At the 2025 Cannes International Creativity Festival, Vidhya Srinivasan, Vice President of Google Ads, revealed an important trend: 43% of companies around the world have entered the AI implementation stage, and the revenue growth of leading companies is 60% higher than that of AI startups. The data comes from a new study by Boston Consulting Group (BCG), highlighting the strategic value of AI in digital marketing. Especially in the field of Google Ads, traditional account structures and management methods are facing unprecedented challenges and innovation opportunities. The success story of Spanish hotel chain RIU Hoteles further proves that through AI-driven Demand Gen advertising strategy, companies can achieve amazing results of doubling revenue and increasing ROI by 3 times. This AI-driven marketing revolution is forcing marketers to rethink the operating logic and strategic framework of Google Ads.
In today’s marketing environment, AI has moved from a competitive advantage to a business necessity. A joint study by Google and BCG shows that the revenue growth of leading AI application companies is significantly higher than that of their peers, which is particularly critical in a market environment with tight budgets. The value of AI lies not only in improved efficiency, but also in its ability to handle data scale and complexity that are beyond the reach of humans. Take Google Ads’ smart bidding system as an example. It can analyze hundreds of real-time signals at the same time, including user devices, time periods, locations and even current market trends, to make bidding decisions that are more accurate than manual ones. This capability enables marketers to break through the linear thinking of traditional conversion funnels and intervene precisely at the key moments of influence when consumers are watching streaming media, searching for information or making shopping decisions. AI is no longer just a tool, but has become a strategic asset that reshapes the entire marketing value chain.
The segmentation logic of the traditional Google Ads account structure has become the main obstacle to improving performance in the AI era. The Sephora case is instructive—the brand discovered that its overly segmented account structure based on product lines severely limited the AI system’s learning capabilities and optimization space. When campaigns are cut up too piecemeal, it’s hard for AI models to get enough data to learn effectively, which is like trying to piece together a complete image using scattered fragments. What’s more serious is that the multi-level account structure will lead to data fragmentation, making it impossible for the marketing team to gain a global perspective and make accurate decisions. This structural problem cannot be solved through fine-tuning; a fundamental account restructuring is required to unlock the full potential of AI-driven marketing.
Leading companies around the world have achieved significant results through account simplification strategies. After streamlining 85% of their campaigns, Sephora not only saw a 42% increase in conversion rate, but also a 6% increase in average order value. RIU Hoteles doubled its revenue in 14 countries through its AI-driven Demand Gen strategy. Together, these success stories reveal a key insight: in the age of AI, less is more. The simplified account structure enables the AI system to more comprehensively understand marketing goals and audience behavior, thereby making more accurate optimization decisions. It is worth noting that these companies are not blindly simplifying, but are scientifically reorganizing based on data analysis and subject relevance, integrating the originally scattered data points into coherent learning materials, allowing AI to exert its true predictive and optimization capabilities.
An overly segmented account structure essentially constrains AI’s learning capabilities. When keywords and advertising groups are cut into overly fragmented units, the amount of data in each unit is often insufficient to support effective pattern recognition and prediction by the AI system. It's like trying to teach a student to solve a complex problem by only providing them with fragmented pieces of information. Google Ads AI system requires sufficient data breadth and depth to understand the correlation between user intent and behavior patterns. In practice, the performance of the smart bidding system drops significantly when the ad group contains fewer than 50 conversions. Many businesses fail to realize that their carefully crafted “super targeting” strategies are actually undermining AI’s decision-making capabilities. The solution is to build more flexible topic groups so that AI can discover deep patterns and opportunities that are difficult for humans to detect based on sufficient data.
Complex account structures not only affect AI performance, but also cause serious waste of human resources. Marketing teams are often bogged down in an endless cycle of micromanagement—adjusting bids on individual campaigns, monitoring the performance of decentralized keywords, and coordinating budget allocations across multiple levels. This operating model transforms valuable strategic thinking time into mechanical operational work. According to internal research at Google, an overly complex account structure can cause teams to spend up to 60% of their time on basic maintenance rather than strategic innovation. What’s more serious is that multi-level management often leads to inconsistent strategies and execution deviations, and different teams may make completely different interpretations of the same data. Simplifying the account structure is not only a technical optimization, but also a transformation of organizational effectiveness. It can free the team from the operational quagmire and focus on strategic work that truly creates value.
When data is scattered across dozens or even hundreds of campaigns, even the most professional marketers struggle to gain accurate, global insights. Data fragmentation can lead to several serious problems: first, cross-series comparisons become difficult because the benchmarks set for different series may be inconsistent; second, trend analysis lacks continuity, and short-term fluctuations may be misinterpreted as long-term trends; finally, attribution analysis loses accuracy and cannot truly reflect the complete path of the user journey. These data issues will directly affect the quality of decision-making, leading to improper budget allocation and increased opportunity costs. The solution is to establish a unified data architecture that integrates key indicators into a horizontally comparable analytical framework. This not only improves decision-making accuracy, but also provides AI systems with cleaner and more consistent training data, further enhancing their predictive capabilities.
The first step in account simplification is to establish a solid data foundation. This means reviewing all existing conversion tracking to ensure it accurately reflects business value and not just surface level metrics. For example, for an e-commerce business, the focus should be on revenue rather than just clicks; for a B2B business, the focus should be on tracking high-value leads rather than all form submissions. In practice, it is recommended to use Google Analytics 4 as the central data hub and build a complete user journey map through its powerful event tracking function. At the same time, the value attribution model should be unified to avoid incomparable data caused by different attribution windows used in different advertising campaigns. This stage may involve cleaning and re-labeling of historical data, which is time-consuming but crucial because all subsequent optimization decisions will be based on this data. Once completed, you will have a clean, consistent dataset that enables AI systems to learn and make more accurate predictions.
Traditional keyword-centric grouping logic is no longer applicable in the AI era. Instead, it is replaced by topic classification based on user intent and business goals. In practice, this means combining highly related keywords into the same ad group, even if they appear different on the surface. For example, a flower shop should group "red roses", "12 roses", and "wedding roses" into the same theme group rather than managing them separately. This categorization enables the AI system to better understand core business objectives and make smarter delivery decisions when relevant queries arise. The golden rule of topic categorization is that if two keywords attract a similar type of audience and lead to the same business goals, they belong in the same group. Once the refactoring is complete, you'll notice a general improvement in ad relevance rankings because the system is now able to more accurately match query intent to ad content.
Once you have your data foundation and thematic structure in place, you can start to boldly merge campaigns. Merger criteria should be based on business objectives rather than superficial characteristics—all activities pursuing the same transformation goal should be considered for merger. For example, campaigns for different variations of the same product line, or separate campaigns targeting similar audiences. During the merger process, the key is to maintain consistency in bidding strategies and budget allocations, and avoid forcibly combining activities that use different logics. The practical technique is to conduct a pilot first: select 2-3 series with high correlation but different performance, merge them, monitor the performance changes for 2-4 weeks, and then promote them comprehensively. It was through this gradual merger that Sephora ultimately streamlined 85% of its advertising campaigns. It is worth noting that after the merger, the AI system should be given a sufficient learning period (usually 14-28 days), during which frequent intervention should be avoided to allow the algorithm to fully adapt to the new data structure.
The simplified accounts require a completely new monitoring framework. The traditional evaluation method centered on click-through rate (CTR) is no longer sufficient, and a multi-dimensional performance indicator system should be established. The first layer is business outcome indicators, such as return on investment (ROAS), customer acquisition cost (CAC) etc.; the second layer is AI performance indicators, including learning status, prediction accuracy, etc.; the third layer is the traditional engagement indicators. Monitoring frequency should also be adjusted - business outcome indicators can be reviewed daily but analyzed for trends weekly, while AI indicators need to be observed over a longer period (at least 14 days) to avoid overreaction. In practice, it is recommended to establish automated dashboards, compare key indicators with historical benchmarks and industry standards, and set intelligent alerts rather than fixed thresholds. This monitoring method can not only detect problems in a timely manner, but also will not interfere with the long-term learning process of AI due to short-term fluctuations.
Sephora’s transformation process is of great reference value. Faced with the growth bottleneck in the UK market, its marketing team first conducted a comprehensive account audit and found that the over-segmented product line advertising structure led to three major problems: insufficient learning data, high management complexity, and rigid budget allocation. The solution was to reorganize the structure originally divided by product categories (such as lipstick, eye shadow, etc.) into a thematic structure based on consumption intentions (such as holiday gift giving, daily replenishment, etc.). In terms of technical execution, the team took a phased approach: first merging low-performing campaigns to test the effects, and then gradually expanding to core product lines. The key success factor was unifying the value tracking system, ensuring that all combined campaigns were optimized based on the same conversion metrics. The results exceeded expectations—not only did switching costs drop by 13%, but some overlooked product combination opportunities were unexpectedly discovered, which were the new patterns discovered by AI on a larger data set.
The case of RIU Hoteles demonstrates the power of AI-driven strategies in complex multinational environments. Faced with the diverse demands of 14 different markets, traditional country-specific advertising strategies led to dispersed resources and inefficient learning. The solution is to adopt Google’s Demand Gen strategy, combining lookalike targeting with tROAS-based smart bidding. The key innovation is to centralize high-value customer data from various markets and train a unified AI model to identify the characteristics of potential customers across borders. At the same time, the team significantly simplified its creative strategy, moving from dozens of independent versions to a modular creative combination based on core value propositions, with AI instantly combining the best elements based on audience characteristics. The challenge in the implementation process is to coordinate teams in various countries to accept this centralized decision-making model. The solution is to establish a transparent performance sharing mechanism so that local markets can clearly see the overall benefits. Ultimately, the strategy not only improved efficiency but also uncovered unexpected customer flow patterns across markets, providing new insights into pricing strategies.
Topkee provides integrated digital marketing solutions based on the Google Ads platform. Its service framework covers the entire cycle from early evaluation to later optimization. Through professional website evaluation tools, the team will conduct SEO structural diagnosis on the client's website, including technical level review and content value analysis, and produce a detailed optimization recommendation report. This basic assessment can ensure that subsequent advertising and landing pages are highly coordinated. According to actual data, an optimized website structure can increase advertising conversion rates by more than 40%. The service system has specially designed a TTO management tool, which supports centralized management of multiple accounts, can simultaneously handle complex operations such as media budget allocation and advertising account authorization, and realize cross-platform data tracking through tag ID association.
At the advertising execution level, Topkee has developed a TM tracking module that has more detailed dimensional setting capabilities than traditional UTM parameters. The system allows customization of tracking rules based on 15 variables such as advertising source, media type, creative goals, etc. The generated TMID link can accurately attribute the effectiveness of each channel. The keyword strategy adopts a three-stage research method: first, the core industry vocabulary is built, then the long-tail keywords are expanded through competitor phrase analysis, and finally the matching mode is dynamically adjusted in combination with the intelligent bidding algorithm. The creative production process integrates AI-assisted tools to automatically generate text and image materials that comply with Google's advertising policies, and then a professional design team performs scene-based polishing to ensure that the visual elements are highly consistent with the marketing message.
This AI-driven Google Ads transformation is essentially an evolution of marketing thinking from mechanical execution to strategic intelligence. When Sephora streamlined its advertising campaigns by 85% and achieved a 42% increase in conversion rates, and when RIU Hoteles doubled its revenue, what they demonstrated was not the magic of technology but the power of strategic transformation. AI won’t replace marketers, but marketers who use AI will replace those who resist change. Is your Google Ads account structure ready for the inevitable change? Now is the perfect time to reassess and adjust. If you need professional guidance, our team of consultants is ready to help you design advertising strategies in the AI era to get unprecedented returns on your marketing investment.