Maximize Conversions vs Target CPA — when each fails for SMB Google Ads accounts

Direct answer — Maximize Conversions or Target CPA?

Use Maximize Conversions when primary conversions are clean and you need volume without inventing a fake cost target. Switch to Target CPA only when you have roughly 30+ qualifying conversions per month and a defendable allowable CPA from finance — set near recent efficient CPA, not 40% below it. Maximize Conversions fails when soft primaries teach the algorithm to buy junk forms. Target CPA fails when the target is fantasy, volume is too low, or conversion rate collapsed and you tighten the bid instead of fixing the page.

This insight owns the Max Conv vs tCPA decision and failure modes. For the full Smart Bidding tree (including value-based bidding and learning-phase rules), use the Smart Bidding SMB playbook. Fix measurement first with the conversion tracking setup guide.

What each strategy actually optimizes

Maximize Conversions tells Google Ads: spend the budget to get as many of your primary conversion actions as possible. There is no hard average CPA ceiling (unless you add an optional target/cap where the UI allows — treat that carefully; it is not the same as a mature Target CPA strategy with history).

Target CPA (tCPA) tells Google Ads: get conversions while aiming for an average cost per conversion you set. The system still explores auctions, but an aggressive target restricts which auctions it enters. Set too tight and volume dies. Set too loose and you overpay.

Neither strategy cares about your intentions. Both optimize the conversion actions marked primary and included in campaign goals. Soft thank-you page fires, scroll events, and unfiltered call clicks make Maximize Conversions look like a hero and Target CPA look “broken” when the real problem is the training label — see why Smart Bidding wastes budget on bad conversion data.

Side-by-side — when each wins for SMBs

| Situation | Prefer | Why | |-----------|--------|-----| | Tracking just rebuilt; history is thin | Maximize Conversions | Avoid fantasy targets while the system relearns clean events | | Equal lead values (one service, one lead type) | Maximize Conversions → then tCPA | Build volume, then constrain cost | | Known allowable CPA + ≥~30 clean conv/mo | Target CPA | Finance has a number the auction can respect | | CPA exploding because soft forms are primary | Neither — fix conversions | Bid strategy changes amplify junk | | Volume collapsed after “aggressive” tCPA | Loosen tCPA or return to Max Conv | Target was the constraint, not “Google broke” | | Brand Search mixed with non-brand in one campaign | Split first | Blended CPA lies; tCPA on the blend starves non-brand |

When Maximize Conversions fails

1. Soft primary conversions

Maximize Conversions will efficiently find the cheapest path to whatever you called a conversion. If that path is a form start, a 10-second session, or a thank-you refresh, platform CPA falls while CRM qualified rate collapses. Operators then say “Maximize Conversions is generating junk leads.” The strategy did its job. The conversion definition did not.

Fix: one money primary (qualified lead, booked job, purchase, or offline closed-won). Demote the rest. Re-baseline for 1–2 weeks before judging the strategy.

Audit pattern we write in scorecards: list every primary action, counting method (one vs every), and whether it is included in “Conversions.” If more than one soft event is primary, Max Conv is disqualified as a strategy debate until that inventory is cleaned.

2. Budget far larger than high-intent demand

With a large daily budget and limited high-intent inventory, Maximize Conversions expands into lower-intent queries and placements to fill conversion goals. Search term reports fill with researchy and DIY queries. CPA rises; sales complain.

Fix: tighten geo and negatives, improve landing conversion rate, or reduce budget to match realistic high-intent capacity — do not jump to a punishing Target CPA as the first move unless allowable CPA is already proven.

3. Untracked phone / WhatsApp closes

If 40–70% of closes start on calls or chat and those paths are not conversion actions, Maximize Conversions overweight web forms. Common in US home services, UK trades, and automotive workshops.

Fix: call conversions with duration thresholds; track chat/WhatsApp clicks where relevant; import CRM stages. Then re-evaluate Max Conv vs tCPA.

4. Celebrating platform CPL while MER is flat

Same pattern as PMax attribution inflation: dashboard efficiency without business efficiency. Judge with qualified pipeline and MER / CAC payback, not CPL alone.

5. Constant strategy flip-flops during learning

Teams switch Maximize Conversions → Target CPA → Manual → Maximize Conversions inside ten days because yesterday’s CPA looked ugly. The account never exits learning long enough to produce a fair test. Maximize Conversions then “fails” in the narrative even though the process never gave it a clean window.

Fix: freeze the strategy for 14 days after a tracking rebuild (critical negatives and broken tracking fixes only). Document the hypothesis and review date.

When Target CPA fails

1. Target set far below historical CPA

The classic SMB mistake: last 30 days CPA is £80 (or $120). Someone sets Target CPA to £45 because a case study promised it. Impression share falls, learning sticks, status messages hint at limited performance by bid strategy / target. Team blames Smart Bidding.

Fix: set Target CPA near recent efficient CPA (often within ~10% of trailing reality), then step tighter only after quality holds. If finance needs £45, improve conversion rate and lead quality first — bidding cannot invent margin.

2. Not enough conversion volume

Target CPA needs clean events to learn. As a working SMB floor, treat ~30 qualifying conversions per month at the campaign or portfolio level as the point where tCPA usually stabilizes. Below ~15, Prefer Maximize Conversions (or Manual / Maximize Clicks for diagnosis) until volume or conversion rate improves. Portfolios can pool similar campaigns that share the same conversion definition and economics.

3. Conversion rate dropped, target stayed the same

Landing page slowdown, offer change, or form friction raises required CPA. Keeping the old Target CPA is equivalent to cutting bids in a harder market. Volume collapses; “tCPA stopped working.”

Fix: diagnose CR and search terms first. Temporarily loosen tCPA or return to Maximize Conversions while you fix the page. Use the Landing Page Conversion Gap when the page is the suspect.

4. Wrong conversion definition under a tight target

Target CPA on unqualified form fills forces Google to buy cheap junk efficiently. Sales gets busier with worse leads. The fix is still the primary action — not a tighter number.

5. Brand and non-brand forced under one Target CPA

Branded Search CPA is often a fraction of non-brand. One Target CPA across both lets brand subsidize a blended average while non-brand starves — or conversely lets non-brand blow the target while brand looks “inefficient.”

Fix: separate branded and non-brand campaigns (or clear portfolio scopes) before Target CPA debates.

6. Portfolio Target CPA across mismatched economics

Pooling an emergency callout campaign with a high-ticket retainer campaign under one Target CPA averages incompatible margins. The portfolio “learns” an average that fits neither line. Use portfolios only when conversion definitions and contribution economics are genuinely similar.

How we set Target CPA on SMB accounts (practical sequence)

1. Confirm primaries — money outcomes only; campaign goals exclude micro-noise.
2. Screenshot baseline — 30-day CPA, conversion volume, qualified rate, MER.
3. Run Maximize Conversions (if not already) until you have stable clean volume — typically toward the ~30/mo floor.
4. Set initial tCPA near recent efficient CPA (not aspirational CPA).
5. Hold 7–14 days — critical negatives only; no daily micro-edits.
6. Step 5–10% tighter only if quality and volume remain acceptable; loosen 10–20% if volume collapsed.
7. Kill criteria — if after a honest target and clean data, qualified pipeline and MER worsen for 3–4 weeks, revisit structure, offer, and conversion definition — not another panic flip to Manual CPC forever.

Derive allowable CPA from unit economics (job value × margin × close rates). Full service playbook: break-even CPA for service businesses. Pair with the Break-Even ROAS calculator when you think in return terms, then convert to CPA for lead gen. Finance must agree the conversion definition matches “qualified.”

Worked example (service SMB)

Suppose average closed job contributes £400 after variable costs, and lead→close is 25% on qualified leads. Allowable media CPA on a qualified-lead conversion ≈ £100. If Google’s primary is still “any form submit” with a 10% close rate, the allowable CPA on that looser event is closer to £40 — and Max Conv will happily find £35 junk forms. Align the event with sales before arguing Max Conv vs tCPA.

Optional Max Conv “target” vs real Target CPA

Some UIs let you add a target or bid ceiling while on Maximize Conversions. That is a constraint on exploration, not a substitute for a Target CPA strategy with history. An aggressive optional cap early often reproduces the “too-tight tCPA” failure: learning stalls, volume dies, and operators conclude automation failed.

Rule we use: no harsh optional caps until Maximize Conversions has proven clean volume and CPA is understood. Then graduate to Target CPA with an evidence-based number.

US and UK callouts

US home services / multi-location: call duration thresholds and CRM imports usually matter more than shaving £/$5 off Target CPA. Aggregator auctions raise floors; improving landing conversion rate beats fantasy targets.

UK trades / local services: consent-aware Enhanced Conversions and careful overlap with Local Services Ads matter. Do not average LSA and Search into one Target CPA story without clear goal design. Map pack demand is still a Local SEO problem — not a bid strategy toggle.

Both markets: keep a change log. Accounts without dated strategy changes cannot tell whether Max Conv or tCPA “worked.”

Search vs Performance Max — same choice, different visibility

On Search, Max Conv vs tCPA is diagnosable: search terms, ad groups, and RSA tests show whether the algorithm bought junk intent. Start discipline here.

On Performance Max, the same strategies apply under campaign goals, with less query glass. Do not use a tight Target CPA on PMax to paper over soft lead-gen tracking. Clear PMax readiness and the PMax playbook first; then apply the same Max Conv → tCPA sequence.

Before / after pattern from audits

When a free Google Ads audit finds a Max Conv vs tCPA fight in progress, the written scorecard usually includes:

1. Conversion inventory (primaries, secondaries, include-in-Conversions flags)
2. Trailing 30-day CPA vs proposed Target CPA (gap %)
3. Monthly qualifying conversion count on the strategy’s scope
4. Qualified rate / MER snapshot (client-provided)
5. Recommendation: stay Max Conv / move to tCPA near history / loosen target / rebuild measurement first

Bring matching CRM dates if you request the audit — otherwise the debate stays stuck on platform columns.

Decision tree (print this)

1. Are primary conversions clean and verified this month? No → fix tracking. Do not switch strategies.
2. Do you have ~30+ qualifying conversions/mo on this scope? No → Maximize Conversions (or build volume / CR).
3. Do you have a finance-approved allowable CPA that matches the conversion definition? No → stay on Maximize Conversions; build the economic model.
4. Is recent efficient CPA within ~20% of allowable? Yes → Target CPA near recent CPA. No → improve CR/offer/negatives before harsh targets.
5. After 2–4 weeks, did qualified pipeline and MER hold or improve? Yes → scale budget in 10–20% steps. No → loosen target or return to Max Conv; audit conversion quality.

If you are mid-argument in a Slack thread, the shortest version is: dirty data → neither strategy; thin volume → Max Conv; clean volume + real allowable CPA → tCPA near history. Everything else is usually a landing page, negative, or offer problem wearing a bid-strategy costume.

What to do this week (operator checklist)

1. Export every primary conversion action — demote soft events.
2. Compare last 30 days CPA on Maximize Conversions (or current strategy) to finance’s allowable CPA.
3. If on Target CPA and volume collapsed: check target vs history, budget, and landing CR before flipping strategies again.
4. Score tracking with Tracking Trust Score and structure with Google Ads Health Score.
5. Read the Smart Bidding SMB playbook for value-based bidding, learning-phase rules, and the 30-day cadence.
6. Request a free Google Ads audit if you want a written Max Conv vs tCPA recommendation on the live account.

FAQ

Is Maximize Conversions better than Target CPA?

Neither is universally better. Maximize Conversions wins while you build clean volume without a hard cost ceiling. Target CPA wins when you have enough conversions and a real allowable CPA. Switching too early or with a fantasy target usually kills volume.

When should I switch from Maximize Conversions to Target CPA?

When tracking is clean, you have roughly 30+ qualifying conversions per month on the campaign or portfolio, CPA variance is understandable, and finance has approved an allowable CPA. Set the first target near recent efficient CPA — not a wish number.

Why did my Target CPA kill volume overnight?

Usually the target was far below historical CPA, budget was too low to learn, conversion rate dropped, or the conversion action changed. Loosen the target 10–20%, fix CR, or temporarily return to Maximize Conversions.

Can Target CPA work with low conversion volume?

Rarely well. Below practical volume floors, Prefer Maximize Conversions, improve conversion rate, or pool similar campaigns in a portfolio that shares the same conversion definition. Harsh tCPA on thin data produces restricted serving and permanent “learning” behaviour.

Should lead-gen SMBs always use Target CPA?

No. Many service SMBs should stay on Maximize Conversions longer — especially after a tracking rebuild — until qualified volume is stable. Target CPA on unqualified form fills efficiently buys junk.

Does this apply to Performance Max?

Yes. PMax uses automated bidding toward your goals. The Max Conv vs tCPA logic is the same; visibility is worse. Fix PMax prerequisites and conversion quality before tightening targets.

How often should I change Target CPA?

In small steps (about 5–10%) after a stable observation window — typically weekly at most once learning settles. Daily micro-edits prolong learning and hide cause and effect.

What if Maximize Conversions CPA is already above my allowable CPA?

Do not slam a Target CPA far below that reality on day one. First cut junk queries, fix landing conversion rate, and confirm the conversion is qualified. Then set tCPA near the new efficient CPA. Using Target CPA as a wishful ceiling on a structurally expensive funnel usually produces “limited by bid strategy” and a quiet account.

Should I use portfolio Target CPA?

Yes when several campaigns share the same conversion definition and similar unit economics and each is too thin alone. No when you are averaging brand with non-brand, or emergency jobs with high-ticket retainers, under one number.

Related frameworks

• Smart Bidding SMB playbook — full strategy selection, learning phase, kill criteria
• Google Ads conversion tracking setup — fix primaries before any bid strategy debate
• SMB marketing metrics that matter — MER, CAC payback, pipeline
• Break-Even ROAS calculator — margin math before aggressive targets
• Break-even CPA for service businesses — allowable CPA formula
• Google Ads Health Score — five-minute account diagnostic
• Smart Bidding readiness scorecard — 36 checks before Target CPA
• 47-point Google Ads audit checklist — structured account review

Free Google Ads audit — senior-led scorecard in 24 hours. We will tell you whether to stay on Maximize Conversions or move to Target CPA on your conversion definition — not a generic default.