Google's Aug. 17 Smart Bidding Change Raises Stakes for Growth

Google's August 17 Smart Bidding update moves budget-constrained campaigns using tCPA (target cost per acquisition) and tROAS (target return on ad spend) closer to their stated targets, rather than allowing them to significantly outperform those goals, according to Google's indication cited in the analysis.

That change gives more weight to the numbers advertisers enter. Smart Bidding (Google Ads' automated bidding system) optimizes toward a chosen goal, so the analysis warns: “Smart Bidding is optimized to achieve its targets, not necessarily to grow the business.” Hitting CPA (cost per acquisition) or ROAS (return on ad spend) can coexist with flat revenue, profit and growth.

Google's indication is the platform update cited here. The seven explanations are editorial analysis, not tool signals or study results; the source provides no sample size or measured benchmark.

What changed on August 17?

Before this update, some budget-limited campaigns could significantly outperform their targets. After it, they should land nearer to those numbers. The analysis treats the target field as an instruction rather than a suggestion, which magnifies the cost of a stale target.

Which failures can hide behind healthy Google Ads results?

1. The target may ignore current margin. Advertisers often set CPA or ROAS targets from historical performance, an industry benchmark or a once-safe number. When COGS (cost of goods sold), shipping, overhead, product mix and customer behavior shift, 4.0x ROAS on a 20%-margin line differs from 4.0x on a 60% line. Hitting it may produce break-even or a loss.

2. The easiest conversions may come from existing customers. Returning customers and people already familiar with the brand tend to convert faster and more cheaply than cold prospects, so Smart Bidding favors their queries and audiences. An account can hit every target by repeatedly converting people likely to buy anyway, leaving the customer base no broader.

3. Branded demand may take credit for existing sales. Accounts heavy on branded search or retargeting often post the portfolio's best ROAS. Yet some of that revenue might have arrived through organic search, direct traffic or existing demand. Full Google Ads attribution then lifts reported ROAS while incremental revenue remains flat.

A conversion may be a proxy, not a sale. Business-to-business (B2B) lead-generation accounts often optimize for form fills, PDF downloads or booked calls. Without closed-won feedback, the algorithm learns to find more form fillers rather than buyers and can keep using historical patterns after lead quality changes. A record month of platform conversions may produce no new pipeline.

Efficiency targets can cap reach. Strict targets make Smart Bidding skip costlier auctions, even when they could deliver profitable net-new volume. Efficiency-first setups show high ROAS, low CPA, low impression share (eligible impressions captured), restricted reach and capped volume. Growth-first setups accept moderate ROAS and higher CPA for broader reach, more incremental revenue and higher net profit.

The system may treat every conversion as equal. Without customer-value weighting, immediate CPA optimization favors price-sensitive, low-tier buyers who convert easily and churn quickly. In subscription or software-as-a-service (SaaS) models, where lifetime value matters, month-one cancellations can depress growth while the ad account reports excellent efficiency.

The leak may occur after the click. Google Ads can deliver demand to the front door, but it cannot repair leads sitting untouched for three days, orders for out-of-stock products or stalled onboarding. Revenue can leak between conversion and the balance sheet, leaving no bidding change able to explain or fix the lost growth.

That trade-off also explains how an efficient account can shrink. In effect, account settings impose the ceiling before the market does: the highest-growth auctions may cost more than the easiest conversions.

Which account signals should advertisers question?

The source's diagnostic sequence asks six questions the platform does not automatically raise:

• Target: When was the current tCPA or tROAS set, what supported it, and can anyone connect it to current profit margins?
• Customer mix: What share of conversions and conversion value comes from new versus returning customers, and is that share stable, shrinking or unmonitored?
• Brand contribution: How much reported revenue sits in branded campaigns? When non-brand spend rose, did total business revenue rise too, or did only platform numbers move?
• Conversion definition: Is the primary action a real sale or something that might become one? How often do those conversions become revenue, and does closed-won data (recorded won sales) return to Google Ads?
• Constraint: Is impression share being lost on campaigns hitting target? Does demand limit the account, or does the target?
• Post-click operations: How quickly are leads contacted? How does paid-search close rate compare with other channels? Do stock or capacity issues suppress results?

How should advertisers respond?

The response is an audit, not a blanket bid increase. Advertisers should compare bid targets and conversion goals with financial results, including margin, new-customer revenue and pipeline.

They should revisit targets when COGS, shipping, overhead or product mix change; measure new versus returning customers; weight conversions by customer value; test branded and nonbrand spend against total revenue; and return closed-won outcomes to Google Ads.

Campaign type changes the exposure. Brand-heavy and retargeting-heavy accounts face attribution inflation; B2B lead-generation accounts face proxy-conversion and feedback gaps; subscription and SaaS accounts face customer-value and churn blind spots; product sellers face margin and stock problems. Existing-customer skew can affect any account.

Downstream fixes remain separate from bidding. Fast lead follow-up, available inventory and onboarding capacity can restore value that Google Ads cannot create. The source's warning is direct: “Google Ads can only deliver demand to the front door.” Efficiency targets should therefore reflect profitable growth, not merely successful attainment.

Monitor two linked series after August 17: the gap between campaigns and their tCPA or tROAS settings, and whether new-customer revenue, margin, pipeline and post-click completion improve alongside reported ROAS. The next signal is whether reported efficiency reaches new customers and the balance sheet.