A Cheaper Lead Is Not Always a Better Lead

In banking, insurance and real estate, acquisition campaigns are often optimised to generate as many form submissions as possible at the lowest cost. But this approach relies on a shortcut: assuming that all leads are equal. Working with Allianz, Keyade's teams tested a different approach—giving algorithms the means to prioritise the prospects that contribute most to profitability.

When a Conversion Has No Immediate Value

For an e-commerce website, the value of a conversion is relatively easy to measure. When a user buys a pair of shoes, the purchase price can be sent directly to the advertising platform. The algorithm then knows that a €500 sale does not generate the same value as a €100 sale.

The situation is more complex for businesses that generate leads.

What is the value of a completed insurance form? A loan application? A property valuation request? A sales appointment? At the point of conversion, no revenue has necessarily been generated yet. The prospect's value will only become known later, once they have been qualified or converted, or once their contribution to profitability has been analysed.

Without this information, campaigns are often managed using two metrics: the number of leads generated and their average cost.

Yet Not All Leads Contribute to the Business in the Same Way

This approach leads platforms to treat every completed form in the same way. A lead with little likelihood of becoming a customer counts just as much as a prospect with strong commercial potential.

The algorithm then performs the task it has been given perfectly: generating as many conversions as possible at the lowest cost. But it cannot determine whether those conversions are genuinely valuable to the business.

A campaign can therefore reduce its cost per lead while attracting more prospects who are less likely to convert, purchase less profitable products or generate less value over time.

CPL remains a useful metric, but it only tells part of the story. The cheapest lead is not necessarily the best lead.

For marketing leaders, the real challenge is to bring media performance and commercial performance closer together: rather than simply asking how many leads were generated, they need to identify which ones genuinely create value.

The Solution: Teaching Algorithms What Really Matters

Value Based Bidding addresses this challenge. Rather than simply asking the algorithm to maximise the number of conversions, it encourages it to prioritise those that are most valuable to the business.

The value passed to the platform may correspond to revenue, margin, customer lifetime value or an assessment of a prospect's potential. When the precise financial value of a lead is not immediately available, a weighting that reflects the advertiser's business priorities can also be used.

But activating a bidding strategy in Google Ads is not enough. The real work lies in building the signal that will enable the algorithm to make better decisions.

This first requires an understanding of the advertiser's business: which profiles convert most successfully? Which products, needs or characteristics are associated with greater profitability? What information is already collected through forms, the CRM or sales tools?

The next step is to verify that the platform has sufficiently robust signals to identify the target profiles. A segmentation may be highly relevant within a customer database without being effectively actionable in a media environment.

Finally, the data must be able to flow between systems. The website must distinguish between conversions, assign them a value and pass this information correctly to the platform.

The success of the initiative therefore depends on collaboration between several areas of expertise:

• marketing and acquisition teams, which define the media objectives;

• sales teams, which understand the true value of prospects;

• data analysts, who identify performance differences between segments;

• technical teams, which organise tagging and data flows;

• the agency, which connects business priorities, data and media activation.

"The agency's role is to bridge the gap between tech, business and media," says Axel Montu, SEA Client Lead at WPP Media/Keyade. This interface role is essential: activation is only the culmination of a collective effort to identify the relevant data, make it actionable and measure its actual impact.

This is precisely the approach taken by Axel and the teams working alongside Allianz.

For its SEA campaigns, Allianz wanted to increase the share of a specific type of lead in its home insurance campaign.

Before Value Based Bidding was activated in Google Ads, all completed forms were treated in the same way.

To understand what the initiative was intended to change, let us consider two fictional characters: Hélène and Théo.

Both are searching for home insurance on Google. Hélène is a homeowner who lives with her partner in a third-floor apartment in Nantes. Théo, aged 35, lives with his family in a rented house in Évreux. They enter the same search query, but will not necessarily see the same result. Using the available signals and the values passed through the campaign, the algorithm can prioritise serving an ad to the profile that offers greater value to Allianz.

In the test, a high optimisation value was assigned to the type of lead that Allianz wanted to increase, while a lower value was assigned to other leads. These figures do not represent the prospects' actual financial value. They serve as an instruction to the algorithm, indicating which type of conversion to prioritise without excluding other profiles.

At the end of the test, which used Allianz CRM data from June–July 2025, the proportion of the target leads increased.

Cost per lead rose by +20%. Viewed in isolation, this metric could suggest that performance had deteriorated. However, Allianz's analyses confirmed that the higher value of the leads generated offset this increase and ensured the initiative remained profitable.

The result illustrates the full benefit of value-based optimisation: potentially accepting a higher cost for a conversion when it contributes more to the business.

It also demonstrates that performance cannot be assessed solely within the advertising platform. Google Ads measures leads and their cost; the advertiser is best placed to analyse their conversion and profitability.

How Can Value-Based Optimisation Be Implemented?

For businesses that generate leads, the approach can be structured into five steps.

1. Map the Available Information

The first step is to examine the data available in forms, the CRM and sales tools: type of need, requested product, prospect profile, location, qualification level or outcome of the commercial relationship.

The objective is to identify available, reliable and actionable criteria that can be used to categorise leads according to the value they are likely to generate for the business. The aim is therefore not to collect more data, but to identify the data points that genuinely distinguish the prospects with the greatest potential.

2. Link Segments to Business Outcomes

The next step is to compare conversion rates, revenue, margins or customer lifetime value across the different profiles.

The question is not which lead appears to be the most valuable, but which characteristics can be shown, based on data, to be associated with a stronger economic contribution.

3. Verify That the Platform Can Interpret the Signal

Not all data that is useful to the business can necessarily be used by an advertising algorithm.

An initial use case considered by the teams did not produce conclusive results because the platform lacked a sufficiently robust signal to identify the target profile. This test enabled the teams to refine their methodology and select a more relevant segmentation.

The right use case lies at the intersection of three elements: a demonstrable difference in value, data available to the advertiser and a signal that the platform can activate.

4. Organise Tagging and Data Flows

The selected values must be associated with conversions and then passed correctly to Google Ads.

Google notably recommends checking tagging quality, connecting first-party CRM data and, where relevant, sharing offline conversions.

This step requires close collaboration between marketing, data and technical teams. Without reliable data, even the best algorithm will optimise against an incomplete or inaccurate signal.

5. Test, Learn and Measure Actual Profitability

Value-based optimisation takes time. In Allianz's case, the project was developed over 18 months, including identifying the right use case, technical development, learning periods, testing and results analysis.

It is important to compare consistent periods, account for seasonality and give the algorithm sufficient time to learn. Google recommends feeding the system new values for at least four weeks or three conversion cycles before assessing the transition.

Above all, the final decision must not be based on CPL alone. It must take into account prospect quality, conversion and the value generated for the business.

Redefining Performance

Value Based Bidding is not simply a new way of configuring campaigns. It requires the business to define what a genuinely valuable conversion is and to share that definition across its teams.

The question is no longer simply: how many leads did we generate, and at what cost?

It becomes: which leads did we generate, and how much value did they actually create?

What's Next?

Do you generate leads in insurance, banking, real estate, automotive or B2B services? Our teams can help you identify the segments that contribute most to your business, organise the necessary data and test a media optimisation strategy based on their actual value.

Contact the WPP Media/Keyade teams to explore the opportunities applicable to your acquisition campaigns.