Value-Based Bidding in Google Ads: A Plain-English Guide
. By Ads Daddy.
Not all conversions are equal. A $20 sale and a $2,000 sale both show up as "1 conversion" in a count-based setup. Value-based bidding fixes that. You tell Google what each conversion is actually worth, and the algorithm chases the valuable ones instead of just the numerous ones.
Done well, it stops you paying premium prices for tyre-kickers. Done badly, it feeds Google junk numbers and burns your budget with total confidence. This guide shows you which one you'll get.
TL;DR
- Value-based bidding means Google optimises towards the total value of your conversions, not the number of them. It covers Maximise conversion value and Target ROAS.
- Your values are the steering wheel. Bad values in, bad bids out. Use real revenue or margin, or a sensible proxy for leads.
- Check readiness first. Google's guidance for Target ROAS is at least 15 conversions in 30 days, plus about 3 weeks of uploaded values.
- Test it properly. Run a campaign experiment for at least 30 days. Don't judge it after four days of panic.
- CPA still has a job. When volume matters more than value, it's the better tool.
Table of Contents
- What value-based bidding actually is
- Maximise conversion value vs Target ROAS
- Target CPA vs Target ROAS
- Are you ready? The data check
- Giving your conversions a value
- Value rules: nudging the algorithm
- How to test it without wrecking your account
- Common mistakes (we've seen all of them)
- Perspective
- Want us to run this for you?
- FAQ
What value-based bidding actually is
Value-based bidding is a subset of Google's Smart Bidding. Instead of asking "how do I get the most conversions?", it asks "how do I get the most value from this budget?" Google's own help documentation describes it as optimising for the value your campaigns bring to your business, either by maximising conversion value within a budget or by using Target ROAS.
Think of it like dating apps. A count-based strategy swipes right on everyone. Value-based bidding is the friend who says, "Mate, that one has a mortgage broker's salary and a boat. Prioritise."
Under the hood, Google uses auction-time bidding. For every single auction, its machine learning predicts how likely a click is to convert and how much that conversion might be worth. It sets a unique bid using signals like time of day, browser and operating system, across millions of bids every second. You can't out-click that by hand. What you can do is give it better information.
That's the whole game. The algorithm is the engine. Your conversion values are the fuel.
Maximise conversion value vs Target ROAS
These are the two value-based strategies in Google Ads, and people mix them up constantly.
Maximise conversion value spends your daily budget and tries to get the highest total conversion value it can. No efficiency target. Just "get me the most value for this money."
Target ROAS (return on ad spend) does the same, but with a leash. You set a target, say 400%, meaning $4 of conversion value for every $1 of ad spend. Google then looks for as much value as it can while holding that line. If performance drifts below the target, the system may lower bids to pull it back.
When to use which:
- Start with Maximise conversion value if you want to spend the budget and you're new to value-based bidding. It's the gentler on-ramp.
- Move to Target ROAS when you have a specific efficiency requirement to hit, such as a margin you can't dip below.
- Be careful with aggressive targets. Set the ROAS too high and Google simply stops bidding. You get a lovely ratio and almost no sales. A high ROAS on four orders is not a business.
If you want a refresher on the metric itself, our post on what ROAS is and how it relates to profitability covers the maths.
Target CPA vs Target ROAS
This is the question behind "CPA vs ROAS" and "tCPA vs tROAS". Short answer: CPA treats every conversion the same. ROAS doesn't.
Target CPA optimises for conversion volume regardless of value. Target ROAS chases as much conversion value as possible at a set efficiency. Google's guidance is to choose CPA when volume matters more than telling conversions apart by value.
| Target CPA | Target ROAS | |
|---|---|---|
| What it optimises | Number of conversions | Total conversion value |
| Your target is | Cost per conversion | Value per dollar spent |
| Treats all conversions as equal? | Yes | No |
| Best when | Every lead or sale is worth about the same | Orders or leads vary a lot in value |
| Needs value data? | No | Yes, and it must be accurate |
| Biggest risk | Cheap, low-quality conversions | Bad values or a target set too tight |
A quick gut check. If every enquiry you receive is worth roughly the same, stick with CPA. If your average order sits at $60 but a chunk of customers spend $600, CPA is blind to the difference. ROAS isn't.
We go deeper on picking between strategies in how bidding strategies can maximise your ad results.
Are you ready? The data check
Value-based bidding is not a magic switch. It's a hungry algorithm, and it needs feeding before it performs.
Google's readiness guidance says you're ready to test when you've uploaded values for at least 3 weeks, or 1 to 2 conversion cycles, whichever is longer. For Target ROAS, you also want at least 15 conversions in 30 days at the conversion tracking level.
Demand Gen campaigns are hungrier. Google's guidance asks for at least 50 conversions with value in the past 35 days, including 10 in the last 7 days, or 100 across all Demand Gen campaigns in the account over 35 days.
Your tracking checklist:
- Track conversions properly with the Google tag or Google Tag Manager.
- Report at least two distinct, non-zero values across your conversion goals. If everything is worth $1, there's nothing to learn from.
- Use enhanced conversions for more accurate measurement.
- Send data often, ideally daily, so the model isn't steering on stale numbers.
- Skip zero-value conversions. Google advises removing them or using conversion adjustments instead.
All of the above is in Google's value-based bidding best practices. If your tracking is shaky, fix that first. Our guide to conversion tracking for Aussie e-commerce is a good place to start.
Giving your conversions a value
This is where most accounts fall over. Not the bidding. The values.
E-commerce: easy mode
If you run a Shopify store, the purchase value is already in the cart. Pass the revenue through. Better still, pass margin, because a $200 sale at 10% margin is worse than a $90 sale at 60%. Sorry, big baskets.
Lead gen: where it gets interesting
A form fill has no price tag. So you invent a sensible one. Google suggests starting from what a conversion is actually worth, or using proxy values such as lead scores when exact figures are hard to pin down.
To estimate a lead's value, Google's guidance points to deal revenue, profit margin and lead-to-deal rates, plus extras like repeat business, word of mouth and lifetime customer value. No exact numbers? Use conservative estimates. Google says an estimate beats having no value at all.
A worked example. Say you're a finance broker:
- Average commission per settled deal: $3,000
- One in 10 qualified enquiries settles
- Estimated value of a qualified enquiry: $300
- A raw, unqualified form fill that rarely progresses: $30
Now Google has a reason to prefer the $300 lead over ten $30 ones. Same ad spend, very different business.
The next level is importing offline outcomes. When your CRM says a lead became a paying customer, send that value back to Google. That's when the algorithm learns what a good lead really looks like, not just what a form fill looks like. Our piece on why you should track assisted conversions explains why the supporting actors deserve credit too.
Value rules: nudging the algorithm
Sometimes you know something Google doesn't. Maybe customers in one city are worth more, or one device brings higher-spending buyers. That's the job of conversion value rules in Google Ads. They let you adjust the value assigned to a conversion based on audience, location or device, so bidding reflects what you know about your business.
Use them sparingly. Start with one rule built on data you trust, like a location where average order value is clearly higher. Pile on five guesses and you've built a Jenga tower.
How to test it without wrecking your account
Switching strategies mid-campaign and hoping for the best is the marketing equivalent of changing your car's engine at highway speed. Use a campaign experiment instead.
Google's guidance for testing value-based bidding with campaign experiments is:
- Split your campaign into a base arm (your current strategy) and a trial arm (value-based bidding).
- Allow a ramp-up of 2 weeks or 1 to 2 conversion cycles, whichever is longer.
- Run it for at least 30 days.
- Ignore the most recent days where fewer than 90% of conversions have been reported yet.
- Avoid budget constraints on Target ROAS tests. If the budget is the bottleneck, you're testing the budget, not the strategy.
Then judge on what matters to your bank account: total value, profit and cost per outcome, not just the ROAS figure.
For wider context on squeezing more from your spend, see our steps to optimise ad campaigns for higher ROI.
Common mistakes (we've seen all of them)
- Uploading fake precision. Values like $47.13 invented from thin air. Keep it simple and honest.
- Leaving zero-value conversions in. They dilute the signal.
- Setting Target ROAS from a wish. Base it on your actual history and margins, not a number that looks good in a pitch deck.
- Starving the campaign. Tight budgets plus tight ROAS equals very little spend and very few sales.
- Changing everything at once. New creative, new targets and new strategy in one week means you'll learn nothing.
- Measuring after three days. The algorithm is still learning. Be patient, then be ruthless.
- Optimising for the wrong value. If revenue looks great but margin is thin, you're growing a business that loses money faster.
For how this fits with audience selection, have a look at our notes on ad targeting strategies for better ROI.
Perspective
Most businesses that come to us aren't short of clicks. They're short of the right conversations. Value-based bidding is the cleanest way we know to point Google at the customers who matter, but only when the numbers going in are honest.
What we do differently is boring on purpose. We sort tracking first, agree on what a lead or sale is genuinely worth, and only then let the algorithm loose. For lead gen, that often means feeding back what happened after the form fill, not just the fill itself.
Our view: you shouldn't have to become a Google Ads mechanic to benefit from this. That's the point of hands-on management with integrated tracking and lead delivery. You run the business. We babysit the bidding.
Adrian
Want us to run this for you?
Value-based bidding rewards the people who set it up properly, and punishes everyone else. If you'd rather skip the learning curve, our team manages Google Ads, Meta Ads, YouTube Ads, LinkedIn Ads and Microsoft Ads with the tracking sorted from day one.
Running lead gen in finance? Have a look at our lead generation service and our pay-per-lead options, with car finance leads from A$70 per lead and ABN finance leads from A$170 per lead. Book a call and we'll tell you straight whether value-based bidding suits your account right now.
FAQ
What is a value-based bidding strategy?
It's a Google Ads approach where the system optimises for the total value of your conversions, not just how many you get. The two main options are Maximise conversion value and Target ROAS. You supply the values, and Google adjusts bids auction by auction.
Should I use Maximise conversion value or Target ROAS?
Start with Maximise conversion value if you want to spend your budget and get comfortable with value-based bidding. Switch to Target ROAS when you have a specific return you must hit. Just don't set the target so high that Google stops spending.
What's the difference between CPA and ROAS bidding?
CPA bidding aims for conversions at a target cost and treats each one as equal. ROAS bidding aims for conversion value at a target return, so it can tell a big sale from a small one. Use CPA when conversions are worth about the same, and ROAS when they aren't.
How much data do I need before I switch?
Google's guidance is to upload values for at least 3 weeks or 1 to 2 conversion cycles, whichever is longer. For Target ROAS, aim for at least 15 conversions in 30 days. If you're below that, tighten up your tracking before testing.
Can lead generation businesses use value-based bidding?
Yes. Assign each lead type a value based on deal size, close rate and margin, then use conservative estimates where you can't be exact. Even better, import offline results from your CRM so Google learns which leads actually became customers.