A decade ago, running Google Ads meant sitting in the account adjusting bids by hand — nudging a keyword up 20 cents because it converted well on mobile, pulling another down because it wasted money after 10pm. Today, Google’s machine-learning bidding does all of that in every single auction, using signals no human could process in real time. That is Smart Bidding, and for most Singapore SMEs it is no longer optional — it is how the platform is designed to be run. But “let Google set the bids” hides a lot of important choices: which of the four strategies to use, how much conversion data you need before they work, what target to set, and how not to sabotage the system by fiddling. This guide explains Smart Bidding in plain English for 2026 — including the June 2026 naming changes and an August 2026 update that will affect budget-limited campaigns — so you can choose the right strategy instead of guessing.
If you’re still getting your bearings with the platform, our complete Google Ads guide for Singapore SMEs sets the scene. This post goes deep on the one decision that most directly controls how your budget is spent.
What Smart Bidding actually is
Smart Bidding is Google’s term for the automated bid strategies that use its AI to optimise for conversions or conversion value in every auction — a capability Google calls “auction-time bidding.” Instead of you setting a fixed bid, Google predicts, for each individual search, how likely it is to convert and how valuable that conversion would be, then sets a bid accordingly. It reads a huge range of signals a manual bidder simply cannot — device, physical location, time of day and day of week, browser, operating system, language, whether the user is on a remarketing list, the exact search query, and more — and it combines them in real time.
Four strategies count as Smart Bidding: Maximize conversions, Maximize conversion value, Target CPA (target cost per action) and Target ROAS (target return on ad spend). The key thing that unites them: they all bid at auction time using conversion data. That last part is the catch — without conversion tracking set up correctly, Smart Bidding is flying blind. Getting your conversion tracking right is a prerequisite, not a nice-to-have, and it’s the foundation of everything we cover in performance marketing.
The four strategies, and when to use each
| Strategy | What it optimises for | Best for | Needs a target? |
|---|---|---|---|
| Maximize conversions | the most conversions your budget can buy | lead gen; building early data | No (spends full budget) |
| Maximize conversion value | the most total value from your budget | e-commerce with varied order values | No (spends full budget) |
| Target CPA | as many conversions as possible at a target cost each | lead gen with a known cost-per-lead goal | Yes (your target cost per conversion) |
| Target ROAS | the most value at a target return on spend | e-commerce with reliable revenue tracking | Yes (your target return, e.g. 400%) |
The simplest way to choose: do your conversions all carry roughly the same value, or different values?
- If every conversion is worth about the same — a typical lead-generation business where each enquiry form is broadly equivalent — use Maximize conversions, then graduate to Target CPA once you know your acceptable cost per lead.
- If conversions carry different values — typical for e-commerce, where one order is $40 and another is $400 — use Maximize conversion value, then graduate to Target ROAS once you can track revenue reliably.
Treat Maximize conversions and Maximize conversion value as transitional strategies: they spend your full budget without an efficiency constraint, which is exactly what you want early on to build conversion history, but not what you want forever. Once you have enough data, moving to Target CPA or Target ROAS gives Google an efficiency goal to hold to.
The June 2026 naming change (nothing to panic about)
If you logged into an account recently and saw the bid strategy labels look different, you’re not imagining it. Starting June 2026, Google updated how these strategies are labelled: “Maximize conversions with a Target CPA” is now simply “Target CPA,” and “Maximize conversion value with a Target ROAS” is now simply “Target ROAS.” This is a naming and organisation change only — the underlying bidding behaviour is exactly the same. Nothing about how your campaigns bid changed; only the words in the interface did. It’s worth knowing so you don’t go hunting for a “lost” strategy that was merely renamed.
How much conversion data you actually need
This is where most SME campaigns go wrong: they switch on Target CPA or Target ROAS before the account has enough conversions for Google’s model to learn from, then wonder why performance is erratic. The rough guidance:
- Target CPA: Google’s baseline is at least 15 conversions in the last 30 days; in practice, 30 or more per month gives noticeably more stable performance.
- Target ROAS: because it also has to predict conversion value, it needs more data still — commonly cited as 50+ conversions a month for reliable results. Accounts with variable order values and fewer than 30–50 monthly conversions often see erratic Target ROAS performance.
Below those thresholds, Maximize conversions is the safer choice while you build history, because it has lower data requirements. There is no shame in staying on Maximize conversions for a few months — it is the correct call for a young account, not a fallback.
There is also a learning period to respect. Whenever you switch strategies or make a significant change, Google’s system recalibrates for roughly one to two weeks, and performance is volatile during that window. Crucially, big changes — a budget or target adjustment above about 15–20%, pausing ads, or switching strategy — can reset the learning period. The practical rule we give clients: change targets in small steps (no more than ~15–20% at a time) and then leave the campaign alone for a couple of weeks. Constant fiddling keeps a campaign permanently in learning mode, which is the enemy of stable results.
Choosing a strategy: a decision path for a Singapore SME
For a lead-generation SME — a renovation firm, a clinic, a B2B services company — the usual journey is: launch on Maximize conversions, let it gather 30-plus conversions, then switch to Target CPA at a cost per lead you can afford. For an online store, it’s: launch on Maximize conversion value, then move to Target ROAS once revenue tracking is solid and you’re clearing 50+ conversions a month. If you run Performance Max as well, note that it uses the same Smart Bidding logic under the hood — Maximize conversions/value with optional CPA/ROAS targets — so the same data thresholds apply.
Is manual bidding ever still the right call?
Occasionally, yes — and it’s worth being honest about when. Smart Bidding needs conversion data to work; if a campaign genuinely can’t generate conversion signals, the AI has nothing to optimise toward. The clearest cases are a brand-new account with zero conversion history and no way to import past data, or a campaign whose goal isn’t a trackable online conversion at all (pure brand awareness, for instance, where you care about impressions and reach rather than form fills). In those situations, starting on Maximize clicks or a manual CPC while you install and validate conversion tracking is defensible — not because manual is better, but because Smart Bidding literally can’t function without the data.
For the overwhelming majority of Singapore SME lead-gen and e-commerce accounts, though, the honest answer is that manual bidding leaves performance on the table. A human simply cannot adjust bids per auction across device, location, time, audience and query the way auction-time bidding does. The right sequence is almost always: get conversion tracking working, run Maximize conversions to build data, then move to a target strategy — not stay on manual out of a desire for “control” that, in practice, means slower, coarser bidding.
Portfolio strategies for multi-campaign accounts
If you run several campaigns that share a business goal — say three Search campaigns all chasing leads at a similar cost per lead — a portfolio bid strategy lets Google optimise across all of them toward one shared Target CPA or Target ROAS, rather than treating each in isolation. The advantage is pooled data: three campaigns that individually sit below the conversion threshold can, combined, give the algorithm enough signal to bid well. For a smaller Singapore SME with a handful of tightly related campaigns, a portfolio Target CPA is often the difference between “not enough data to be stable” and “enough data to trust.” The trade-off is that a portfolio will move budget and bids toward whichever campaign is converting most efficiently, so use it only when the campaigns really do share a goal — not to force unrelated objectives under one roof.
Setting a target that helps rather than starves
The most common Target CPA/ROAS mistake is setting the target too aggressively. If your real cost per lead has been running at SGD 40 and you set a Target CPA of SGD 20, you’re effectively telling Google to only bid on the cheapest, rarest auctions — and your volume can collapse. Similarly, setting Target ROAS far above what the account has historically achieved starves the campaign of impressions. Start your target close to your recent actual performance, then tighten it gradually (again, ~15–20% steps) as the system proves it can hit the goal. Ambition is fine; ambition faster than the data can support is self-defeating.
Two further tips from the accounts we run: make sure your Quality Score and landing pages are healthy before you blame the bid strategy — Smart Bidding can only work with the auction position your relevance and Ad Rank earn — and don’t judge a strategy inside its learning period. Give any change a clean two to three weeks before deciding it isn’t working.
The August 2026 update every advertiser should know
Google has announced that starting 17 August 2026, it is updating its bidding systems to deliver more consistent, predictable performance for campaigns that are limited by budget. If your Target CPA or Target ROAS campaigns are budget-constrained, Google has flagged that these changes may cause temporary performance and traffic fluctuations around that date. This is not something you need to act on pre-emptively, but it is something to watch: if you see a wobble in mid-to-late August, check whether the campaign is budget-limited before assuming something broke. Knowing the update is coming saves you from over-reacting — and over-reacting, as we’ve seen, just resets the learning period.
Smart Bidding and the Singapore cost picture
Smart Bidding changes how you bid, not the underlying economics of the auction. The cost of Google Ads in Singapore — commonly SGD 1.50–6.00 per click for many sectors, and higher for competitive fields like legal, finance, aesthetics and B2B — is still governed by competition, your Quality Score and Ad Rank. What Smart Bidding does is spend that money more intelligently within those constraints, chasing conversions rather than clicks. One Singapore-specific reminder that applies regardless of bid strategy: GST at 9% has applied to Google Ads spend in Singapore since 1 January 2024, on both media and any management fee, for accounts with a Singapore business address. Budget the gross figure. And remember that Smart Bidding is only as good as the conversion data you feed it — if you’re not sure your tracking is accurate, that’s the first thing to fix, not the bid strategy.
Common Smart Bidding mistakes
- Switching to a target strategy too early, before the account has the 15–30+ (CPA) or 50+ (ROAS) monthly conversions the model needs.
- Setting the target too aggressively, strangling volume by only bidding on the cheapest auctions.
- Changing budgets or targets constantly, keeping the campaign permanently in learning mode.
- Judging performance during the learning period instead of waiting two to three weeks.
- Feeding it bad conversion data — duplicate conversions, wrong values, or untracked leads — so the AI optimises toward the wrong outcome.
- Blaming the bid strategy for a Quality Score or landing-page problem that no amount of clever bidding can fix.
The honest bottom line
Smart Bidding is genuinely powerful, and for most Singapore SMEs it will out-perform manual bidding — but only when it’s fed clean conversion data, given enough volume to learn from, and left alone long enough to do so. Start on Maximize conversions or Maximize conversion value to build history, graduate to Target CPA or Target ROAS once you have the data, set targets close to reality, and resist the urge to fiddle. Do that, and the machine will spend your budget more shrewdly in a day than a manual bidder could in a month. Get the fundamentals wrong — no data, unrealistic targets, constant changes — and even the best AI in the world can’t save the account.
Not sure whether your bid strategy is helping or hurting? We’ll review your conversion data, volume and targets and tell you honestly what to run. Talk to us or explore our Google Ads management and the results in our case studies.
FAQ
What is Smart Bidding in Google Ads?
Smart Bidding is Google’s set of automated, AI-driven bid strategies — Maximize conversions, Maximize conversion value, Target CPA and Target ROAS — that set a bid in every auction using live signals like device, location, time and intent to optimise for conversions or conversion value.
Which Smart Bidding strategy should a Singapore SME use?
If your conversions are similar in value (lead gen), start with Maximize conversions and move to Target CPA once you have 30+ conversions a month. If they vary in value (e-commerce), start with Maximize conversion value and move to Target ROAS at 50+ conversions a month.
How many conversions do I need for Smart Bidding to work?
Target CPA needs at least 15 conversions in 30 days (30+ is more stable); Target ROAS needs more, commonly 50+ a month, because it also predicts conversion value. Below that, use Maximize conversions while you build history.
Did Google rename its bid strategies in 2026?
Yes. From June 2026, “Maximize conversions with a Target CPA” became simply “Target CPA,” and “Maximize conversion value with a Target ROAS” became “Target ROAS.” It’s a labelling change only — the bidding behaviour is unchanged.
What is the learning period, and why does it matter?
When you switch or significantly change a Smart Bidding strategy, Google recalibrates for about one to two weeks, and performance is volatile during that time. Big changes (over ~15–20%) can reset it, so make small adjustments and wait before judging results.
Will the August 2026 bidding update affect me?
Only if your Target CPA or Target ROAS campaigns are limited by budget. Google’s update from 17 August 2026 may cause temporary fluctuations for budget-constrained campaigns, so watch for a short-term wobble around that date rather than over-reacting to it.
