There is no universal conversion count at which Manual CPC suddenly becomes better than Target ROAS. Google publishes eligibility guidance for Target ROAS by campaign type, but that is not a guarantee that the strategy will outperform manual bidding, nor a promise that every eligible campaign has reliable value data.
For Search and Shopping, Google currently lists at least 15 conversions in the past 30 days at the conversion tracking level for Target ROAS. Other campaign types have different requirements. Google also advises that Search campaigns with newly added or changed conversion values wait one to two conversion cycles for values to populate at a similar rate before adopting value bidding. Treat those as platform setup guidance, not as a “winner threshold.” Check the current page and account eligibility before acting. About Target ROAS bidding
The decision depends on what the campaign is measuring and how trustworthy the values are. Manual CPC gives the advertiser control over maximum cost-per-click bids. Target ROAS uses reported conversion values to adjust bids at auction time toward an average return target. If the value signal is missing, delayed, duplicated, or disconnected from business value, a polished ROAS number can mislead either the bidder or the analyst.
First separate three different questions
Can the campaign select Target ROAS? Check current eligibility for that campaign type and its conversion setup. Eligibility is a product requirement, not a performance verdict.
Is Target ROAS receiving useful data? Confirm the primary conversion actions, value definitions, currency, transaction IDs where applicable, and conversion delay. If every lead has the same placeholder value despite very different sales outcomes, the system is not receiving the value distinctions the business cares about.
Which strategy produces the better business result? Answer with a fair comparison against a defined objective: verified revenue or contribution margin for ecommerce, or qualified and progressed leads for services. Do not decide from CPC, reported ROAS, or a short unsettled date range alone.
Google describes Manual CPC as a method for setting the maximum CPC yourself. Its guidance also suggests Maximize Clicks when you want an automated way to drive clicks and do not know which keywords or placements are profitable. Manual control can suit a specific control need, but it transfers ongoing bid management to the advertiser. It does not inherently create better lead quality or lower acquisition cost. Manual CPC bidding
Check conversion-value readiness before comparing
Use this audit before changing a live bidding strategy:
| Check | Evidence to inspect | Decision if it fails |
|---|---|---|
| Outcome definition | Is the action an order, qualified lead, or another agreed business stage? | Fix the goal definition before choosing an automated value target. |
| Conversion accuracy | Do test and live diagnostics show the intended action once, with correct values and currency? | Repair the measurement path and verify it before the comparison. |
| Value meaning | Are values actual revenue/margin or a documented proxy tied to later outcomes? | Do not call the output ROAS when values are arbitrary. |
| Data timing | How long after click do outcomes arrive, and are values uploaded consistently? | Extend the analysis window and account for the lag. |
| Campaign type | What current Target ROAS requirements apply to this specific type and account? | Check the current official guidance and UI; do not reuse a threshold from another type. |
| Business economics | What margin, fulfilment, cancellation, and repeat-purchase rules matter? | Agree on a business-derived target rather than copying a generic percentage. |
For Search and Shopping, the published 15-conversion figure can help identify a listed prerequisite, but conversions with no meaningful value still do not create a useful ROAS objective. Google’s value-based bidding guidance says Target ROAS should have at least 15 conversions in the past 30 days at the conversion tracking level and recommends reporting two or more different values for the goal. It also explains how conversion delay and regular offline uploads affect learning. Smart Bidding with value-based bidding for Search and Shopping
Do not turn this into “manual until 15, automated after 15.” A campaign may meet an eligibility bar and still need measurement work. Another campaign may have fewer events but use an available strategy that fits its goal. The right question is whether the platform receives enough timely, accurate signal for the decision you want it to make—and whether the comparison can detect a business-relevant difference.
Design a fair test, not a before-and-after story
If the account and campaign support an experiment, define the hypothesis and compare arms that differ in bidding strategy while keeping the offer, landing page, targeting, budget conditions, conversion goals, and time period as comparable as practical. Record the changes that cannot be held constant. Avoid changing creative, keywords, location, and conversion tracking at the same time; otherwise the result cannot isolate bidding.
Choose one primary outcome and a few guardrails before launch. For a retail campaign, that could be verified conversion value or contribution after returns, with spend and order volume as guardrails. For B2B, it could be qualified pipeline value, with valid lead rate and cost per qualified lead as guardrails. If downstream outcomes arrive slowly, define how the lag will be handled and avoid treating recent incomplete data as final.
Google’s Smart Bidding test guidance recommends simplifying the test and changing one variable at a time. Its suggested experiment timing is an initial ramp-up period followed by at least 30 days of running, with recent days excluded when less than 90% of conversions have been reported. That is guidance for Google Ads experiment design; check the current page, account capabilities, and actual conversion cycle rather than applying a fixed timeline blindly. Test your bid strategy
For observational comparisons where a formal experiment is unavailable, label the result as observational. Match the periods for seasonality and sales lag, annotate budget and target changes, and avoid attributing all differences to the strategy. Google notes that conversion delay can make recent Smart Bidding performance look weaker until later outcomes are reported, and advises allowing one to two conversion cycles after major bidding or budget changes before assessing. Tips on measuring Smart Bidding performance
Know when Manual CPC may be the practical choice
Manual CPC may be a reasonable temporary or continuing fit when the business needs direct maximum-click bids for a tightly scoped Search campaign, the team can review and maintain keyword-level bids, and the immediate goal is controlled traffic rather than value optimization. It can also be considered when conversion value tracking is not ready, but that does not remove the need for an agreed business outcome or a plan to improve measurement.
Target ROAS may be worth evaluating when the campaign has eligible settings, reliably reported conversion values that reflect relative business worth, enough regular outcomes for meaningful assessment, and a target grounded in margin and historical evidence. Google’s guidance recommends using historical value-per-cost data and allowing one to two conversion cycles after target changes; a target set too high can constrain traffic. Eligibility and setup still do not guarantee results.
Use the strategy that fits the real goal and the account’s evidence. If the data is weak, mark the result “not ready” or “inconclusive” and state why. Do not invent a numeric cutoff that Google does not prescribe as the point where one strategy beats the other.
Bidding decision card
Campaign type and network: [Search / Shopping / other; verify strategy availability]
Business outcome: [verified order value / qualified pipeline / other]
Primary conversion action(s): [exact account actions and status]
Value definition and currency: [actual / proxy; calculation]
Recent eligible volume: [date window and conversion-level source]
Conversion delay: [observed distribution and maturity rule]
Current strategy and target: [captured from account]
Alternative strategy and hypothesis: [specific change to test]
Primary result and guardrails: [metrics and business thresholds]
Test method and timing: [supported experiment or observational method]
Known limitations: [seasonality, low sample, missing outcomes, other changes]
Short-video script: Fifteen conversions is not a winner signal
Hook: “Google lists a Target ROAS eligibility requirement. That does not mean the strategy wins at conversion number 15.”
Visual: Show two checklists: “eligible to use” and “proven better for this business.”
Voiceover: “Verify the campaign type, conversion values, data quality, and delay. Then compare against a business outcome in a fair test. A setup threshold is not a performance guarantee.”
Close: “Use the account’s evidence—not a magic cutoff.”
Educational carousel: Manual CPC vs Target ROAS without fake thresholds
- Cover: “Manual CPC or Target ROAS? Skip the magic number.”
- Eligibility asks whether a strategy can be selected—not whether it will win.
- For Search and Shopping, Google currently lists 15 conversions in 30 days at the conversion tracking level; other campaign types differ.
- Check whether the values represent different business outcomes and are sent reliably.
- Include conversion lag, order quality, margins, and refunds in the decision.
- Test bidding as the one changing variable; hold other factors steady where practical.
- Report an inconclusive result when the test is too small or not mature.
Field experiment thread: Test bidding with a real business outcome
1/ Record the campaign type, current strategy, conversion actions, value rules, eligible conversion count, and the account’s exact reporting window.
2/ Audit whether the values are accurate and distinct enough to support a value goal. Confirm the business outcome and currency with the people who own sales or finance.
3/ Write one hypothesis: “For [campaign], Target ROAS will [business outcome] versus Manual CPC while meeting [guardrails], because [evidence].” If the hypothesis has no measurable business outcome, refine it.
4/ Use a supported experiment when possible. Change bidding only; keep budgets, targeting, creative, landing page, and conversion actions stable where the account permits. Document unavoidable differences.
5/ Set a mature comparison window based on conversion delay and Google’s current experiment recommendations. Exclude recent incomplete data; define a stop rule for tracking failure or business risk.
6/ Decide: adopt, iterate, retain control, or inconclusive. Do not use the official eligibility count as proof of a winning result.
Tool-neutral prompt pack: Review a bidding-strategy proposal
Prompt 1 — Eligibility vs outcome
Given this Google Ads campaign type, current conversion setup, and current official documentation [details/URLs], separate product eligibility requirements from evidence that the strategy will improve business performance. Flag account settings that must be verified. Do not invent a universal conversion threshold.
Prompt 2 — Value-signal audit
Audit these redacted conversion definitions, values, and timestamps [data] for missing, duplicated, delayed, or non-differentiating signals. State what Target ROAS could meaningfully optimize toward and which business facts remain unknown. Do not infer profitability from ad-platform ROAS alone.
Prompt 3 — Experiment plan
Design a controlled Manual CPC versus Target ROAS test using this campaign and business objective [data]. Specify the single variable, primary outcome, guardrails, conversion-lag rule, practical constraints, stop condition, and inconclusive criteria. Use only current official experiment guidance supplied here [sources].
Stop-doing swipe file: Replace bid-strategy folklore
Stop: “Manual CPC wins before 30 conversions.”
Use: “Choose based on the campaign goal, measurement readiness, and a fair comparison. Platform eligibility rules are not performance cutoffs.”
Stop: “At exactly 15 conversions, switch to Target ROAS.”
Use: “Check the current requirement for this campaign type; then confirm accurate, useful values and mature data.”
Stop: “A higher platform ROAS proves higher profit.”
Use: “Reconcile attributed conversion value with margin, refunds, fulfilment costs, and the business outcome.”
Stop: “Change the bidding strategy and landing page together to get faster results.”
Use: “Isolate bidding where possible so the result can answer the question.”
Stop: “The recent week is worse, so automated bidding failed.”
Use: “Check conversion delay and allow the comparison window to mature before deciding.”
Hook–Value–Template: Make the bid decision auditable
Hook: “Fifteen conversions can be a setup requirement. It cannot tell you which bid strategy makes more money.”
Value: Keep eligibility, data readiness, and business performance as separate checks. Test bidding against qualified outcomes or verified value, with an explicit lag rule.
Decision template: “We evaluated [campaign type] for [period]. Target ROAS eligibility was [verified status and source]. The campaign recorded [count] [defined conversions] with [value definition, currency, and delay]. We compared [strategy A] with [strategy B] using [test/observational method], holding [factors] constant. The mature business outcome was [result] against [guardrails]. Limitations: [sample, lag, confounders]. Decision: [adopt / continue / retain / inconclusive] because [evidence].”
Example: “A Search campaign met the documented Target ROAS eligibility guidance, but the company’s lead values were still flat placeholders and qualified outcomes arrived weeks later. The team did not treat eligibility as proof that Target ROAS would outperform. It fixed the value mapping, waited for the updated values to populate, then planned a controlled comparison with qualified pipeline value and a mature reporting window.”
The useful cutoff is not a slogan. It is the point at which your measurement, eligible setup, business value, and test design support a decision you can defend.