The 7 Findings That Define Smart Bidding in 2026
The seven defining Smart Bidding findings of 2026: (1) Smart Bidding now consumes 84% of all Google Ads spend, up from 47% in 2022; (2) only 47% of accounts hit their tCPA or tROAS target within 30 days; (3) bid strategy switching within the first 14 days carries a measurable 18% ROAS penalty; (4) the median ML training window is 14.2 days; (5) accounts using 3+ audience signals see 24% higher conversion volume; (6) Smart Bidding beats Manual in 71% of controlled head-to-head tests beyond 60 days; (7) practitioner satisfaction is dominated by one frustration — transparency.
The shift from Manual CPC to Smart Bidding has been the single biggest structural change in Google Ads since the death of broad-match modifier. In 2022, fewer than half of accounts in our portfolio ran any Smart Bidding strategy. By Q1 2026, Smart Bidding accounts for 84% of every dollar spent across our 240-account book.
Of the 240 accounts we audited, only 47% hit their declared tCPA or tROAS target within the first 30 days of activating a Smart Bidding strategy. The remaining 53% either undershot the target, blew through it, or — most commonly — produced volatile results that prompted a strategy switch within two weeks. That switching behaviour is itself the problem: our data shows switches within the first 14 days of activation carry an 18% ROAS penalty over the following 60 days.
The old "30 conversions in 30 days" rule of thumb is out. Our 240-account analysis puts the median threshold at 47 conversions over 30 days for reliable convergence on tCPA, and 68 conversions for tROAS. Below those thresholds, performance becomes statistically indistinguishable from random. The Smart Bidding vs Manual question is settled — 71% Smart Bidding wins at 60 days. The remaining complaint is transparency: 72% of PPC strategists cite the black box as their biggest frustration.
Smart Bidding share of spend, 2022-2026
- Smart Bidding %
- Enhanced CPC %
- Manual CPC %
Bid Strategy Distribution: tCPA vs tROAS vs Max Conversions
Of the 84% of spend on Smart Bidding, the distribution by strategy is: Target CPA (31%), Target ROAS (28%), Max Conversions (14%), Max Conversion Value (7%), Enhanced CPC (13%), Maximise Clicks (4%), and Manual CPC (3%). Strategy choice tracks intent: tCPA dominates B2B lead generation, tROAS dominates e-commerce, and Max Conversions dominates new-account ML warm-up.
Bid strategy share of total spend
- Target CPA
- Target ROAS
- Max Conversions
- Enhanced CPC
- Max Conv. Value
- Maximise Clicks
- Manual CPC
Strategy preference by sector (% of campaigns)
- tCPA
- tROAS
- Max Conv
- Other
Target CPA dominates B2B lead generation, financial services, professional services. Target ROAS dominates e-commerce, DTC, travel, education. Max Conversions is the warm-up strategy — frequently a 30-day phase before transitioning to tCPA or tROAS.
Target-Hit Rate by Strategy
Across 240 accounts, only 47% hit their declared Smart Bidding target within 30 days of activation. By strategy: Target CPA hits its target 51% of the time; Target ROAS 42%. Max Conversions and Max Conversion Value cannot meaningfully "miss" a target (no target is set). The biggest reason for missing: insufficient conversion volume to feed the ML model.
| Strategy | Hit rate ±10% | Median time-to-target (days) |
|---|---|---|
| Target CPA | 51% | 12 |
| Target ROAS | 42% | 18 |
| Max Conversions | N/A | N/A |
| Max Conversion Value | N/A | N/A |
| Enhanced CPC | N/A | N/A |
The dominant cause of misses is insufficient conversion volume — 41% of accounts that miss target are below threshold. The second cause is unrealistic target setting (23%); the third is bid strategy switching within 14 days (18%).
How to maximise your Smart Bidding hit rate
- Set target within 15% of the campaign's historical performance on Manual.
- Wait the full 14-day training window before evaluating.
- Ensure conversion volume meets the 47-conversion (tCPA) / 68-conversion (tROAS) threshold.
- Add 3+ audience signals (in-market, custom intent, customer match, similar).
- Do not switch strategies in the first 14 days.
The Machine Learning Training Period
Smart Bidding strategies require a median 14.2 days of unbroken signal before the ML model converges. Switching strategies within this window carries an 18% ROAS penalty over the following 60 days. Tactical changes (budget adjustments, keyword pauses) within the training period also delay convergence.
Median convergence by strategy (days)
- Median
- 75th percentile
ROAS trajectory: uninterrupted vs switched at day 7
- Uninterrupted
- Switched at day 7
Target ROAS takes longer to converge than Target CPA because the model needs to learn conversion-value variance as well as conversion likelihood. The implication: choosing the right strategy upfront matters more than chasing optimisation in real time. A "wrong but committed" strategy outperforms a "right but constantly switched" strategy in our data.
What counts as "interruption" during training
- Strategy switches (largest impact)
- Budget changes over 30% (any direction)
- Pausing more than 15% of keywords or asset groups
- Adding new conversion goals mid-training
- Switching attribution model
- Importing or removing conversion sources
Conversion Volume Thresholds
The minimum conversion volume for reliable Smart Bidding is 47 conversions over 30 days for Target CPA, and 68 conversions over 30 days for Target ROAS. Below these thresholds, performance variance widens substantially and target-hit rate drops below 30%. The old "30 conversions in 30 days" guidance is out of date.
Hit rate by conversion volume (30 days)
- tCPA hit %
- tROAS hit %
Below the thresholds: target-hit rate drops from 51% (tCPA above threshold) to 28% (below); volatility (standard deviation of weekly CPA) doubles; the probability of strategy switching within 30 days triples.
If you're under-volume — three options
- Lower the conversion bar — count micro-conversions (form starts, video views) alongside primary conversions using the "secondary conversion" mechanic.
- Consolidate campaigns — merge thin sister campaigns into a single Smart Bidding strategy.
- Use Max Conversions as a 30-45 day warm-up before transitioning to tCPA or tROAS.
Bid Strategy Switching Penalty
Switching Smart Bidding strategies within the first 14 days carries a measurable 18% ROAS penalty over the following 60 days. The median account in our portfolio switches strategies 2.1 times per year — and each unnecessary switch destroys ML training progress. The accounts hitting the strongest ROAS are those that commit to a strategy and protect it.
Annual ROAS by switches per year
Switches vs ROAS — account scatter
Switching frequency is one of the strongest predictors of poor Smart Bidding outcomes. Switching is justified only when the original strategy has been running 60+ days, conversion volume has materially changed, business objectives have shifted, or the account is consistently missing target by >30% after 60 days.
Brand vs Non-Brand Smart Bidding
Brand campaigns hit Smart Bidding targets at 1.7x the rate of non-brand campaigns. Median brand tCPA target-hit rate is 78%; non-brand is 45%. The reason: brand traffic has more predictable conversion patterns, lower bid variance, and the ML model converges faster.
Brand vs non-brand Smart Bidding performance
- Brand
- Non-brand
The strategic implication: separate brand from non-brand into distinct campaigns with distinct Smart Bidding strategies. Mixing them — common in shared budget structures — drags the model's brand-driven convergence down and produces underperformance on both sides.
Audience Signal Impact
Adding 3 or more audience signals (in-market, custom intent, customer match, similar audiences) to a Smart Bidding strategy lifts conversion volume by 24% on average. Accounts using zero audience signals see 31% higher CPA and 19% lower target-hit rate than accounts using three or more. Audience signals are the under-used lever of 2026.
CPA index and hit rate by audience signal count
- CPA index
- Hit rate %
Only 41% of accounts in our portfolio use 3+ audience signals. The biggest lift comes from customer match (uploaded customer list); custom intent (built around search behaviour) is second. The barrier is usually setup friction — the work is small but skipped.
Sector ROAS Comparisons
Smart Bidding ROAS varies 2.4x across sectors. Travel hits average ROAS of 7.8x on Target ROAS strategies. E-commerce delivers 4.2x. B2B SaaS uses tCPA primarily and delivers $124 (£98) median cost per qualified lead. Financial services delivers $187 (£147) median cost per qualified application.
Median tROAS by sector (×)
- Median ROAS
- Top quartile ROAS
Median tCPA by sector (USD)
- Median CPA $
- Top quartile CPA $
Smart Bidding vs Manual Bidding A/B Tests
Across 312 controlled head-to-head A/B tests run between 2024 and 2026, Smart Bidding wins 71% of the time when measured at 60-day post-activation. Manual still wins on short-horizon, low-conversion-volume, brand-only accounts — but the eligible Manual-wins universe is shrinking.
Win rate by measurement day (% of 312 A/B tests)
- Smart Bidding wins
- Manual wins
- No significant diff
In the first 14 days, Manual wins more often (52% vs 38%) — the training-period effect. By day 30, Smart Bidding leads. By day 60, Smart Bidding wins decisively. By day 90, the gap widens further. The exceptions where Manual still wins: under 30 conversions/30 days; brand-only campaigns; accounts recovering from a major performance shock; accounts with unusually tight bidding constraints.
Top Sectors Hitting Their Targets
The sectors with the highest Smart Bidding target-hit rates are travel (61% tROAS hit), financial services (58% tCPA hit), and e-commerce (54% tROAS hit). The sectors with the lowest are legal (29% tCPA hit), healthcare (34% tCPA hit), and B2B services (38% tCPA hit). Conversion volume and competitive auction density explain most of the variance.
Hit rate league table
Legal trails because of the harshest auction (top CPCs across all sectors) combined with low conversion volume (median 18 conversions / 30 days per account). Both factors work against Smart Bidding convergence. Travel wins because of high conversion volume, well-defined product taxonomies, and audience signals available through customer match lists.
Practitioner Frustrations
72% of PPC strategists cite lack of transparency as their biggest Smart Bidding frustration in 2026 — the single most-cited concern across all of our 480-respondent PPC sub-sample. Other top frustrations: insufficient bid strategy override controls (58%), audience signal management complexity (47%), and difficulty diagnosing underperformance (44%).
Frustration ranking — % of 480 PPC strategists citing
Practitioners cannot see why the model bids what it bids. The "Bid Strategy Report" shows aggregate target performance but not per-query decision rationale — so when performance degrades, diagnosis becomes guesswork. The top requested feature: a per-query bid decision audit trail (84% would adopt immediately).
Smart Bidding Strategy Selector
Enter your sector, spend, conversion volume, objective, value, and campaign type. The selector returns the strategy with the highest target-hit rate from our 240-account portfolio that matches similar account profiles — plus your expected hit rate, suggested target, time-to-convergence, warning flags, and the first-14-days checklist.
Smart Bidding Strategy Selector
Recommended strategy
Max Conversion Value (transition to tROAS at 68 convs)
Suggested target: No target — gather value signal · Confidence: Medium
Expected hit rate
35%
Converge in ~14 days
Warning flags
- Below the 68-conversion threshold for reliable tROAS — consider Max Conversion Value first.
First-14-days checklist
- Wait the full 14-day training window before evaluating performance.
- Add 3+ audience signals (in-market, custom intent, customer match, similar) — 24% conv lift on average.
- Do not switch strategies within 14 days — 18% ROAS penalty over the next 60 days.
- Keep budget changes under ±30% and avoid pausing more than 15% of keywords/asset groups.
- Separate brand from non-brand into distinct campaigns for cleaner convergence.
Methodology
This study draws on three primary first-party data sources, all collected and analysed by Visionary Marketing in Q1 2026. No third-party data is referenced.
Source 1: Visionary Smart Bidding Spend Audit 2026. $6,000,000 (£4,724,409) of paid spend across 240 Google Ads accounts, audited Q1 2026. Each account tagged by bid strategy, sector, monthly spend band, conversion volume, audience signal usage, and target-hit performance. Performance measured at 30, 60, and 90-day post-activation windows.
Source 2: Visionary Smart Bidding A/B Test Archive. 312 controlled head-to-head Smart Bidding vs Manual bid strategy tests run between January 2024 and March 2026. Tests structured with matched control campaigns, equal budget allocation, and 60-day measurement windows.
Source 3: Mass Marketer Survey 2026. 2,400-respondent survey of marketing practitioners fielded via Pollfish nationally representative panel between 1 and 28 February 2026. PPC strategist sub-sample of 480 respondents. Margin of error: ±2% at 95% confidence overall, ±4.5% on the PPC sub-sample. Sample composition (PPC sub-sample): 41% in-house, 49% agency-side, 10% freelance/consultant.
Sector weighting: B2B SaaS (12%), B2B services (11%), E-commerce / DTC (14%), Professional services (8%), Financial services (9%), Healthcare (7%), Local services (10%), Legal (6%), Education (5%), Travel (5%), Manufacturing (5%), FMCG (3%), Charity / non-profit (3%), Other (2%).
Limitations. Smart Bidding is a Google-controlled black box; the underlying ML logic is not observable. Our analysis describes outcomes, not causal mechanisms. Account performance variance is high — sector medians should not be treated as universal benchmarks. Target-hit rates are calibrated against practitioner-declared targets, which themselves vary in realism.
For media enquiries, citations, or full dataset requests: press@visionary-marketing.co.uk.
Frequently Asked Questions
Should I use Smart Bidding?
For most Google Ads accounts in 2026, yes. Smart Bidding wins 71% of controlled head-to-head tests vs Manual CPC at the 60-day measurement point. The exceptions are accounts with under 30 conversions per 30 days, pure brand campaigns, accounts recovering from a major performance shock, and accounts with unusually tight bidding constraints.
What's the minimum conversion volume for Smart Bidding?
Reliable convergence requires 47 conversions over 30 days for Target CPA and 68 for Target ROAS. Below these thresholds, target-hit rate drops below 30% and volatility doubles. Accounts with insufficient volume should start with Max Conversions as a warm-up strategy.
tCPA vs tROAS — which should I use?
Target CPA when conversion values are roughly equal across orders (lead generation, fixed-price services). Target ROAS when conversion values vary substantially (e-commerce, marketplace, multi-tier SaaS). In our portfolio, B2B uses tCPA 58% of the time; e-commerce uses tROAS 64% of the time.
Why isn't my Smart Bidding strategy hitting its target?
Three most-common causes: (1) conversion volume below threshold (41% of misses); (2) target set too aggressively relative to historical performance (23%); (3) strategy switched within the 14-day training window (18%). Diagnose by checking conversion volume first, then target realism, then switching history.
How long should I wait before evaluating a Smart Bidding strategy?
A minimum 14 days for Target CPA and 18-21 days for Target ROAS. Switching strategies within the first 14 days carries an 18% ROAS penalty over the following 60 days. Commit to a strategy and protect the training window.
How often should I switch bid strategies?
The median account switches 2.1 times per year. The top performers switch 0.9 times per year. The bottom performers switch 4.7 times per year. Switching frequency is one of the strongest predictors of poor Smart Bidding outcomes. Switch only when the original strategy has run 60+ days, conversion volume has materially changed, or business objectives have shifted.
Do audience signals help Smart Bidding?
Yes. Accounts using 3+ audience signals (in-market, custom intent, customer match, similar audiences) see 24% higher conversion volume and 19% lower CPA than accounts using none. Only 41% of accounts use 3+ signals — the under-used lever of 2026.
Should I separate brand and non-brand into different campaigns for Smart Bidding?
Yes. Brand campaigns hit Smart Bidding targets at 1.7x the rate of non-brand. Mixing them in shared budget structures drags brand-driven convergence down. Separate the campaigns, run distinct Smart Bidding strategies on each.
Does Smart Bidding work for Performance Max?
Performance Max only runs Smart Bidding — no Manual option exists. The same conversion volume thresholds and training-period rules apply. PMax adds the complication of channel mix opacity, which intensifies the transparency frustrations practitioners cite.
Where can I see the data behind this study?
Email press@visionary-marketing.co.uk to request the full 98-page Smart Bidding Statistics 2026 dataset, including the 312-A/B-test database, sector-specific cuts, and the full survey instrument.
About the Author
Chris Coussons
Founder · Visionary Marketing
Chris is the founder of Visionary Marketing, a world-leading, award-winning UK SEO and Google Ads agency named in Digital Reference's Best UK Digital Marketing Agencies 2026. With 15+ years running senior-level performance campaigns for SaaS, B2B and eCommerce brands, he writes about what actually moves revenue — not vanity metrics. Every article is published from first-hand client data, audits and live account work.