Google has been pushing advertisers toward automated bidding for years, and in 2026 the pressure has intensified further. Manual CPC is still available but increasingly treated by Google as a legacy option. Smart bidding has become the default recommendation across virtually every campaign type.

The good news: smart bidding actually works — when it has the data, the right goal settings, and a properly structured campaign. The bad news: a poorly configured smart bidding setup can spend a lot of money with very little to show for it.

Here's where things stand, what's working, and how to make the right choice for your specific campaigns.

The Current Smart Bidding Landscape

Google's smart bidding portfolio includes four main options:

Target CPA (tCPA) — Optimizes bids to get conversions at or below your specified cost per acquisition. Strong for lead generation campaigns where you have clear conversion data.

Target ROAS (tROAS) — Optimizes to hit a specific return on ad spend. The right choice for e-commerce and revenue-focused campaigns.

Maximize Conversions — Spends your budget to get as many conversions as possible without a specific cost target. Good for campaigns with limited conversion history that need to build data faster.

Maximize Conversion Value — Spends your budget to maximize total conversion value (revenue or tracked value). E-commerce play, similar intent to tROAS but without a specific ROAS target constraint.

What's Working in 2026

tCPA Is Still the Workhorse for Lead Gen

For service businesses and B2B lead generation campaigns, Target CPA remains the most reliable smart bidding strategy — when you have the data to support it.

The data threshold matters. Google recommends at least 30-50 conversions per month (some sources say 50+) before tCPA will optimize reliably. Below that threshold, the algorithm doesn't have enough signal, and you'll see erratic results.

What's working well: Setting tCPA targets based on actual historical data, not aspirational goals. If your last 90 days of data shows a $45 cost per lead, starting tCPA at $45-50 and gradually tightening over time produces better results than launching at an aggressive $25 target that the algorithm can't hit.

Common mistake: Setting tCPA too low and wondering why impressions drop. If your target is unrealistically low relative to your market's actual CPC landscape, Google simply stops serving your ads to achieve it.

Maximize Conversions as a Learning Bridge

For new campaigns or for campaigns migrating to smart bidding from manual, Maximize Conversions (without a tCPA constraint) is a useful bridge strategy. It allows the algorithm to collect conversion data without being constrained by a cost target it can't yet reliably hit.

The workflow: start on Maximize Conversions for 30-60 days, accumulate conversion data, then migrate to tCPA once you have enough signal.

What's changed in 2026: Google has improved the data learning speed, so the migration window can often be compressed. Campaigns are reaching reliable smart bidding performance faster than they did two or three years ago.

tROAS for E-Commerce Is More Nuanced Than It Looks

Target ROAS sounds straightforward — set your target, let Google optimize. In practice, it's the bidding strategy that requires the most thoughtful configuration.

What's working: Product-level tROAS for Shopping campaigns, particularly in segmenting high-margin products into separate campaigns with more aggressive ROAS targets and low-margin products into campaigns where the target is set to break-even. This hybrid approach captures the efficiency gains of smart bidding while preventing budget from being wasted on your lowest-value SKUs.

What's not working: A single tROAS target applied across an entire mixed-product catalog. The algorithm will bias toward easier conversions at lower order values to hit the target, often at the expense of your higher-value transactions.

Enhanced CPC: The Unsung Middle Ground

Enhanced CPC (eCPC) is technically a manual bidding strategy with a smart layer — it adjusts your manual bids up or down based on conversion likelihood signals. It's often overlooked in discussions about smart bidding, but in 2026 it remains valuable for:

• Campaigns with conversion volumes too low for full smart bidding
• Advertisers who want to maintain more direct bid control
• Accounts in competitive auctions where smart bidding tends to overbid

eCPC gives you the benefits of conversion-based optimization without fully ceding bid control. For smaller accounts and lower-volume campaigns, it often outperforms tCPA.

When Manual CPC Still Makes Sense

Manual CPC isn't dead. There are specific scenarios where it's still the right choice:

Brand campaigns: Your brand name searches are already pre-qualified intent signals. You don't need Google's algorithm to figure out conversion likelihood on a search for your own company name. Manual CPC on brand campaigns typically delivers better efficiency than smart bidding.

Very low conversion volume accounts: If you're doing fewer than 10-15 conversions per month, smart bidding will be unreliable and often expensive. Manual CPC with careful keyword selection and negative keyword management frequently outperforms smart bidding below this threshold.

Competitor targeting campaigns: These campaigns require nuanced bid management that algorithmic bidding doesn't handle well. Manual control lets you push on specific competitor terms strategically.

The Portfolio Bidding Strategy Worth Testing

One underutilized feature in 2026: portfolio bid strategies applied across multiple campaigns. Instead of each campaign having its own smart bidding optimization, portfolio strategies pool conversion data across campaigns targeting similar goals.

This is particularly valuable for accounts where individual campaigns have insufficient conversion volume for reliable smart bidding but the account as a whole has enough data. Pooling the signals across campaigns can enable smart bidding performance that wouldn't be possible with per-campaign optimization.

Practical Setup Recommendations

Regardless of strategy, these fundamentals make any bidding approach work better:

Conversion tracking must be clean. Smart bidding is only as good as what you're telling it to optimize toward. If your conversion events include junk signals (page views, button clicks that don't represent real intent), the algorithm will optimize toward those junk signals. Audit your conversion events before trusting any automated bidding.

Seasonality adjustments are a must. If you're running a holiday sale, a product launch, or any event that will temporarily change conversion patterns, use Google's seasonality adjustment feature to prevent the algorithm from over-correcting during the anomaly.

Budget shouldn't constrain the algorithm. Smart bidding optimization is undermined when campaigns are significantly budget-constrained. If your campaign is hitting budget limits daily, the algorithm can't optimize freely. Either increase budget or reduce the number of campaigns drawing from the same budget.

Don't change bids too often. Smart bidding needs learning time. Making significant target changes every few days prevents the algorithm from stabilizing. Change targets by no more than 15-20% at a time, and wait at least 7-14 days between significant adjustments.

The Bottom Line

Smart bidding in 2026 is mature enough that the question isn't "manual vs. automated" — it's "which automated strategy, configured how, for which campaign type."

Get the conversion data infrastructure right. Choose the strategy that matches your campaign goal and conversion volume. Set targets based on data, not wishful thinking. And resist the urge to make constant adjustments.

---

*Wondering if your current agency is getting the most from your Google Ads spend? [Check out our PPC agency audit checklist](/blog/ppc-agency-audit-checklist) to find out.*