TrackAd and NOA: Prescriptive AI Revolutionizes Digital Acquisition

Data Gaps on Platforms: The Puzzle of Digital Acquisition

In today's digital marketing landscape, each platform generates its own performance indicators, creating a major challenge for advertisers. Strategic decisions are often made in isolation, channel by channel, relying on metrics that do not uniformly measure conversions. This siloed approach leads to biased budget allocations, where funds are often directed to channels that appear to be the most effective, typically those at the bottom of the conversion funnel. This sometimes comes at the expense of other channels that may be equally important. According to Guillaume Le Roy, CEO of TrackAd, this situation results in a loss of 21% of advertising budgets invested, as they do not generate the expected return on investment.

Campaigns aimed at increasing awareness or enhancing consideration suffer particularly from this approach. These types of campaigns do not produce immediate conversions, complicating their evaluation in a performance-driven model. Guillaume Le Roy emphasizes that the main risk is a misunderstanding of the actual contribution of each advertising lever.

The opacity of the tools provided by platforms like Meta and Google exacerbates this problem. Their automated solutions, such as Advantage+ and Performance Max, promise improved performance but at the cost of lost control and clarity for advertisers. This can lead to insufficient segmentation of campaigns, even when managed by algorithms. The danger lies in concentrating the entire budget on a single PMAX campaign, disregarding price variations, seasonality, or target audience, which limits the ability to test and adjust budgets based on the specific performance of each campaign.

Moreover, partial performance tracking, the disappearance of third-party cookies, and closed environments further complicate the situation. Cutting a channel that appears to be underperforming may actually mean losing a discreet but essential contributor to conversion.

The Importance of Prescriptive and Independent AI

To address these challenges, the use of prescriptive and independent AI proves crucial. This technology allows for transcending the logic specific to each platform, providing a unified view of the media mix and budget allocations aligned with actual profitability.

Guillaume Le Roy explains that while a human is limited to analyzing a few dimensions at a time, AI can cross-reference multiple factors: channel, campaign, device, audience, geography, seasonality, history, and commercial performance. The solution developed by TrackAd, named NOA (Next-gen Optimized Ads), takes this logic even further. NOA evaluates acquisition campaigns based on their role in the customer journey and their specific objectives. Thus, the expected return on investment from a campaign on Google DV360 differs from that of a remarketing campaign on Meta.

The goal of AI is not to optimize each campaign in isolation but to streamline the entire budget mix. This technology offers three key capabilities:

• Identifying marginal returns, i.e., the threshold at which a channel becomes less profitable.
• Detecting pockets of waste, such as active campaigns with no added value, over-invested geographic areas, or saturated audiences.
• Simulating arbitration scenarios, allowing for anticipation of the impact of budget rebalancing before implementation.

Thus, we move from descriptive management to predictive management, as explained by Guillaume Le Roy.

Managing Acquisition with AI: TrackAd's Best Practices

To make the most of AI in managing advertising campaigns, TrackAd offers several best practices:

• Start with business objectives rather than platform metrics. For example, Google’s ROAS (Return on Ad Spend) is not sufficient to measure overall profitability.
• Understand your marketing mix before analyzing it. Each lever has a distinct role, and analyzing them uniformly leads to erroneous conclusions.
• Centralize data to provide AI with a unified view of sources, a condition essential for its relevance.
• Use AI as a co-pilot rather than as autopilot. AI suggests, but it is up to humans to decide.
• Constantly test, measure, and challenge AI recommendations, which should not be considered absolute truths.
• Be willing to question your beliefs, as AI can reveal uncomfortable truths.

TrackAd Bets on Prescriptive AI to Manage Acquisition Across Multiple Channels

TrackAd has developed NOA (Next-gen Optimized Ads) based on these principles. This solution combines an adaptive algorithm and generative AI to provide actionable and prioritized recommendations. According to Guillaume Le Roy, NOA analyzes performance and directly identifies anomalies, opportunities, or inconsistencies in campaigns, thereby accelerating decision-making.

In practical terms, NOA can recommend reallocating part of the budget from an underperforming source to a channel offering a higher marginal return, or signal that a consideration campaign is generating too small a share of new visitors. The analysis is contextualized: a consideration campaign is not evaluated using the same indicators as a conversion campaign.

NOA is natively integrated into the TrackAd platform, allowing it to leverage all available data, whether from media, analytics, or commercial data, without additional configuration. Its learning is continuous, and recommendations are refined over time based on each client's specifics. NOA also checks the technical consistency of sources, detects tracking errors that could skew analysis, and ensures that the KPIs associated with each campaign correspond to its type.