Most Malaysian accounts still bid as though every conversion is identical. A brochure request, a quote on a RM 60,000 fit-out, and a job applicant using the contact form all arrive in the interface as one conversion each. Target CPA then does what it was told: it buys more of whatever is cheapest, which is almost never the fit-out enquiry.
That is the gap value-based bidding in Google Ads is built to close. Setting the numbers themselves is a separate job, and our guide to setting Google Ads conversion values covers where the figures come from. This page is the layer above it: choosing a bid strategy that acts on those values, and knowing when your account has enough data to survive the switch. ZenWeb runs this migration often enough to know where it goes wrong.
Before the Malaysian numbers, here is Google's own walkthrough of the sequence.
1. What Value-Based Bidding in Google Ads Actually Changes
Quick Answer: It changes what the auction optimises for. Target CPA counts conversions and tries to make each one cheap. Value-based bidding reads the ringgit figure on each conversion and spends where the return is bigger, so it will happily pay three times the usual cost for the right lead.
Two strategies sit under this heading, both inside the wider Smart Bidding family: Maximise conversion value, and Maximise conversion value with a Target ROAS. Google's documentation on value-based bidding for Search and Shopping describes the first as spending the budget for the most total value, and the second as doing the same under an efficiency constraint.
None of this is a switch you flip. It sits on measurement, which is why our Google Ads management treats a value-based bidding migration as a tracking project first. Still choosing between the two families? Target CPA versus Target ROAS is the head-to-head.
The difference from volume bidding shows up in three places:
- Bids stop being flat across conversion types. A quote request worth RM 900 gets a bid the enquiry form worth RM 40 never will, in the same ad group.
- Cost per lead usually rises. That is the strategy working, not failing. Judged on cost per lead alone it looks like a regression.
- Bad data gets amplified. With volume bidding a wrong value is cosmetic. Here it is an instruction.
Key takeaway: Value-based bidding does not make leads cheaper. It makes the account spend where the money is, which usually means a higher cost per lead and a better bank balance.
Not sure your conversions are worth what you think?
We audit the value on every conversion action before touching a bid strategy.
See how we run Google Ads for Malaysian SMEs →2. What Is a Lead Worth at Each Stage of Your Pipeline?
Quick Answer: Take the gross profit on a closed sale, then multiply it by the probability of reaching a sale from that stage. A raw enquiry that closes 5% of the time on a RM 7,600 profit is worth RM 380. That number, not the deal size, is what belongs on the conversion action.
Sending deal revenue is the most common mistake we see. It teaches the account the wrong lesson twice: once by ignoring margin, once by treating a form fill as a sale. Stage-weighted profit fixes both, starting from a clear definition of what counts as a conversion.
| Industry | Raw Enquiry | Qualified Lead | Quote Issued | Closed Sale |
|---|---|---|---|---|
| Aircond & home services | 18 | 45 | 120 | 380 |
| Dental & aesthetics | 35 | 95 | 260 | 900 |
| Legal & professional | 60 | 180 | 520 | 2,100 |
| Interior fit-out | 90 | 310 | 1,150 | 6,500 |
| Industrial B2B supply | 55 | 240 | 900 | 5,200 |
Source: ZenWeb client tracking across 12 industries, Malaysia, 2024–2026. Licence.
The ratio matters more than the ringgit. Interior fit-out runs roughly 1:72 from enquiry to closed sale; aircond services run about 1:21, which is why home-services accounts survive volume bidding far better than project businesses do. If you do not know your own close rates yet, checking whether your Google Ads are profitable comes first.
Key takeaway: The value you send is gross profit multiplied by the odds of getting there. Deal revenue on a form fill is not a value — it is a guess with a currency symbol.
3. First-Sale Profit or Lifetime Value: Which Number Do You Send?
Quick Answer: Send first-sale gross profit. Lifetime value belongs in your budget decision, not in the conversion action, because it is a forecast the bidding algorithm cannot verify for months. The exception is a subscription or service-contract business where the second payment is contractual rather than hoped for.
The argument for customer lifetime value is honest enough: a dental patient is worth years of visits, not one scale and polish. The problem is timing. Value-based bidding in Google Ads reacts within days, and a lifetime figure gives every early lead a number nobody can test for a year. If the retention assumption is optimistic, the account overbids for twelve months before anyone notices.
A workable middle ground for Malaysian SMEs:
- Bid on first-sale gross profit. Verifiable inside 30 to 60 days, which is the window the algorithm learns in.
- Budget on lifetime value. Let it decide how much you are willing to spend in total, not what any individual bid should be.
- Count a second sale only when it is contracted. A 12-month maintenance agreement signed at the point of sale is real revenue. A "they usually come back" is not.
If the values look right but reporting disagrees with your accounts, the fault is upstream. Wrong conversion values in GA4 and enhanced conversions failing to match both produce plausible numbers that bid badly.
Key takeaway: Bid on what you can prove within two months. Keep lifetime value for the budget conversation, where being wrong is cheaper.
4. Conversion Value Rules: Adjusting Value Without Touching the Tag
Quick Answer: Conversion value rules adjust a conversion's value at auction time based on location, device or audience, without a developer touching your tracking code. For a Malaysian business with uneven geography or a strong repeat-customer list, they are the fastest route to values that reflect reality.
These rules sit on top of whatever your conversion values already say. Google's own description of conversion value rules is that they adjust reported value and Smart Bidding optimisation together, in real time, using geography, device or audience as the condition. Add, multiply or set a specific value, at account or campaign level.
Three rules earn their place in most Malaysian accounts:
- Geography, where service cost differs. A Klang Valley job with no travel cost is worth more than the same job in Kuantan. Multiply down for the distant states rather than excluding them.
- Existing customers, where the sale is easier. Upload the customer list, then multiply value up for that audience. The close rate justifies the higher bid.
- Device, only with evidence. Do not apply a mobile multiplier because it feels right. Check your own close rate by device first.
The caution is that rules stack. Google documents the impact of value rules on Smart Bidding, and a rule on top of an inflated base value compounds the error. Fix the base first, then layer rules — the order that also keeps broad match with smart bidding honest.
Key takeaway: Value rules are a multiplier on whatever you already send. Get the base number honest before you start adjusting it by state or audience.
5. Do You Have Enough Conversions to Run Value-Based Bidding?
Quick Answer: Google requires at least 15 conversions in the past 30 days before Target ROAS is available on Search and Shopping. In practice a Malaysian SME account wants 30 or more before the bidding settles, because value data is noisier than conversion counts.
The published minimum and the workable minimum are different numbers. Google's page on Target ROAS bidding sets the Search and Shopping requirement at 15 conversions in the past 30 days — an eligibility gate, not the point where the strategy behaves. Below it, fixing an account with too few conversions is the real job.
| Conversions / Month | Bidding Stability | Weeks to Settle | What to Run |
|---|---|---|---|
| Under 15 | Not viable | Target CPA or Maximise conversions | |
| 15–29 | 8–10 | Max conversion value, no target | |
| 30–59 | 5–7 | Max conversion value, then Target ROAS | |
| 60–149 | 3–5 | Target ROAS | |
| 150 or more | 2–3 | Target ROAS, portfolio strategy |
Source: ZenWeb operational data, Malaysian SME campaigns under management, 2024–2026. Licence.
Eligible and ready are two different numbers, and Google only publishes the first one.
The middle rows are where most Malaysian SMEs sit, and they are the rows that get abandoned early. Eight weeks of unstable bidding feels like failure by week three.
Key takeaway: Fifteen conversions a month gets you in the door. Thirty gets you a strategy that behaves. Below fifteen, fix conversion volume first.
Want to know if your account is ready for Target ROAS?
We check conversion volume, value integrity and offline import before recommending a switch.
Compare Target CPA and Target ROAS →6. How to Migrate to Target ROAS Without Resetting Learning
Quick Answer: Get values flowing for a full 30 days before changing anything, move to Maximise conversion value without a target, then add a Target ROAS at roughly 80% of your last 30 days of actual ROAS. Changing budget and strategy in the same week causes the crash people blame on the strategy.
How to move a Malaysian SME account onto value-based bidding
Budget six to eight weeks. The first four are measurement work with no visible change in the account.
- Put a distinct value on every conversion action. Stage-weighted gross profit, per the table above. One flat value makes the strategy behave like Target CPA.
- Connect the offline half of the pipeline. Import closed sales from your CRM so the algorithm learns which enquiries became money. Offline lead conversion is the mechanism.
- Let 30 days of valued conversions accumulate. Do not touch the bid strategy here. You are building the history the algorithm reads on day one.
- Switch to Maximise conversion value with no target. Keep the conversion goal the same so the learning carries over, then leave it two weeks.
- Read your 30-day ROAS, then set the target below it. Around 80% of the achieved figure. A target above your own history is an instruction to stop spending.
- Adjust by 10 to 15%, fortnightly at most. Faster moves push the campaign back into learning, the self-inflicted version of a ROAS drop.
Step two is the one accounts skip, and it decides whether any of this works. Offline conversions failing to import looks like a bidding problem for weeks before anyone checks the upload. Google is also moving these uploads to the Data Manager API from 15 June 2026, so anything on the older path needs re-plumbing this year.
Key takeaway: Thirty days of valued data, then Maximise conversion value, then a target below your own history. Change one thing at a time.
7. What Actually Happens in the First Eight Weeks
Quick Answer: Conversion volume falls and cost per lead climbs from week one, while total conversion value dips before recovering past baseline around week four. Accounts that judge the switch at week two see only the bad half of that curve, and revert.
The table indexes the three numbers that move, with the 30 days before the switch set at 100. Read it alongside how a failing bid strategy behaves — the two look identical for a fortnight.
| Metric (Index) | Wk 1 | Wk 2 | Wk 3 | Wk 4 | Wk 6 | Wk 8 |
|---|---|---|---|---|---|---|
| Total conversion value | 92 | 88 | 97 | 106 | 118 | 127 |
| Conversion volume | 96 | 89 | 85 | 83 | 81 | 79 |
| Cost per lead | 104 | 112 | 118 | 121 | 124 | 126 |
Source: ZenWeb-managed campaigns, Malaysia, 2024–2026. Baseline 100 = 30 days before switch. Licence.
Week two is the trap: value down 12%, leads down 11%, cost per lead up 12%, and the meeting where someone says "put it back". By week eight the account is producing 21% fewer leads for 27% more value, which is the whole point. Judge it on the value row.
Key takeaway: Fewer leads at a higher cost is the expected shape of a working migration. Give it eight weeks and read the value row.
8. Where Value-Based Bidding Fails in Malaysian Accounts
Quick Answer: Almost every failure traces back to the values, not the algorithm. Revenue sent instead of margin, one flat value on every action, offline sales never imported, and a target above your own history cover most accounts we are asked to rescue.
Each has a distinct fingerprint, and most surface first in conversion tracking, not in the bidding column.
| Failure Mode | What You See | Root Cause | Fix |
|---|---|---|---|
| Inflated value | ROAS looks excellent, profit does not move | Sending revenue instead of gross profit | Re-send margin, reset target to match |
| Flat value | Bidding behaves identically to Target CPA | One value applied to every conversion action | Stage-weighted values per action |
| Missing offline half | Buys enquiries that never close | Sales recorded in the CRM, never imported | Daily offline conversion upload |
| Target set too high | Impression share collapses within days | Target ROAS above the account's own history | Reset to 80% of achieved 30-day ROAS |
| Double counting | Overbids on repeat form submitters | Every conversion counted on a lead form | Count one conversion per click |
Source: ZenWeb client sample, Malaysian SME accounts, 2024–2026. Licence.
Four of the five are measurement faults wearing a bidding costume, so the fix starts in call tracking and offline lead conversion. The pattern repeats across campaign types: Performance Max asset groups reward accurate value signals for the same reason, and exporters reaching overseas buyers need per-country values before any target makes sense.
Key takeaway: If value-based bidding is underperforming, audit the values before you touch the target. The algorithm is usually doing exactly what it was told.
9. Conclusion: Bid on Profit, Not on Volume
Quick Answer: Put stage-weighted gross profit on every conversion action, import closed sales daily, wait 30 days, move to Maximise conversion value, then add a target below your own achieved ROAS. Judge the result at week eight on total value, not on cost per lead.
Value-based bidding in Google Ads is not a smarter Target CPA. It asks the auction where the money is, rather than how cheap a conversion can be, and the account answers only as well as your data lets it.
Start narrow. Take one campaign with real conversion volume, put honest stage values on its conversion actions, connect the CRM, and leave it a month. If the value line climbs at week eight while the lead count falls, widen it. If not, the problem is a number you sent. That sequence is what our Google Ads management is built around.
Two things to settle first. Value bidding assumes the demand underneath it is complete, which in Malaysia means covering Bahasa Malaysia as well as English keywords. And none of it works without a usable pipeline record, so whether you need a CRM comes first.
Ready to bid on profit instead of lead count?
Book a free 30-minute strategy session. We'll review the values on your conversion actions, check whether your account has the volume for Target ROAS, and give you a 90-day migration plan with realistic ROAS targets.
Get my free strategy session →
10. Frequently Asked Questions
1. What is value-based bidding in Google Ads?
It is a group of Smart Bidding strategies that optimise for the total ringgit value of conversions rather than the count. The two are Maximise conversion value, which spends the budget for the most value it can find, and Maximise conversion value with a Target ROAS, which does the same under an efficiency constraint you set.
2. How many conversions do I need before I can use Target ROAS?
Google requires at least 15 conversions in the past 30 days for Search and Shopping campaigns. In practice, Malaysian SME accounts want 30 or more a month before the bidding settles, because value data varies more than a plain conversion count.
3. Should I send revenue or profit as the conversion value?
Send gross profit, weighted by the probability of the lead reaching a sale. Revenue ignores margin, so the account learns to chase high-turnover low-margin work. A raw enquiry with a 5% close rate on RM 7,600 of profit is worth RM 380.
4. Will value-based bidding make my leads more expensive?
Usually yes, and that is the strategy working. In managed accounts, cost per lead typically rises around 20 to 25% within eight weeks while total conversion value rises faster. Judge the switch on cost per lead alone and you will revert a change that was making you money.
5. Do I need offline conversion imports to run value-based bidding?
For lead generation, effectively yes. Without closed sales imported from your CRM, the algorithm only ever sees form fills and cannot tell a serious buyer from a browser. E-commerce is the exception, because the sale and its value are captured on the website at the point of purchase.
