Target return on ad spend (ROAS) is heralded as the holy grail of PPC advertising.
In contrast to click-based and conversion-based bidding, this strategy is designed to optimize financial business outcomes.
While undoubtedly high on the Google Ads evolutionary scale, Target ROAS requires a rigorous setup before realizing its full potential.
Advertisers who switched from Target CPA to Target ROAS saw a 14% increase in conversion value at a similar return on ad spend, per Google’s internal data from March 2021.
Google claims that advertisers who upgraded from Smart Shopping to Maximize Conversion Value and Target ROAS achieved as much as a 30% uplift in conversion value.
Results can vary from one business to another and across different sectors of the economy.
Your success with value-based bidding depends on how well it aligns with your business model and implementation quality.
This article outlines the key considerations in adopting Target ROAS to help you evaluate whether value-based bidding is a good fit for your business.
Target ROAS: An overview
Target ROAS, or tROAS, is a Google Ads value-based bidding strategy designed to maximize conversion value within your targeted return on ad spend.
As a Smart Bidding strategy, Target ROAS leverages a multitude of contextual and audience signals in combination with historical first-party data.
Google uses its advanced predictive modeling to estimate the prospective conversion value of users and automatically adjusts your bids in line with your ROAS target.
The higher you set your target, the lower the AI will bid and vice versa.
In practice, some conversions will yield a higher ROAS than others. Google then factors that into its calculations and recalibrates bidding to maintain your desired ROAS.
Bidding based on potential customer value
![An example of value-based bidding. Three values are attributed to three different customers: £100, £300 and £500. Under conversion bidding, companies bid £10 each. Under value-based bidding, companies would bid £5, £10 and £15 respectively.](https://searchengineland.com/wp-content/seloads/2023/09/image-41.png)
Upgrading from a conversion-based to a value-based strategy represents a shift from optimizing for the highest number of conversions to prioritizing the most valuable customers, according to Ginny Marvin, Ads Product Liaison at Google.
As a consequence of this shift, advertisers should anticipate a trade-off between volume and value.
Therefore, Target ROAS is generally more likely to return a higher total conversion value but lower conversion volume than Target CPA.
Here are five key considerations to help assess your business readiness for tROAS in Google Ads.
1. Variability in sales value
Before delving into the more technical requirements of value-based bidding, it might be useful to weigh up the size of the opportunity.
Considering the variability in sales value will give you some indication of the potential upside that Target ROAS could bring to your business.
At its core, value-based bidding aims to optimize toward high-value conversion outcomes and away from low-value conversion outcomes.
If your business has high variability in sales value within the same product or service category, you’re better placed to reap the rewards of Target ROAS.
Consider an ecommerce store selling products worth $20, $50 and $100.
All things being equal, this store is more likely to benefit from value-based bidding than a store that only sells products worth $50. That’s because the algorithm can drive more $100 sales and fewer $20 sales.
The disparity in value creates the opportunity to optimize for more valuable conversion outcomes.
Conversely, the store in the second example lacks the same capacity for optimization as all products are worth the same.
Value-based bidding is further amplified by greater variance or spread in conversion value.
In our initial example, there’s a moderate level of variance. For instance, the variance would be considerably higher if the products were worth $5, $50 and $500.
A broader distribution in conversion values gives the AI more room to find efficiencies and maximize overall conversion value.
The principle of variability applies to any assigned conversion value, be it revenue, gross profit, or another value estimate unique to your business.
Using value-based bidding in low variability scenarios
What if your products or services are priced similarly? Could you still benefit from value-based bidding?
Even if your prices are uniform, the profit margins may differ. Different customers may buy varying quantities at different frequencies and repeat rates.
In other words, if conversion value variability is low from a revenue perspective, it may not be through the lens of gross profit or customer lifetime value (CLV). We’ll explore the implications of each of these options shortly.
Suppose every sale in your business generates the same conversion value, irrespective of the financial measure you associate with it.
In this scenario, you would assign an identical value to each conversion.
This is similar to Target CPA, but instead of telling Google what you’re willing to pay for a conversion, you’re defining how much a conversion is worth and using the ROAS target as your lever.
The main difference is that with value-based bidding, your bids are pegged to your returns.
The AI will automatically adjust bids with the conversion value against your ROAS target.
Therefore, Target ROAS provides an automation benefit, even when conversions don’t fluctuate in value.
2. Sales volume
Another key consideration is the number of sales your business generates each month.
This will tell you whether you can accumulate sufficient conversion data on an ongoing basis to meet the minimum conversion thresholds.
Target ROAS requires minimum conversion thresholds to provide Google with adequate data to make statistically reliable bidding decisions.
This data allows the AI to spot patterns, establish correlations, and draw meaningful insights that fuel machine learning.
Without sufficient conversion data, the AI would base its analyses on smaller and potentially unrepresentative samples, which could compromise its predictive capabilities.
Larger data sets provide Google more opportunities to learn and bid more effectively.
Most campaign types require at least 15 conversions per campaign in the previous 30 days to run Target ROAS.
However, minimum thresholds can vary by campaign type, as shown in the table below. Note that your conversions must include valid values to qualify toward the threshold.
Minimum Target ROAS conversion thresholds by campaign type
Campaign type | Minimum conversions | Qualification period |
Search campaigns | At least 15 conversions | Last 30 days |
Shopping campaigns | At least 15 conversions | Last 30 days |
Display campaigns | At least 15 conversions | Last 30 days |
Video action campaigns | At least 30 conversions | Last 30 days |
Discovery campaigns | At least 75 conversions | Last 30 days |
App campaigns | At least 300 conversions | Last 30 days |
For new or small campaigns with insufficient conversion data, you can start with Maximize Conversion Value, which has no minimum requirements.
Then upgrade to Target ROAS once you reach the required threshold for your respective campaign type.
An important decision is what conversion event to use as your primary conversion action.
When deciding, you must consider your sales volume in conjunction with the length of your sales cycle.
Get the daily newsletter search marketers rely on.
3. Sales cycle length
The duration of your business cycle will dictate the speed at which you can assign values to different conversions and feed those back into Google.
The faster you can import conversion values, the sooner the AI can factor that data into its computations.
Sales cycles can vary significantly depending on the industry you’re in.
In lead gen, the B2B sales cycle is typically between 60 to 90 days due to the cost, complexity and multiple decision-makers involved.
B2C tends to have a much shorter cycle, ranging from a few minutes to a few weeks.
In contrast, B2B ecommerce generally has a shorter cycle than traditional B2B sales.
Finally, B2C ecommerce would see the shortest sales cycle due to its direct-to-consumer transactions and relatively lower cost.
Consider the average time your prospect clicked your ad and the conversion event. Note that the conversion needs to occur within 90 days of the click. Otherwise, it would fall outside the lookback window, and Google won’t be able to attribute that value back to the original click.
You must choose the primary conversion action most likely to drive optimal performance.
As a rule of thumb, Google recommends optimizing for the conversion event that occurs furthest down your marketing funnel and meets the minimum eligibility criteria.
Here are some general guidelines to inform your decision
Scenario 1: High sales volume, short sales cycle
If your sales cycle is around two weeks or less and you generate at least 100 sales a month, then you’re in a position to optimize for sales or closed deals.
If you’re running lead gen campaigns, consider secondary observational conversion actions, such as:
- Sales Qualified Leads (SQLs).
- Marketing Qualified Leads (MQLs).
- Form submissions.
- Calls.
Alternatively, if you’re running ecommerce campaigns, consider including begin checkout, add to cart, or newsletter subscriptions as secondary conversions.
Scenario 2: Low sales volume, long sales cycle
If your lead gen sales cycle is up to three months and you generate at least 30 sales a month, consider using either SQLs or MQLs as your primary conversion action.
It may also be worthwhile to include sales, form submissions, and calls as secondary conversions for enhanced visibility of your funnel.
In ecommerce, consider using either begin checkout or add to cart as your primary conversion action, while adding sales and subscriptions as secondary conversions.
Scenario 3: Low sales volume, long sales cycle, and long lead qualification time
If your lead qualification time takes more than 30 days:
- Use form submissions and calls as primary conversions.
- Consider importing sales, SQLs, MQLs, and page interactions as secondary conversions.
Recommended conversion actions in lead generation
Scenario | Sales Cycle | Sales vol. | Primary | Secondary |
High sales volume, short sales cycle | ~2 weeks | 100/month | Closed deals | SQLs/MQLs, form subs. and calls |
Low sales volume, long sales cycle | 3 months | 30/month | SQLs/MQLs | Closed deals, form subs. and calls |
Low sales volume, long sales cycle + long lead qual. time (~30 days) | 3 months | 30/month | Form subs. and calls | SQLs/MQLs, closed deals, page engagement |
Recommended conversion actions in ecommerce
Scenario | Sales Cycle | Sales vol. | Primary | Secondary |
High sales volume, short sales cycle | ~2 weeks | 100/month | Sales | Add to cart, begin checkout |
Low sales volume, long sales cycle | 3 months | 30/month | Begin checkout, add to cart | Sales, newsletter subscriptions |
Google recommends importing your entire conversion funnel for improved visibility while focusing on a single primary conversion action for bid optimization.
An exception to this could be using both form submissions and calls as primary conversion actions, provided there is no duplication
Mapping your marketing funnel can be a good starting point to visualize your prospects’ key actions along their conversion path.
Google’s conversion value calculator provides a quick way to calculate the average values of the conversion actions at every stage of your customer journey.
4. Data accuracy
So far, we’ve emphasized the importance of variability, volume, and frequency. However, the quality of your data will determine the degree of your success.
In the context of Target ROAS, data quality is the extent to which your conversion values accurately reflect their economic worth to your business.
Target ROAS relies on accurate input to deliver your target return on ad spend.
If your conversion values don’t correspond to their true monetary value, neither will the AI’s bidding nor your campaign results.
The “garbage in, garbage out” principle firmly applies here. No matter how advanced the algorithm, a low-quality input is unlikely to produce a high-quality output.
This leads us to the next important decision: What conversion values do you feed the AI?
As a general rule, your acquisition strategy should align with your business goals. Here are several approaches you could take:
Optimizing for proxy values
If you can’t measure or assign transaction-specific values, you can still run Target ROAS using static proxy values.
This is straightforward as it doesn’t necessitate a complex conversion tracking configuration.
Instead, you assign a fixed value to your primary conversion actions, meaning every conversion will account for the same value.
However, you can dynamically adjust your values using rules based on criteria such as location, device, or audience.
If your sales values typically fluctuate, proxy values won’t accurately reflect the true economic value of your conversions.
As such, using proxies is the simplest but most limiting way to go about value-based bidding.
Optimizing for revenue
If your business goal is to maximize the gross value of sales, consider using revenue conversion values. This will require importing dynamic conversion values and an accurate estimate of the revenue driven by each conversion.
By aligning your value-based bidding with revenue, the AI will aim to maximize the total revenue generated within your ROAS target.
Besides driving top-line growth, this may also be suitable for market share expansion or promoting a new product.
A drawback of revenue optimization is its neglect of profitability. They will likely have different profit margins if you offer a wide range of products or services.
However, this difference will not be considered by the AI, which could lead to an overemphasis on high-revenue but low-margin products or services.
Optimizing for profit
If your business prioritizes the bottom line, consider assigning values that closely mirror your gross profit. To calculate gross profit, deduct the cost of goods sold from your sales revenue.
Since ad spend is also a cost of sale, you can use custom columns in your Google Ads account to subtract ad spend from your conversion value (i.e., All conversion value – Cost).
Note that Target ROAS will still optimize for the value in the All conversion value column.
By optimizing for profit, the AI will deploy your budget towards the most financially favorable outcomes.
In the short term, this should yield the highest total gross dollar amount, assuming accurate values, sufficient volume, and timely data imports.
Remember that optimizing for profit could come at the expense of lower conversion volumes.
Moreover, focusing on profit may overlook potential opportunities to grow your customer base or expand your reach.
Lastly, measuring and tracking the true profitability of each conversion can be especially challenging due to the various cost factors at play.
Optimizing for customer lifetime value (CLV)
If you’re aiming to maximize long-term profitability, consider using predicted customer lifetime value (CLV).
This requires assigning a forecasted value to each conversion based on the total expected value over the entire course of the customer relationship.
CLV generally incorporates average order value, purchase frequency, retention rate, and customer acquisition and retention costs. The relative weighting and exact calculation methods can vary across industries.
Like optimizing for short-term profit, this will also likely limit your reach to a smaller conversion pool. Furthermore, accurately estimating long-term profit can be exponentially more complex.
Over the long run, CLV optimization has the potential to deliver the highest return on investment.
But tread carefully. This strategy banks on spending money today and recovering it years into the future.
The delay in feedback on financial performance could prove costly should your initial projections turn out to be incorrect.
While CLV offers significant potential upside, it also comes with considerable uncertainty and upfront costs, making it something of a leveraged bet.
Given these risks, it may be prudent to test CLV-based bidding only after successfully validating a proof of concept using Target ROAS aligned with revenue or profit.
5. Data infrastructure
Hopefully, by now, you have an idea of how to tackle value-based bidding for your specific use case. Assuming your business ticks all the boxes, the next key consideration is data logistics.
More specifically, what systems will you need in place to streamline your marketing data and does your business have the capacity to accommodate your requirements?
Implementing Target ROAS will require a reliable way to collect, store and import data back into Google regularly. You can do this manually, automatically, or combine the two depending on your strategy.
Here are the three main tracking options available:
Manual conversion tracking
Manual conversion tracking allows you to assign a static conversion value for each conversion action inside Google Ads.
This can easily be set and modified on the platform without the need for technical expertise or third-party software.
As noted earlier, this is an imprecise way to track value since static conversion values don’t account for variations in purchase value.
Assuming that your conversion values fluctuate, this is why it’s a suboptimal conversion tracking method.
Tag-based conversion tracking
Tag-based conversion tracking relies on a Google-generated JavaScript code snippet (the “tag”) embedded on your website.
When a user completes a conversion action, the tag captures the associated conversion value and sends it back to Google.
Ecommerce businesses most commonly use this tracking method as it provides a way to dynamically adjust the conversion value to match the actual order value.
This usually means that the conversion values correspond to revenue rather than profit, as the tag pulls the monetary amount the customer spent on a transaction.
It’s also possible to track profit using the tag-based method as long as the profit value is known and accessible when the tag fires.
This may require deep integration with your inventory systems and third-party software to accurately calculate the profit for every sale in real time.
So, while it’s possible to track profit, it may not be practical or feasible for most businesses due to the added complexity involved.
Setting up tag-based conversion tracking requires technical proficiency and can be challenging for businesses with various products or services.
Another limitation of tag-based tracking is its dependence on cookies to attribute conversions back to the ad click.
When a user rejects, blocks, or deletes cookies, this can result in data gaps, negatively impacting your optimization.
Find out more about tag-based conversion tracking in this Google Ads help documentation.
Offline conversion tracking
Offline conversion tracking uses a Google Click Identifier (GCLID) to track offline conversion outcomes following a user’s interaction with your ad.
The GCLID is a unique string of characters that Google automatically appends to your destination URLs.
This tracking method requires that you capture the GCLID parameter alongside the lead or customer’s details and store them in your CRM database.
You can import that data back to the platform once a conversion value is assigned. Google will then use the GCLID to associate the conversion value back to the correct click.
You can import offline conversions manually inside the Google UI or schedule a recurring upload via Google Sheets, HTTPS or SFTP.
Alternatively, you can automate this process using Google Ads API, which would require developer input.
It’s worth checking whether your existing CRM can be directly integrated with Google Ads, as this could save you significant time and effort.
Offline conversion tracking can be a reliable and comprehensive way to track conversion outcomes.
It also offers flexibility when assigning values that best align with your business objectives.
Moreover, it allows you to retract and restate values you’ve already uploaded to reflect returned orders, canceled bookings, or failed deals.
Depending on the degree of complexity, you may require technical resources to get this up and running.
A downside of this approach is its reliance on a clear connection between the ad click and the offline conversion.
In practice, this will not always be possible due to the length of the customer journey or the nature of the conversion itself.
Regardless of the tracking method you choose, it’s important to ensure that your processing of user and customer data complies with local and international data protection and privacy laws in your jurisdictions.
The primary purpose of importing first-party conversion data into the platform is to guide the AI’s bidding decisions.
However, linking the conversion value to the exact click that drove it also unlocks Google Ads’ full reporting capabilities.
This allows you to track profitability down to granular details such as search terms, ads, or placements, to name a few.
Assessing your business’s readiness for Target ROAS
A successful value-based bidding strategy comes down to your data’s variability, volume, speed, and accuracy and the infrastructure needed to support your marketing operation.
- Variability is the degree to which your conversions fluctuate in value.
- Volume refers to the amount of conversion data that you generate.
- Speed measures how promptly you can feed the data back to the AI.
- While accuracy is the extent to which your data reflects the true economic value to your business.
- Your infrastructure represents the technical foundation to collect, store, and import conversion data to the Google Ads platform.
We’ve established that AI thrives on data, but to fully harness Target ROAS, it’s vital to strike the right balance between quality and quantity.
While Google recommends optimizing for the conversion furthest down your funnel that meets the eligibility criteria, that may not always be the best approach.
Depending on the quality of your data, you may see better results by optimizing for a conversion higher up the funnel that provides the AI with more data points.
Sometimes, feeding the AI with a lot of ‘good’ data can outweigh feeding it the bare minimum of “great” data.
Equally, a smaller pool of accurately calculated conversion values may outperform a larger pool of less accurately calculated ones. It’s up to you to adapt your strategy to your unique business circumstances and test your way to success.
So, to ROAS or not to ROAS? That is the question only you can answer.
Theoretical evaluation is a good place to start. But to know how effective value-based bidding can be, you’d have to press that live button and find out.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.
Related stories
New on Search Engine Land