Smart Bidding Strategies Explained: tCPA, tROAS, MaxConv & More

Smart Bidding is the part of Google Ads where most advertisers either ride a wave or get crushed by one. Pick the right bidding strategy and the algorithm scales your account. Pick the wrong one and you watch CPCs triple while conversions flatline.

The reason this happens is rarely the algorithm. It is almost always the advertiser choosing a strategy that does not match their data, their goal, or their stage. Smart Bidding strategies are powerful, but each one has specific requirements to work properly.

This guide walks through every Smart Bidding strategy in plain English, when to use each one, how to transition between them without crashing performance, and the silent traps that catch most accounts. No theory — only the decisions we make on client accounts every week.

What Smart Bidding actually is

Smart Bidding is the family of automated bidding strategies inside Google Ads. The algorithm sets each bid in real time based on signals like device, location, time, audience, query intent, and dozens of others a human cannot evaluate that fast.

You give it a goal — get the most conversions, hit a target cost per acquisition, hit a target return on ad spend — and the AI does the bidding work. Your job becomes choosing the right goal and feeding the algorithm clean data.

The bidding strategies that fall under Smart Bidding are: Maximise Conversions, Maximise Conversion Value, Target CPA, Target ROAS, and Enhanced CPC (the older, partially-automated option). Manual CPC is not Smart Bidding.

The five Smart Bidding strategies — when to use each

Each strategy is designed for a specific stage and goal. Pick the one that matches yours, not the one that sounds most sophisticated.

Maximise Conversions

Goal: get the highest possible number of conversions within the daily budget.

Best for: new accounts, new campaigns, or accounts re-establishing baseline data after a major change. Also good when conversion volume matters more than cost per conversion — for example a paid pilot campaign measuring demand.

Requirements: you need conversion tracking working correctly. No volume minimum to start, but the algorithm needs around 30 conversions per month to optimise well.

Watch out for: this strategy will spend the full daily budget every day. If your budget is loose, you can pay €80 for a lead that should cost €20.

Maximise Conversion Value

Goal: get the highest total conversion value within the daily budget. Used when your conversions have varying values (e-commerce purchases, lead-quality scores).

Best for: e-commerce accounts with proper revenue tracking. Lead-gen accounts that pass back closed-deal values via offline conversion import.

Requirements: conversion value tracking on every conversion event. Without revenue data per conversion, this strategy behaves like Maximise Conversions.

We use this on accounts where lifetime value differs meaningfully between conversion types — for example our Marseille cosmetics e-commerce case study where Performance Max revenue was the right metric, not conversion count.

Target CPA (tCPA)

Goal: hit a specific cost per acquisition on average.

Best for: established lead-gen accounts with at least 30 conversions per month and a known acceptable CPA. Service businesses, SaaS lead gen, healthcare practices.

Requirements: minimum 30 conversions in the last 30 days. Conversion tracking accurate. Some history with Maximise Conversions or eCPC first so the algorithm has data to anchor on.

Watch out for: setting an aggressive tCPA on day one of a new campaign. The algorithm will starve the campaign of impressions trying to find conversions at that price. Always start with Maximise Conversions, let it run 2 to 4 weeks, see your average CPA, then switch to tCPA at 90 to 100 percent of that average.

Target ROAS (tROAS)

Goal: hit a specific return on ad spend (revenue divided by ad cost).

Best for: e-commerce with consistent average order value, SaaS with predictable contract values, any account where revenue per conversion is tracked accurately.

Requirements: minimum 50 conversions with value data in the last 30 days. Stable revenue tracking. Some prior history with Maximise Conversion Value.

Watch out for: aggressive tROAS targets choke spend. If your steady-state ROAS is 4x, set the target at 3.5x to start, not 6x.

Enhanced CPC (eCPC)

Goal: blend manual bidding with limited automation. Google adjusts your manual bids by up to ±30 percent based on conversion likelihood.

Best for: legacy accounts transitioning from full manual CPC. Some advertisers use it as a stepping stone before going to Maximise Conversions.

Requirements: minimal — works at any volume.

Watch out for: Google is phasing eCPC out for most account types. Treat it as a temporary bridge, not a long-term strategy.

The decision tree we use on client accounts

When we audit an account, the bidding strategy decision follows this exact path.

• Less than 15 conversions per month: keep Manual CPC or Enhanced CPC. Smart Bidding lacks data to work.
• 15 to 30 conversions per month: switch to Maximise Conversions, no target. Let the algorithm learn for 4 weeks.
• 30 to 50 conversions per month, lead gen: Maximise Conversions or early-stage Target CPA at the historical average.
• 50+ conversions per month, lead gen with consistent CPA goal: Target CPA, target at 95–100 percent of the historical average.
• 50+ conversions per month, e-commerce with revenue tracking: Maximise Conversion Value for 2–4 weeks, then Target ROAS at 80 percent of baseline.
• High-budget e-commerce with strict profitability: Target ROAS, dialed in monthly.

This is not gospel. Context matters. But it is a reliable starting framework.

The hidden requirement — conversion data quality

Every Smart Bidding strategy depends on accurate, complete conversion tracking. The most expensive bidding mistake is setting a target before tracking is correct.

Before switching to any Smart Bidding strategy, verify:

• Conversion event fires once per real conversion (not duplicates from refresh)
• Conversion value is correct and matches your CMS or CRM
• Enhanced Conversions are enabled with hashed first-party data
• Offline conversions are imported for lead-gen accounts where closing happens later
• GA4 and Google Ads agree on conversion counts within 5 percent

If any of those are off, fix tracking first, switch bidding second. We covered the full tracking setup in our conversion rate optimization guide.

How to switch between bidding strategies without crashing

Switching bidding strategies resets some of the learning data. The algorithm needs time to re-learn, and during that period performance can swing.

Here is the protocol we follow to minimise disruption.

Step 1 — Document the baseline

Before switching, write down the past 30-day metrics: conversions, CPA, ROAS, average CPC, impression share. These are your benchmarks.

Step 2 — Pick the right moment

Avoid switching during major sales periods, product launches, or budget changes. Switch during a "normal" two-week window so the algorithm learns from typical data.

Step 3 — Set the target conservatively

If switching to Target CPA, set the target at 100 to 110 percent of historical CPA. Tightening from there is easier than recovering from a starved campaign.

If switching to Target ROAS, set the target at 80 percent of historical ROAS. Aggressive targets are the number-one reason ROAS strategies fail.

Step 4 — Watch the first 7 days

Conversions usually dip 10 to 30 percent in the first week as the algorithm recalibrates. This is normal. Do not panic-revert.

If conversions drop more than 50 percent after 7 days, something is wrong. Check that conversion tracking still fires. Check that bid changes did not coincide with a budget cut.

Step 5 — Adjust the target every 4 weeks

After the algorithm stabilises (week 4), tighten the target by 5 to 10 percent if performance is solid. Loosen by 5 to 10 percent if volume is too low.

Never change the target by more than 15 percent at a time. Big swings reset learning.

Common Smart Bidding mistakes

These are the patterns we see on most underperforming accounts.

Setting an aggressive tCPA on day one. The algorithm cannot magically find leads at a price below your historical average if the underlying data does not exist.

Switching strategies every 2 weeks. Each switch wipes part of the learning data. Pick a strategy and let it run at least 6 weeks before evaluating.

Mixing manual and smart bidding within the same campaign. Smart Bidding takes over all bid management. Do not also adjust device, location, or audience bid modifiers at the same time the algorithm is trying to learn — those signals are now redundant and confusing.

Ignoring conversion lag. B2B leads that close 30 to 90 days later need offline conversion import. Without it, the algorithm learns from the wrong success signal.

Setting Target ROAS too aggressive. The most common reason ROAS strategies stall. Set at 80 percent of baseline, tighten gradually.

When Smart Bidding does not work

Smart Bidding fails in specific cases. Knowing them saves you weeks of confusion.

• Too little volume. Under 15 conversions a month, the algorithm has no signal. Stay manual or use Maximise Clicks.
• Wildly inconsistent conversion value. If your conversion values vary from €10 to €10,000 with no pattern, the algorithm struggles. Segment campaigns by value tier.
• Heavy seasonality with short cycles. If demand swings 5x within a single week (e.g. Black Friday), Smart Bidding lags. Switch to Maximise Conversions and manage budget aggressively.
• Brand new product or service. Zero historical data, zero signals. Build baseline with Maximise Conversions first.

How Smart Bidding interacts with other levers

Smart Bidding is one piece of the puzzle. The bidding strategy can only optimise within the data and creative you give it.

A campaign with Target ROAS but weak landing pages will plateau. The algorithm finds the best clicks it can, but conversion rate caps performance. The same applies to weak ad copy, poor Quality Score, or aggressive negative keywords blocking valid traffic.

If you switch bidding strategies and performance stalls, check the inputs before blaming the algorithm. Often the bidding choice was right and the issue is upstream.

A 30-day Smart Bidding optimisation plan

If you are currently on Manual CPC or eCPC and want to migrate to Smart Bidding, follow this sequence.

Days 1–7 — Verify conversion tracking, fire test conversions, confirm Google Ads and GA4 agree. Enable Enhanced Conversions if not active.

Days 8–14 — Switch to Maximise Conversions, no target. Budget set to 1.5x your previous daily spend so the algorithm has room.

Days 15–28 — Let it run. Do not change targeting, ad copy, or budget. Watch daily but only adjust if something breaks.

Day 29 onwards — Evaluate. If conversion volume is steady and CPA is acceptable, optionally switch to Target CPA at 100 percent of the 30-day average. If conversion volume is too low, hold on Maximise Conversions or revert.

Frequently asked questions

What is the best Smart Bidding strategy for a new campaign?

Maximise Conversions, no target. Let the algorithm build a baseline for 4 weeks before adding any target.

Should I use Target CPA or Maximise Conversions long-term?

Target CPA if you have a strict CPA goal and 50+ conversions per month. Maximise Conversions if you want maximum volume and can absorb variable CPA.

How do I know what tROAS to set?

Look at the average ROAS over the past 60 days on Maximise Conversion Value. Set the target at 80 percent of that to start. Tighten 5 to 10 percent monthly.

Why did my conversions drop after switching to Target CPA?

Three usual reasons: target too aggressive, not enough historical data, or tracking broke at the same moment. Verify each before reverting.

Can I use Smart Bidding with Performance Max?

Performance Max uses Smart Bidding by default — typically Maximise Conversion Value with optional target ROAS. You cannot use Manual CPC inside Performance Max. We covered the full setup in our Performance Max best practices guide.

How often does Smart Bidding update its model?

The algorithm updates continuously, but visible changes in bid behaviour take 7 to 14 days after a meaningful change. Plan reviews on a 2-week cadence, not daily.

Get a Smart Bidding audit

If your bidding feels off — costs creeping up, volume dropping, ROAS unpredictable — the issue is usually a mismatch between strategy and data. We audit accounts free of charge, pull the bidding strategy decision tree on your actual numbers, and deliver a recommendation in 48 hours.

Book a free 30-minute audit. We screen-share, look at your bidding setup, and walk you through what we would change.

Or explore our Google Ads service for the full system we run on accounts spending €5K to €50K monthly.