Most Google Ads accounts start by counting conversions. A form submission is one conversion, a phone call is one conversion, a purchase is one conversion. That is a reasonable start, but it hides an uncomfortable fact: some conversions are worth far more to your business than others. A renovation enquiry for a whole four-room HDB flat and a request for a single cabinet repair both count as one lead, yet only one of them pays the rent.
Value-based bidding fixes this by telling Google Ads how much each conversion is worth, so that its automated bidding can chase value rather than volume. This guide explains how value-based bidding works in plain terms, which bid strategies use it, how a service business without an online checkout can assign sensible values, and the steps and pitfalls to know before you switch.
What is value-based bidding?
Value-based bidding is a family of Google Ads Smart Bidding strategies that optimise for the total value of conversions rather than the number of conversions. Google’s overview of Smart Bidding using value-based bidding describes the idea: you report a value for each conversion, and bidding uses those values to decide how much each auction is worth to you.
Compare the two approaches:
- Conversion-based bidding (Maximise conversions, target CPA) treats every conversion as equal and tries to get as many as possible, or as many as possible at an average cost.
- Value-based bidding (Maximise conversion value, target ROAS) tries to get as much conversion value as possible, or as much as possible at an average return on ad spend.
If every customer is worth roughly the same to you, conversion-based bidding is often enough. If customer value varies widely, value-based bidding gives the system the information it needs to prefer the better ones. Our wider guide to Google Ads bidding strategies covers the full range.
How does value-based bidding work?
Under the bonnet, the process repeats in four stages every time your ad is eligible for an auction.
Stage 1: You report conversion values
Each conversion action is given a value. That can be a fixed amount (every enquiry is worth SGD 80), a dynamic amount passed from your website (the actual basket total), or a value uploaded later from your CRM once you know what the customer bought. Google’s help page about conversion values explains the options.
Stage 2: Google predicts the likely value of each auction
For every search or placement, Smart Bidding uses signals such as device, location, time of day, the search query and audience membership to predict both the chance of a conversion and its likely value. It learns these patterns from the values you have reported in the past.
Stage 3: The bid is set to match your goal
With Maximise conversion value, Google tries to get the most total value within your budget. With a target ROAS, Google’s Target ROAS help page explains that it sets bids to maximise conversion value while trying to keep the average return on ad spend equal to your target. A high-value prediction earns a higher bid; a low-value prediction earns a lower bid or none.
Stage 4: Results feed back into the model
As new conversions and values arrive, the model updates. This is why accurate, timely values matter so much: whatever you report becomes the definition of a good customer.
Target ROAS is calculated as conversion value divided by cost, expressed as a percentage. Google’s own example is simple: if a conversion is worth US$5 and costs US$1, the return is 500%. Our explainer on return on ad spend walks through ROAS in more detail.
Choosing between the value-based bid strategies
The table compares the main options. Names can change; Google has said it is relabelling some value-based strategies in the interface without changing how they bid, so check the current labels in your account.
| Strategy | Optimises for | Best when | Main trade-off |
|---|---|---|---|
| Maximise conversions | Number of conversions | Conversions are similar in value or you are still building data | Ignores differences in value |
| Target CPA | Number of conversions at an average cost | You know what one conversion is worth and values are similar | Can favour cheap, low-value leads |
| Maximise conversion value | Total conversion value within budget | You have reliable values and want as much value as the budget can buy | No efficiency limit, so returns can vary |
| Target ROAS | Conversion value at an average return | You have steady data and a clear return target | A target set too high can restrict traffic and volume |
Google’s Maximise conversion value help page suggests using Target ROAS if you have a specific return goal. Its Target ROAS guidance also notes that lowering the target tends to increase volume, while raising it tends to increase efficiency at the risk of entering fewer auctions.
Is your account ready for value-based bidding?
Value-based bidding works best when certain conditions are in place. Before switching, check that:
- Conversion tracking is accurate. Duplicate conversions, test submissions or spam leads will mislead the model.
- You have enough conversions with values. Google’s Target ROAS documentation recommends at least 15 conversions in the last 30 days at the conversion tracking level, and more data generally produces more stable results.
- Values genuinely differ. If every sale is roughly the same, value-based bidding adds complexity without much benefit.
- Values reflect profit, not just revenue, where possible. A high-revenue, low-margin product may be worth less to you than its price suggests.
- You can wait for learning. Strategy changes need time to settle, so avoid switching just before a peak season such as the year-end sales.
Assigning values when you sell services, not products
An online shop can pass the order total automatically. Most Singapore SMEs, however, generate leads by form, phone or WhatsApp and close the sale later. Value-based bidding still works, but you need a reasoned estimate of what each type of lead is worth.
A practical formula is: lead value = likelihood the lead becomes a customer × average gross profit per customer. The table below is an illustrative example for a fictional renovation firm, not real data.
| Conversion type | Illustrative close rate | Illustrative average gross profit (SGD) | Value to report (SGD) |
|---|---|---|---|
| Whole-flat renovation enquiry | 10% | 8,000 | 800 |
| Kitchen-only enquiry | 15% | 2,000 | 300 |
| Minor repair enquiry | 30% | 200 | 60 |
| Brochure download | 2% | 5,000 | 100 |
Even rough figures like these give Smart Bidding far more useful information than treating every lead as equal. Revisit them each quarter using your actual sales records.
Refining values with rules and offline data
Two features make lead values more accurate over time:
- Conversion value rules. Google’s page about conversion value rules explains how to adjust values by conditions such as location, device or audience. For example, if enquiries from a particular region consistently turn into larger jobs, you can raise their value.
- Offline conversion imports and enhanced conversions for leads. These connect a lead that started with an ad click to what actually happened later in your CRM, so you can report qualified leads and closed sales with their real values. Google’s conversion values best practices recommend uploading such data frequently, ideally daily.
Enhanced conversions use hashed customer information such as email addresses. Make sure your privacy notice and consent practices cover this use under the PDPA; our guide to PDPA rules for marketing sets out the basics. To get the measurement foundations right first, see our guide to measuring digital marketing leads in Singapore.
A step-by-step plan for moving to value-based bidding
- Week 1 – Clean up conversions (owner: marketing lead). Remove duplicates, mark soft actions as secondary and make sure each primary action represents a real lead or sale.
- Week 1–2 – Estimate values (owner: business owner with sales or finance). Use the last six to twelve months of sales records to estimate close rates and average gross profit for each lead type.
- Week 2 – Report values (owner: marketing lead with web developer). Add fixed or dynamic values to each conversion action, then check they appear correctly in Google Ads reports.
- Weeks 3–6 – Collect value data (owner: marketing lead). Keep your current bid strategy while values accumulate, so the model has history to learn from.
- Week 6 – Switch or test (owner: marketing lead, approved by business owner). Move to Maximise conversion value, or run a campaign experiment comparing it with your current strategy.
- Weeks 7–12 – Add a target and refine (owner: marketing lead). Once results settle, consider a target ROAS close to what the campaign already achieves, add offline imports and review values quarterly.
If you run Performance Max, the same principles apply; our guide to optimising Performance Max campaigns covers the campaign-specific details.
How AI can help with value-based bidding
Value-based bidding is already an AI-driven system, so the most useful extra help is in feeding it better information and checking what it does. AI tools can help analyse past sales to estimate lead values, match CRM records to ad clicks, flag sudden changes in return on ad spend and summarise which campaigns produce high-value customers.
Keep a person in charge of the inputs. A model can only learn from the values you give it, and a wrong value spreads quickly into bidding. Budget and target changes should be approved by someone who understands the business. Marketing2u’s AI Ads Agent supports this by syncing Google Ads data read-only and flagging anomalies and wasted spend for review, while the AI Sales Agent scores incoming leads from 0 to 100 with a stated reason, which can inform how you value different lead types.
Common value-based bidding mistakes
- Switching before you have enough data. A strategy with too few valued conversions behaves erratically.
- Using made-up or flat values. Giving every lead the same value adds nothing; giving values with no basis in sales records misleads the system.
- Setting the target ROAS too high on day one. An ambitious target can starve the campaign of traffic. Start close to current performance.
- Counting revenue when margins differ widely. Optimising for revenue can push spend towards low-margin products.
- Ignoring lead quality after the click. Without offline data, the model cannot learn which leads became customers.
- Judging results too quickly. Allow time for learning, and compare like-for-like periods, not a festive month against a quiet one.
Is value-based bidding worth it for your business?
Value-based bidding is worth considering when your customers vary meaningfully in value and you can measure that value with reasonable accuracy. It lets Google Ads spend more where better customers are likely and less where they are not. It is not a shortcut: it depends entirely on clean tracking, honest values and enough data.
A sensible next step is to list your conversion types and write down what each is really worth to you, using last year’s sales. That exercise is useful even if you never change your bid strategy, and it is the foundation value-based bidding needs. To plan the spend around it, see our guide on setting a PPC budget.
Frequently asked questions
What is the difference between value-based bidding and target CPA?
Target CPA treats every conversion as equal and aims for an average cost per conversion. Value-based bidding uses the value you assign to each conversion and aims to maximise total value or hit a return on ad spend target. If your customers differ a lot in value, value-based bidding gives Google more useful information.
Can I use value-based bidding if I do not sell online?
Yes. Lead-generation businesses can assign estimated values to each type of enquiry based on close rates and average profit, then refine them with offline conversion imports from a CRM. The values do not need to be perfect, but they should be based on real sales records.
How many conversions do I need for Target ROAS?
Google’s Target ROAS documentation recommends at least 15 conversions in the last 30 days at the conversion tracking level. More conversions generally give the model more to learn from and steadier results. With less data, Maximise conversion value without a target is usually the safer starting point.
What is a good target ROAS?
There is no universal figure. A useful starting point is the return your campaign already achieves, adjusted gradually towards a level that covers your margins and overheads. Setting the target far above current performance often reduces traffic and total value.
Sources and further reading
- Google Ads Help: About Smart Bidding using value-based bidding
- Google Ads Help: About Target ROAS bidding
- Google Ads Help: About Maximise conversion value bidding
- Google Ads Help: About conversion value rules
- Google Ads Help: Conversion values best practices
Related guides
- GA4 E-commerce Tracking: A Setup Guide for Online Stores
- 12 Google Ads Bidding Strategies For Every Campaign Type
- What Is Return On Ad Spend (ROAS): Guide To Smarter Ad Spend
- 10 Ways To Optimise Performance Max Campaigns In Google Ads For Real Results
- What Is PPC Bidding and How Does It Work?
Browse every article in our Paid Advertising guides.