Google Ads automation is powerful, but not foolproof. Learn the signs it’s on track – and the signals it’s steering performance wrong. Smart Bidding, Performance Max, and responsive search ads (RSAs) can all deliver efficiency, but only if they’re optimizing for the right signals. The issue isn’t that automation makes mistakes. It’s that those mistakes compound over time. Left unchecked, that drift can quietly inflate your CPAs, waste spend, or flood your pipeline with junk leads. Automation isn’t the enemy, though. The real challenge is knowing when it’s helping and when it’s hurting your campaigns. Here’s how to tell. When automation is actually failing These are cases where automation isn’t just constrained by your inputs. It’s actively pushing performance in the wrong direction. Performance Max cannibalization The issue PMax often prioritizes cheap, easy traffic – especially branded queries or high-intent searches you intended to capture with Search campaigns. Even with brand exclusions, Google still serves impressions against brand queries, inflating reported performance and giving the illusion of efficiency. On top of that, when PMax and Search campaigns overlap, Google’s auction rules give PMax priority, meaning carefully built Search campaigns can lose impressions they should own. A clear sign this is happening: if you see Search Lost IS (rank) rising in your Search campaigns while PMax spend increases, it’s likely PMax is siphoning traffic. Recommendation Use brand exclusions and negatives in PMax to block queries you want Search to own. Segment brand and non-brand campaigns so you can track each cleanly. And to monitor branded traffic specifically, tools like the PMax Brand Traffic Analyzer (by Smarter Ecommerce) can help. Dig deeper: Performance Max vs. Search campaigns: New data reveals substantial search term overlap Auto-applied recommendations (AAR) rewriting structure The issue AARs can quietly restructure your campaigns without you even noticing. This includes: Adding broad match keywords. “Upgrading” existing keywords to broader match types. Adding new keywords that are sometimes irrelevant to your targeting. Google has framed these “optimizations” as efficiency improvements, but the issue is that they can destabilize performance. Broad keywords open the door to irrelevant queries, which then can spike CPA and waste budget. Recommendation First, opt out of AARs and manually review all recommendations moving forward. Second, audit the changes that have already been made by going to Campaigns > Recommendations > Auto Apply > History. From there, you can see what change happened on what date, which allows you to go back to your campaign data and see if there are any performance correlations. Dig deeper: Top Google Ads recommendations you should always ignore, use, or evaluate Modeled conversions inflating numbers The issue Modeled conversions can climb while real sales or MQLs stay flat. For example, you may see a surge in reported leads or purchases in your ads account, but when you look at your CRM, the numbers don’t match up. This happens because Google uses modeling to estimate conversions where direct measurement isn’t possible. If Google doesn’t have full tracking, it fills gaps by estimating conversions it can’t directly track, based on patterns in observable data. When left unchecked, the automation will double down on these patterns (because it assumes they’re correct), wasting budget on traffic that looks good but won’t convert. Recommendation Tell the automation what matters most to your business. Import offline or qualified conversions (via Enhanced Conversions, manual uploads, or CRM integration). This will ensure that Google optimizes for real revenue and not modeled noise. When automation is boxed in: Reading the signals Not every warning in Google means automation is failing. Sometimes the system is limited by the goals, budget, or inputs you’ve set – and it’s simply flagging that. These diagnostic signals help you understand when to adjust your setup instead of blaming the algorithm. Limited statuses (red vs. yellow) The issue A Limited status doesn’t always mean your campaign is broken. If you see a red Limited label, this means your settings are too strict. That could mean that your CPA or ROAS targets are unrealistic, your budget is too low, etc. Seeing a yellow Limited label is more of a caution sign. It’s usually tied to low volume, limited data, or the campaign is still learning. Recommendation If the status is red, loosen constraints gradually: raise your budget and ease up CPA/ROAS targets by 10–15%. If the status is yellow, don’t panic. This is Google’s version of telling you that they could use more money, if possible, but it’s not vital to your campaign’s success. Responsive search ads (RSAs) inputs The issue RSAs are built in real-time from the headlines and descriptions you have already provided Google. At a minimum, advertisers are required to write 3 headlines with a maximum of 15 (and up to 4 descriptions). The fewer the assets you give the system, the less flexibility it will have. On the other hand, if you’re running a small budget and give the RSAs all 15 headlines and 4 descriptions, there is no way Google will be able to collect enough data to figure out which combinations actually work. The automation isn’t failing with either. You’ve either given it too little information or too much with too little spending. Recommendation Match asset volume to the budget allocated to the campaign. If you’re unsure, aim to write between 8-10 headlines and 2-4 descriptions. If each headline/description isn’t distinct, don’t use it. Conversion reporting lag and attribution issues The issue Sometimes, Google Ads reports fewer conversions than your business actually sees. This isn’t necessarily an automation failure. It’s often just a matter of when the conversion is counted. By default, Google reports conversions on the day of the click, not the day the actual conversion happened. That means if you check performance mid-week, you might see fewer conversions than your campaign has actually generated because Google attributes them back to the click date. The data usually “catches up” as lagging conversions are processed. Recommendation Use the Conversions (by conversion time) column alongside the standard conversion column. Conversions (by conversion time) column This helps you separate true performance drops from simple reporting delays. If discrepancies persist beyond a few days, investigate the tracking setup or import accuracy. Just don’t assume automation is broken just because of timing gaps. Get the newsletter search marketers rely on. See terms. Where to look in the Google Ads UI Automation leaves a clear trail within Google Ads if you know where to look. Here are some reports and columns to help spot when automation is drifting. Bid Strategy report: Top signals The issue The bid strategy report shows some of the signals Smart Bidding relies on when there is enough data. The “top signals” can sometimes make sense, and at other times, they can be a bit misleading. If the algorithm relies on weak signals (e.g., broad search themes and a lack of first-party data), its optimizations will be weak, too. Bid Strategy report: Top signals Recommendation Make checking your Top Signals a regular activity. If they don’t align with your business, fix the inputs. Improve conversion tracking. Import offline conversions. Reevaluate search themes. Add customer/remarketing lists. Expand your negative keyword list(s). Impression share metrics The issue When a campaign underdelivers, it’s tempting to assume automation is failing, but looking at Impression Share (IS) metrics tends to reveal the real bottleneck. By looking at Search Lost IS (budget), Search Lost IS (rank), and Absolute Top IS together, you can separate automation problems from structural or competitive ones. How to use IS metrics as a diagnostic tool. Budget problem High Lost IS (budget) + low Lost IS (rank): Your campaign isn’t struggling. It just doesn’t have enough budget to run properly. Recommendation: Raise the budget or accept capped volume. Targets too aggressive High Lost IS (rank) + low Absolute Top IS: If your Lost IS (rank) is high and your budget is adequate, your CPA/ROAS targets are likely too aggressive, causing Smart Bidding to underbid in auctions. Recommendation: Loosen targets gradually (10-15%). Scripts to keep automation honest Scripts give you early warnings so you can step in before wasted spend piles up. Anomaly detection The issue: Automation can suddenly overspend or underspend when conditions in the marketplace change, but you often won’t notice until reporting lags. Recommendation: Use an anomaly detection script to flag unusual swings in spend, clicks, or conversions so you can investigate quickly. Query quality (N-gram analysis) The issue: Broad match and PMax can drift into irrelevant themes (“free,” “jobs,” “definition”), wasting budget on low-quality queries. Recommendation: Run an N-gram script to surface recurring poor-quality terms and add them as negatives before automation optimizes toward them. Budget pacing The issue: Google won’t exceed your monthly cap, but daily spend will be uneven. Pacing scripts help you spot front-loading. Recommendation: A pacing script shows you how spend is distributed so you can adjust daily budgets mid-month or hold back funds when performance is weak. Turning automation into an asset Automation rarely fails in dramatic ways – it drifts. Your job isn’t to fight it, but to supervise it: Supply the right signals. Track when it goes off course. Step in before wasted spend compounds. The diagnostics we covered – impression share, attribution checks, PMax insights, and scripts – help you separate real failures from cases where automation is simply following your inputs. The key takeaway: automation is powerful, but not self-policing. With the right guardrails and oversight, it becomes an asset instead of a liability.

 

As retail media networks mature, success hinges on mastering both global scale and neighborhood-level precision.

Retail media networks are projected to be worth $179.5 billion by 2025, but capturing share and achieving long-term success won’t hinge solely on growing their customer base. With over 200 retail media networks now competing for advertiser attention, the landscape has become increasingly complex and crowded. The RMNs that stand out will be those taking a differentiated approach to meeting the evolving needs of advertisers.

The industry’s concentration creates interesting dynamics. While some platforms have achieved significant scale, nearly 70% of RMN buyers cite “complexity in the buying process” as their biggest obstacle. That tension, between explosive growth and operational complexity, is forcing the industry to evolve beyond traditional approaches.

As the landscape matures, which strategies will define the next wave of growth: global expansion, hyperlocal targeting, or both?

The evolution of retail media platforms

To understand where the industry is heading, it’s worth examining how successful platforms are addressing advertisers’ core challenges. Lack of measurement standards across platforms continues to frustrate advertisers who want to compare performance across networks. Manual processes dominate smaller networks, making campaign management inefficient and time-consuming.

At the same time, most retailers lack the digital footprint necessary for standalone success. This has created opportunities for platforms that can solve multiple problems simultaneously: standardization, automation, and scale.

DoorDash represents an interesting case study in this evolution. The platform has built its advertising capabilities around reaching consumers at their moment of local need across multiple categories. With more than 42 million monthly active consumers as of December 2024, DoorDash provides scale and access to high-intent shoppers across various categories spanning restaurants, groceries and retail.

The company’s approach demonstrates how platforms can address advertiser pain points through technology. DoorDash’s recent platform announcement showcases this evolution: the company now serves advertisers with new AI-powered tools and expanded capabilities. Through its acquisition of ad tech platform Symbiosys, a next-generation retail media platform, brands can expand their reach into digital channels, such as search, social, and display, and retailers can extend the breadth of their retail media networks.

Global expansion meets local precision

International expansion presents both opportunities and challenges for retail media networks. Europe’s retail media industry is projected to surpass €31 billion by 2028,. This creates opportunities for networks that can solve the technology puzzle of operating across multiple geographies.

The challenge lies in building platforms that work seamlessly across countries while maintaining local relevance. International expansion requires handling different currencies, regulations, and cultural contexts—capabilities that many networks struggle to develop.

DoorDash’s acquisition of Wolt illustrates how platforms can achieve global scale while maintaining local connections. The integration enables brands to manage campaigns across Europe and the U.S. through a single interface—exactly the kind of operational efficiency that overwhelmed advertisers seek.

The combined entity now operates across more than 30 countries, with DoorDash and Wolt Ads crossing an annualized advertising revenue run rate of more than $1 billion in 2024. What makes this expansion compelling isn’t just the scale—it’s how the integration maintains neighborhood-level precision across diverse geographies.

Wolt has transformed from a food delivery platform into what it describes as a multi-category “shopping mall in people’s pockets.”

The hyperlocal advantage: context beats demographics

Here’s what’s really changing the game: the shift from demographic targeting to contextual precision. Privacy regulations favor contextual targeting over behavioral tracking, but that’s not the only reason smart networks are going hyperlocal.

Location-based intent signals provide dramatically higher conversion probability than traditional demographics. Real-time contextual data—weather patterns, local events, proximity to fulfillment—influences purchase decisions in immediate, actionable ways that broad demographic targeting simply can’t match.

DoorDash built its entire advertising model around this insight, reaching consumers at the exact moment of local need across multiple categories. The platform provides scale and access to high-intent shoppers with contextual precision. A recent innovation that exemplifies this approach is Dayparting for CPG brands, which enables advertisers to target users in their local time zones—a level of time-based precision that distinguishes hyperlocal platforms from broader retail media networks.

In one example, Unilever applied Dayparting to focus on late-night and weekend windows for its ice cream campaigns, aligning ad delivery with peak demand periods. Over a two-week period, 77% of attributed sales were new-to-brand, demonstrating the power of contextual timing in driving incremental reach.

Major brands, including Unilever, Coca-Cola, and Heineken, utilize both DoorDash and Wolt platforms for hyperlocal targeting, proving the model is effective for both endemic and non-endemic advertisers seeking neighborhood-level precision.

Technology evolution: measurement and automation

The technical requirements for next-generation retail media networks extend far beyond basic advertising capabilities. Self-serve functionality has become standard for international geographies—not because it’s trendy, but because manual campaign management doesn’t scale across dozens of countries with different currencies, regulations, and cultural contexts.

Cross-country campaign management requires unified dashboards that manage complexity while maintaining simplicity for advertisers. Automation isn’t optional anymore; it’s necessary to compete with established players who’ve built machine learning into their core operations.

But here’s what’s really transforming measurement: new attribution methodologies that go beyond traditional ROAS. When platforms can integrate fulfillment data with advertising exposure, they enable real-time performance tracking that connects ad spend to actual business outcomes rather than just clicks and impressions.

Progress on standardization continues through IAB guidelines addressing measurement consistency, alongside industry pushes for technical integration standards. The challenge lies in balancing standardization with differentiation—networks need to offer easy integration and consistent measurement while maintaining unique value propositions.

In a move toward addressing advertisers’ need for measurement consistency, DoorDash recognized that restaurant brands valued both click and impression-based attribution for their sponsored listing ads, and recently introduced impression-based attribution and reporting in Ads Manager. This has enabled restaurant brands to gain a deeper understanding of performance and results driven on DoorDash.

Global technology challenges add another layer of complexity: multi-currency transactions, local payment methods, regulatory compliance across countries, and cultural adaptation while maintaining platform consistency. These aren’t afterthoughts for international platforms, they’re core competencies that determine success or failure.

Industry outlook: consolidation and opportunity

Retail media is heading toward consolidation, but not in the way most people expect. Hyperlocal networks are positioned to capture share from undifferentiated RMNs that compete solely on inventory volume. Geographic specialization is becoming a viable alternative to traditional scale-focused approaches.

Simultaneously, community impact measurement is gaining importance for brand strategy. Marketers are discovering that advertising dollars spent on local commerce platforms create multiplier effects—supporting neighborhood businesses and strengthening local economies in ways that traditional e-commerce advertising doesn’t achieve.

The networks that understand this dynamic, that can offer global platform capabilities with genuine local industry expertise, are the ones positioned to define retail media’s next chapter. Success requires technology integration that enables contextual and location-based targeting, plus measurement solutions that prove incrementality beyond traditional metrics.

The path forward

As retail media networks mature, success lies not in choosing between global scale and local relevance, but in achieving both simultaneously. The DoorDash-Wolt combination provides a compelling blueprint, demonstrating how technology platforms can enable international expansion while deepening neighborhood-level connections.

For marketers navigating this evolution, the fundamental question shifts from “where should we advertise?” to “how can we reach consumers at their moment of need?” Networks that answer this effectively—through global reach, hyperlocal precision, or ideally both, will write retail media’s next chapter.Interested to learn more about DoorDash Ads?