only under-converts.
2. Measurement debt. Many call-only setups counted every dial with no duration floor — spam trained Maximize Conversions.
3. Account spaghetti. Parallel call-only + RSA campaigns fighting for the same queries without shared negatives or consistent goals.
4. Platform dependency. When Google sunsets the format, accounts with no RSA/call-asset muscle memory lose volume overnight.

Migration is an opportunity to keep phone convenience and add a real lander for users who will not dial cold.

Migration blueprint (one sprint)

1. Inventory what is live today

Export campaigns using call-only / call ads. Note: budgets, bid strategies, geo, keywords/themes, conversion actions marked primary, call reporting volume, and whether website call tracking already exists.

Screenshot call conversion settings (duration, primary/secondary). You will need them after the creative swap.

2. Decide the replacement structure

Default for most service SMBs: keep (or create) Search RSA campaigns on the same intent structure; enable call assets at account or campaign level with correct numbers and hours; send traffic to a fast mobile lander with click-to-call + form.

When a dedicated “calls” campaign still makes sense: rare — usually as a temporary bridge with the same RSA + call assets stack, not as a forever call-only island. Prefer intent structure (brand / service / emergency) over “format structure.”

PMax: can remain, but do not use deprecation as an excuse to dump all Search into PMax. Fix Search RSA + call assets first; PMax prerequisites stay in the PMax playbook.

3. Build RSA before you pause call-only

Ship RSA with:

• Pin carefully (do not over-pin); include call-led headlines (“Call for same-day…”) and proof headlines.
• Call assets with the number humans answer (tracking number OK if it forwards correctly — see call tracking guide).
• Sitelinks to service/pricing/areas; callouts; structured snippets where relevant.
• Final URL to a mobile lander that loads fast and shows the phone above the fold.

4. Align conversion goals before the cutover

If call-only used a “Calls from ads” conversion with no duration floor, fix that before or during migration — otherwise RSA + call assets will inherit the same junk signal (primary vs secondary, bad conversion data).

Recommended: qualified call (duration threshold) as money primary when phone closes jobs; form submit secondary or co-designed per multi-path rules; never two primaries for one human journey.

5. Cut over with a controlled pause

Pause call-only ads/campaigns after RSA + assets are eligible and test calls appear in Ads. Do not pause on Friday afternoon with no weekend coverage.

Expect a short learning blip if query traffic moves between campaigns — freeze major Target CPA edits for several days (learning phase).

6. Validate in 48 hours

• Call asset showing in Ads preview / ad strength tools where available
• Test tap-to-call from ad → conversion past threshold
• Website click-to-call still tracked if that path matters
• No duplicate primary on the same call
• Mobile lander phone visible without hunting

Full QA scripts live in the call tracking guide.

RSA + call assets vs call-only — what you gain and lose

Gain: sitelinks and proof; form backup when users will not dial; creative testing surface; less platform sunset risk; easier shared negatives with the rest of Search; lander analytics for message match.

Lose: the ultra-minimal “one button, no page” experience. Mitigate with click-to-call sticky bars, short landers, and call-led RSA copy — not by clinging to a deprecated ad type.

Myth: “RSA cannot get calls.” False. Call assets + mobile SERP behavior still drive large phone share for service queries. If calls collapse after migration, check assets eligibility, hours, number quality, and conversion settings — not “RSA killed phone.”

Myth: “We should move everything to Performance Max because call ads died.” False. PMax without call instrumentation and clean goals amplifies the same measurement problems at larger surface area.

Myth: “Ad Strength Excellent means phone demand is safe.” False. Ad Strength does not prove call assets are eligible, staffed, or converting past your duration floor. Run the test-call script anyway.

Practical parity target: within 14 days, qualified calls (over threshold) within ~15–20% of the pre-migration weekly average after normalizing for seasonality — or explain the gap with staffing, spam filter tightening, or query mix changes. If calls fell 50%+ with no staffing change, the migration broke assets or tracking.

Conversion and Smart Bidding pitfalls during migration

1. Silent loss of call conversion action when the old campaign dies and nobody attached the action to the new RSA campaigns’ goals.
2. Every call asset dial = primary recreates spam CPA.
3. Form-only goals after migration because “RSA has a landing page now” — phone becomes invisible again (multi-path).
4. Aggressive Target CPA cut the day after cutover because volume dipped for 72 hours.
5. Double counting website call conversions + calls-from-ads on one conversation.
6. Wrong business hours on call assets → missed calls overnight counted weirdly or users dial dead lines.
7. Tracking number on call asset that does not forward to a staffed line — Quality and lead flow both die.
8. Bid strategy flip on cutover day (e.g. Maximize Clicks → Target CPA) stacked on creative change — you will not know which lever hurt.

Fix measurement with the call tracking playbook; use this insight for the creative cutover sequence. If Smart Bidding was already trained on spammy call-only conversions, expect CPA to look “worse” when you raise the duration floor — that is honesty. Re-baseline against booked jobs for two weeks before judging the RSA stack a failure.

Keyword and negative hygiene during the swap

Call-only campaigns often collected a private negative list and weird match-type habits. When you pause them, port negatives into the RSA campaigns the same day or you will pay for the junk queries call-only had already blocked.

Also reconcile duplicate keywords across call-only and RSA Search. Two campaigns bidding the same emergency terms with different goals is how you get auction confusion and reporting that never matches sales anecdotes.

Brand terms: keep brand RSA with call assets if branded calls are real demand — do not assume call-only was the only brand vehicle. Non-brand: confirm the RSA lander matches the query class (emergency vs quote vs service-area) so you do not inherit call-only’s thin relevance on research queries.

Copy and UX checklist for phone-first RSA

• At least two headlines that invite the call without sounding like a scam (“Call for same-day…”, service + geo).
• Proof in descriptions: response time, license/insurance, price posture if honest.
• Lander: phone in header + sticky mobile bar; form still present; no click maze.
• Match emergency vs booked-appointment intent — emergency queries want dial; research queries want proof then dial.
• Do not remove the form “to force calls” — you will lose silent converters and break multi-path reporting.
• Keep consent and recording notices consistent with your call-tracking vendor setup when calls are recorded.
• Test the sticky bar on iOS Safari — the most common place DNI or tel: links silently fail after a redesign.

Multi-location and franchise notes

Call assets must show the right number for the geo the campaign targets. A national RSA with one HQ number while ads target five cities recreates the old call-only mess with worse politics.

Prefer location-aware campaigns or location assets / correctly scoped numbers. Keep GBP NAP stable; do not put rotating DNI numbers into GBP (call tracking guide).

US / UK operator notes

US home services: high mobile call share — migration is urgent when banners appear. Watch aggregator keywords; call assets on brand and high-intent service terms first.

UK trades: same pattern; confirm hours and local numbers. If Local Services Ads also run, avoid double-counting the same phone lead across products without a rule.

APAC: phone may share the throne with WhatsApp — migrate call ads to RSA + call assets, but do not pretend chat is solved by call assets. Keep the multi-path map honest.

14-day migration plan

Days 1–2: Inventory call-only campaigns; export keywords, negatives, budgets, conversion actions.
Days 2–4: Fix call conversion duration/primary rules; confirm vendor/Google call plumbing.
Days 4–7: Build RSA + call assets + lander phone UX; QA previews.
Day 7–8: Soft launch RSA alongside call-only (shared negatives); compare early CTR/call volume.
Day 8–10: Pause call-only; monitor Diagnostics and call conversions.
Days 10–14: Tune assets/copy; re-baseline CPA vs booked jobs; only then scale budgets.

Staff weekend coverage on cutover week. Phone-first businesses feel migration failures as empty Saturdays — not as a dashboard footnote.

Common failure patterns

• Pausing call-only before RSA is eligible / call assets approved.
• RSA live with no call asset (“we’ll add extensions later”).
• Keeping spam-friendly call conversions through the migration.
• Moving budget to PMax as a panic move.
• Changing Target CPA, creative, and tracking on the same day.
• Landers that hide the phone behind chat widgets.
• Franchise HQ number on city campaigns.
• Declaring victory on Ad Strength while test calls never hit Ads.

FAQ

Are Google Ads call ads deprecated?

Google has been deprecating or limiting dedicated call-only / call ad creatives in favor of RSA with call assets. Check your account notifications for hard dates. Plan the RSA + call assets destination even if your account still allows call-only temporarily.

What replaces call-only ads?

Responsive Search Ads with call assets, plus a mobile landing page that supports click-to-call (and usually a form). Keep qualified call conversions in your campaign goals.

Will we lose call volume after migrating?

Not if call assets are correctly configured, numbers are staffed, and conversion tracking still records qualified calls. Temporary dips during learning are normal; permanent cliffs usually mean missing assets, bad hours, or broken forwarding.

Should call conversions stay primary after we have RSA landing pages?

If phone still closes a material share of jobs, yes — with a duration or disposition filter. Do not switch to form-only primaries just because a lander exists.

Is this the same as setting up CallRail?

No. This insight is the creative/campaign migration. CallRail/DNI/spam thresholds are the measurement OS in the call tracking guide.

Can we keep one call-only campaign “for emergencies”?

Only as a short bridge if the UI still allows it. Architect the long-term stack as RSA + call assets so a future sunset does not strand emergency demand.

Do call assets work with Performance Max?

Phone demand can appear across surfaces; measurement and goals still need qualified call (and/or OCI) actions. Deprecation is not a reason to skip Search RSA fundamentals.

How do we avoid double-counting during migration?

One human call → one primary conversion. Do not run unrestricted calls-from-ads and website call conversions as dual primaries on the same conversation. See multi-path rules.

What should we tell the client or boss?

“Call-only is going away as a format. We are moving to RSA with call buttons, keeping phone tracking with quality filters, and validating calls before we change budgets.” Show the 14-day plan — not a vague “Google update.”

When should we change Target CPA?

After cutover validation and a short learning window — not on day one of the pause. Re-baseline against booked jobs, not against the old spam-inflated call count.

What if our account still allows call-only creatives?

Still build the RSA + call assets destination now. Soft-launch beside call-only, port negatives, then pause call-only on your schedule — not Google’s surprise banner schedule. Early migration beats emergency migration.

How do we report success to a non-marketer owner?

Show weekly qualified calls, booked jobs from phone, and spam rate — before vs after. Skip Ad Strength screenshots as the hero proof. If qualified calls held and spam fell, the migration worked even if raw dials dropped.

What to do next

1. Inventory call-only / call ad usage this week.
2. Fix call conversion duration and primary flags.
3. Launch RSA + call assets + mobile phone UX.
4. Pause legacy call ads after test calls pass.
5. Deepen measurement with the call tracking playbook.

Get a free Google Ads audit — we will map call-only exposure, RSA readiness, and whether your call conversions are safe for Smart Bidding.

Call ads deprecation is not the death of phone leads. It is the end of a single-purpose creative. Migrate the format, keep the measurement honest, and stop letting a sunset notice become a quiet lead-gen outage.