Product feed management is often the unsung hero behind successful ecommerce campaigns.

A well-optimized product feed ensures your ads appear in the right place, at the right time, and in front of the right audience. But when managing thousands—or even millions—of SKUs across multiple platforms, maintaining a high-quality feed can become increasingly complex and time-consuming. 

But what should you do when you suddenly find yourself with a ridiculously large product feed that’s disorganized, missing key data, and generally an all-round mess?

In a recent episode of the Paid Media Lab podcast, we spoke with Nick Handley, Head of Paid Media Performance at Impression Digital, about a large-scale marketing strategy developed for an automotive client. 

Straight away, Nick realised the client’s product feed was ‘unfathomably large’, and incredibly disorganized. The product feed was missing key data, such as titles and product images, severely hindering the performance of the client’s dynamic ads. 

Nick took us through his step-by-step process to analyze and optimize the feed, essentially turning rags to riches (over 30k leads per month, and a 55% decrease in cost per lead)

We also spoke to Nick about a range of other key marketing topics–from how to measure the true impact of your marketing efforts, to effectively integrating cross-platform data to optimize campaigns and improve results. 

Watch the full episode here–or keep reading for the written takeaways from the session. 

Key topics & timestamps:

00:00 - Intro

02:33 - Nick’s approach to a challenging automotive client

05:08 - Dealing with an unorganized product feed

08:12 - Vehicle ads overview

10:10 - Creating personalised feeds for Meta

15:24 - Getting quality first-party information

17:12 - Measuring the holistic impact of marketing

20:07 - Using incrementality & uplift tests to measure impact 

22:24 - Key advice & the importance of feed quality

25:16 - Final thoughts

For all of Nick’s tips, tricks, and insights on product feed management (and more), watch the full episode of the Paid Media Lab podcast above – and make sure to subscribe to the Lunio YouTube channel to watch full episodes of the Paid Media Lab podcast, featuring insights from top PPC professionals. 

Alternatively, follow the podcast on Spotify, Apple Podcasts, or Amazon Music to get new episodes delivered straight to your feed.

The challenge of scale in the automotive industry

Nick’s client, a major automotive brand, faced a common challenge in the industry: managing a vast inventory of products, with over 340,000 cars on their site. The sheer scale of unorganized products presented significant difficulties for paid media campaigns. 

The client's initial goal was to generate 20,000 leads per month through paid media, while also reducing their cost per lead. To achieve this, they needed to address their worse-for-wear product feed–and fast.

The initial feed was in disarray: Missing titles and basic information, which hampered dynamic ad performance on paid social and inventory management for Google Ads.

Automation and personalization

To rectify this, Nick’s core strategy involved automating feed rules and implementing personalization. Instead of showing generic ads, the focus was on displaying cars that matched a user’s intent and past interactions. 

This personalization extended to ad creatives on social media, while Google Ads utilized vehicle ads to enhance visual appeal.

Fixing the feed

During our discussion, Nick emphasized that the importance of a well-managed product feed cannot be overstated, as it significantly drives performance for paid media. His process of fixing a feed with a large product catalog starts with the following steps:

1. Internal understanding: Getting the client to recognize the importance of the feed.

2. Initial analysis: Using internal tools to assess the current state of the feed. This includes:

◦ Checking for missing titles.

◦ Identifying titles that are cut off or poorly formatted.

◦ Verifying the quality of product images.

◦ Looking for missing unique identifiers (GINs).

◦ Identifying pricing errors or other issues leading to product disapprovals.

3. Data organization: Prioritizing getting the data in order to create a complete feed.

4. Supplemental feeds: Using third-party tools to create supplemental feeds where internal data is insufficient.

◦ Tools such as Feedonomics can integrate data from platforms like Shopify, Magento, and others to Google Ads and Meta.

◦ These tools allow for creating rules to fill in missing data, such as titles, by pulling information from other fields or supplemental sheets.

5. Third-party integrators: Utilizing platforms like Feedonomics to automate feed updates.

Using Google’s vehicle ads

Nick's team was among the first in the UK to use vehicle ads on Google Ads. Vehicle ads are similar to shopping ads, but are specifically designed for showing cars dynamically. These ads appear visually appealing at the top of the Google search results page.

• Visual appeal: Vehicle ads provide a more visually appealing alternative to traditional text ads, increasing user engagement and performance.

• Functionality: Vehicle ads work through Google's Performance Max (PMax) campaigns, using a vehicle feed within the Merchant Center.

• Testing: The ad format allows for testing of different assets to create high-performing ads.

Whether you’re in the automotive industry or not–exploring and leveraging new ad formats is crucial when it comes to enhancing campaign visibility and engagement. 

Personalization on Meta: Dynamic feed overlays

On Meta, Nick’s team faced another challenge–dealers uploading placeholder images, such as "image coming soon". Generally speaking, Meta's algorithm is less than effective when it comes to identifying these incorrect placeholders.

To solve this, Nick’s team innovated using Google's API to detect these errors and creating rules to exclude these products from the Meta feed. This reduced waste and saved approximately £10,000 per month.The next phase was creating a more personalized experience with dynamic feed overlays via the following: 

‍1. Data collection: Utilizing data from interactions across different channels (Google, Pinterest, etc.) via the data layer.

‍2. Audience creation: Building audiences based on customer data, ensuring GDPR compliance.

◦ CRM integrators (e.g Clavio) can be used to categorize users.

‍3. Dynamic overlays: Using third party tools to create dynamic feed overlays that display the most relevant information based on customer preferences (price, condition, mileage).

◦ For example, if someone frequently viewed Jags, the ad shown to them will emphasize price.

◦ This personalized approach led to an average of 30,000 leads per month and reduced the cost per lead by 55%.

Measuring the holistic impact of marketing

Beyond product feed optimization and Meta personalization, accurate measurement is critical for understanding true marketing performance. However, most marketers face a major challenge: platform bias—where advertising platforms tend to over-credit their own channels.

The deprecation of third-party cookies further complicates tracking, making it even harder to get a clear picture of cross-channel impact.To measure marketing effectiveness more accurately, always keep the following in mind: 

• Platform bias awareness: Recognize that platforms optimize their own attribution models to appear more effective, often inflating their contribution to conversions.
• Contribution over attribution: Instead of attributing conversions to a single source, adopt a marketing mix modeling (MMM) or econometrics approach to understand the relative impact of each channel in the broader funnel.
• Brand and performance alignment: As content creators and influencer marketing play a growing role, brand and performance marketing are converging. Understanding how upper-funnel brand efforts contribute to lower-funnel performance is crucial.

Moving from attribution to contribution

To navigate these challenges, Nick recommends shifting from last-click or platform-attributed metrics to a contribution-based measurement model. A key tactic is incrementality testing, which isolates the true impact of marketing efforts on sales.

Incrementality Testing: 

By leveraging geo holdout tests, marketers can create control groups in specific regions where changes (such as increased ad spend) are not implemented. Comparing performance between test and control groups provides undeniable proof of impact, helping to justify budget allocations with data-driven confidence.

Nick’s team applied this approach by increasing Meta ad spend in specific test regions and demonstrated a direct lift in sales, proving Meta’s true incremental value beyond what platform-reported metrics suggested.Key takeaways & actionable adviceFor marketers looking to adopt a data-driven, contribution-focused approach, Nick shared three essential steps:

‍1. Prioritize feed optimization

‍A well-structured product feed is foundational for personalization and performance. Ensure:

•  Complete and accurate data points (pricing, availability, product attributes).
•  No missing content that could disrupt dynamic ad delivery.
•  Backfilling gaps where necessary to maintain consistency across platforms.

‍2. Test, learn, and iterate

‍Hypothesis-driven testing helps refine strategies and prove impact before scaling.

•  Use holdouts or supplemental feeds to measure the effect of changes.
•  Track performance rigorously and adapt based on data-driven insights.

‍3. Start small, scale smart

‍Avoid making sweeping changes all at once. Instead:

•  Implement incremental improvements and validate results before expanding.
•  Build a scalable framework for continuous optimization based on learnings.

By adopting a holistic measurement mindset and using incrementality testing to prove marketing impact, brands can optimize ad spend, maximize efficiency, and drive sustainable growth.

Final thoughts

Nick’s strategy for the automotive client demonstrates the power of combining meticulous feed management with personalization and strategic testing. By focusing on data accuracy and customer intent, marketers can drive significant improvements in lead generation and cost efficiency. 

The insights Nick shared in this episode of the Paid Media Lab offer a practical guide for any business looking to optimize its paid media efforts–particularly those dealing with large product catalogs.

Get Nick’s full insights by watching his episode of the Paid Media Lab on YouTube, or listen on Spotify, Apple Music, or Amazon.