Smart Bidding Strategies Compared for Malaysia

Smart Bidding is the part of a Google Ads account most Malaysian business owners agree to without deciding on. The setup flow offers a bid strategy, the recommended option is already selected, and the campaign goes live. Nobody chose anything.

The consequence turns up a month later, usually as a question with no clean answer: is this working? Cost per enquiry has moved, but so has the volume, the season and the competition. There is no baseline to compare against because the strategy was never picked deliberately in the first place.

There is a second reason to look at this now. Google is changing how target-based bid strategies behave from 17 August 2026, and accounts that have been quietly outperforming their stated target are the ones most likely to feel it. This piece compares the four strategies on what each one actually demands from you, and sets out the conversion volume Google publishes for each. It then explains what changes next month and gives a twelve-week plan for switching without losing the thread. The video below covers the mechanics first.

PART 1 · DIAGNOSE

What Smart Bidding Actually Decides For You

IN BRIEFSmart Bidding sets a bid in every individual auction using signals you cannot see or adjust manually. It does not decide what a conversion is or what it is worth, you do. That division of labour is the whole subject, and it sits underneath every managed Google Ads decision.

Google’s definition is narrow and worth taking literally. Smart Bidding refers to bid strategies that use Google AI to optimise for conversions or conversion value in every auction, a feature Google calls auction-time bidding. Four strategies qualify: Maximize conversions, Target CPA, Maximize conversion value and Target ROAS.

What it buys you is signal coverage. Google lists device, physical location, location intent, day and time, remarketing list membership, ad characteristics, browser, operating system and the actual search query behind the keyword. Those signals are read per auction, in combinations no team could compute by hand.

What it does not buy you is judgment about your own business:

• It cannot tell a good lead from a bad one. It optimises towards the conversion actions you ticked. If a low-quality action is in that column, it will find more of it.
• It cannot invent value. Value-based strategies need conversion values you report yourself, which most Malaysian service businesses have never set up.
• It cannot start without tracking. Google requires conversion tracking before any of these four strategies runs, which makes a clean conversion setup the real prerequisite.

BENCHMARK BRIEFING 1 OF 4

The Four Strategies, Side by Side

IN BRIEFTwo strategies spend your budget and report what it produced. Two hold you to a number you set. The split matters more than the four names, because it decides whether the budget or the target is the thing that constrains delivery.

The third column is the honest filter. Two of these four strategies are unavailable to a business that has never attached a ringgit figure to a conversion, no matter how attractive the reporting looks.

PART 2 · DIAGNOSE

The Data You Need Before You Choose

IN BRIEFGoogle publishes different data thresholds for different strategies and campaign types, and they are lower than most advertisers assume. Target CPA can start with no conversion history at all. Target ROAS on Search cannot. Reading the actual numbers usually settles the argument faster than a debate about what the account is ready for.

Three separate numbers get confused here, and separating them removes most of the confusion.

• The eligibility threshold, the conversion volume Google requires before a strategy can be applied at all.
• The evaluation threshold, the volume Google recommends you have before you judge the result. For Target CPA, Google advises measuring over the last 30 days including at least 30 conversions.
• The learning period, up to three weeks or one to two conversion cycles for the strategy to calibrate after a change.

An account can clear the first and fail the second, which is how a campaign ends up switched on legitimately and judged on nothing. That is a budgeting problem disguised as a bidding problem.

BENCHMARK BRIEFING 2 OF 4

Conversion Volume Google Publishes Per Strategy

IN BRIEFThe published requirements vary by campaign type by a factor of twenty. A Search campaign needs 15 conversions in 30 days for Target ROAS; a Demand Gen campaign needs 50 in 35 days. The same gap shapes what is realistic for a brand starting out on YouTube.

For a Malaysian services business running Search only, the practical reading is encouraging. Fifteen valued conversions a month is a reachable bar. The heavy requirements sit on campaign types most local advertisers are not running yet.

PART 3 · DESIGN

Which Strategy Your Account Has Earned

IN BRIEFWork upward, never downward. Start where your data honestly sits, earn the next step with evidence, and treat each move as a decision with a review date. Choosing a strategy above your data is the most expensive mistake available in a new Google Ads build.

Google supports this sequence explicitly for the final step. Before moving to value-based bidding, it recommends already bidding to the goal with Target CPA, and reporting values for four weeks or three conversion cycles, whichever is longer, before setting a ROAS target.

BENCHMARK BRIEFING 3 OF 4

What Changes on 17 August 2026

IN BRIEFBudget-limited campaigns on Target CPA or Target ROAS will start delivering closer to the target you actually entered. Accounts that have been beating their target are the ones exposed. Anyone running a Performance Max campaign is inside the affected group.

Google’s own worked example is the clearest way to see the exposure. A budget-limited campaign carrying a target of RM100 while actually delivering at RM50 will begin moving towards RM100. Nothing has broken. The account was simply reading an outcome the system no longer promises to repeat.

Google also states it will not adjust your targets or budgets for you. Doing nothing is a choice with a consequence, which makes the next four weeks a genuine decision point for Malaysian accounts running on targets set a year ago.

PART 4 · DEPLOY

Where Malaysian Accounts Get This Wrong

IN BRIEFFour failures account for most disappointing results, and three of them happen outside the bidding settings entirely. Changing strategy again is rarely the fix, it usually just restarts the learning period and destroys the comparison you were building.

• Switching strategy every few weeks. Each change can trigger a learning period of up to three weeks, so an account changed monthly never leaves calibration.
• Setting bid limits to feel safe. Google advises against them for both target strategies because they stop the system bidding at the level the target requires, and on standard strategies they are not available anyway.
• Expecting old bid adjustments to still apply. Under Target ROAS, existing bid adjustments are not used, with the single exception of a device adjustment of -100%.
• Blaming bidding for a landing page problem. The strategy buys the click; the page converts it, which is why the page behind the ad deserves review before the bid strategy does.

There is a budget consequence worth stating plainly too. Google recommends being comfortable spending up to twice your average daily budget on any given day, with the monthly total still capped at your daily budget across the billing cycle. Advertisers who watch daily spend rather than monthly spend often panic on day three of a perfectly normal month.

BENCHMARK BRIEFING 4 OF 4

A Twelve-Week Plan for Switching Strategy

IN BRIEFThree months is the realistic frame: calibration, then a clean measurement window, then a decision. Compressing it produces an opinion rather than a result, which is the same discipline problem behind most accounts that are not converting.

The phase most often skipped is the third. Accounts reach week six, see a number they like or dislike, and act on it before the sample is large enough to mean anything.

PART 5 · DRIVE

Judging the Strategy Honestly

IN BRIEFCompare your achieved cost against the average target the strategy was optimising for, not against the number you typed. The two differ whenever targets changed mid-period, and reading the wrong one is how a perfectly sound account gets rebuilt for no reason.

Google supplies the comparison directly. The bid strategy report shows your actual CPA or ROAS beside the average target the strategy was working towards, along with conversion delay and target simulators. Most accounts never open it, then argue about performance from the campaign table instead.

Two habits separate accounts that get a clean answer from accounts that get an argument:

• Measure a fixed window, not a rolling feeling. Pick the dates in advance, exclude the most recent conversion delay period, and compare like with like.
• Check the business, not just the account. If enquiries rose and closed sales did not, the problem is downstream of bidding, a diagnosis that usually starts with a structured audit of the site itself.

Then step back to the sequence. A strategy is not a permanent setting; it is a rung. When the data improves, the next rung becomes available, and that is a scheduled decision rather than an experiment worth running on a quiet Friday. The same logic governs how a managed account is reviewed each quarter.

COMMON QUESTIONS

Smart Bidding FAQ

1. Which Smart Bidding strategy should a small Malaysian business start with?

Maximize conversions, in almost every case. It depends on whether your conversion tracking has been running long enough to trust, if it has not, no strategy will help. Starting without a target lets the account produce the cost figure you will later use to set one, which is the missing input in most small accounts.

2. Do I need conversion data before switching to Smart Bidding?

Not always, though it depends on the strategy. Google states advertisers can start using Target CPA with no conversion history, while Target ROAS on Search and Shopping requires at least 15 conversions in the past 30 days. Conversion tracking itself is required before any of the four strategies can run.

3. Will the 17 August 2026 change increase my cost per lead?

It can, and it depends on the gap between your stated target and your actual result. Google’s guidance is that budget-limited campaigns using a target strategy will deliver more consistently towards the target you entered. A campaign quietly beating its target will drift up towards it unless you lower the target first.

4. Why did my cost jump right after I switched strategy?

Usually the learning period, which lasts up to three weeks or one to two conversion cycles while the system calibrates. That said, if the jump persists past calibration, look at the conversion actions included in the bidding column before you look at the bid strategy, a loose definition costs more than a wrong strategy.

5. Can I still use bid adjustments with Smart Bidding?

Mostly not, and the exceptions are narrow. Under Target ROAS your existing bid adjustments are not used, apart from a device adjustment of -100%. Under Target CPA, device adjustments modify the CPA target itself rather than the bid, so a +40% mobile adjustment raises your mobile target by the same proportion.