Executive Overview: A New Era of Automated Bidding
In a move that marks the most significant architectural overhaul to its automated bidding infrastructure in ten years, Microsoft Advertising has officially begun the rollout of its next-generation bidding models. Announced on November 18, this update represents a fundamental shift in how the platform processes user data, leverages predictive signals, and optimizes for conversions.
For advertisers, this is not merely a background update; it is a fundamental change in the "brain" of the platform. By utilizing a vastly expanded set of signals to predict the probability and value of a conversion, Microsoft aims to move beyond traditional algorithmic constraints. As these models reach more accounts, campaigns will begin to exhibit new spending behaviors, necessitating a proactive review of current strategies. This transition is part of a broader, multi-year roadmap for Microsoft Advertising, with further significant enhancements slated for 2027.
Detailed Chronology: The Rollout and Its Mechanics
The deployment of these upgraded models is occurring automatically across the Microsoft Advertising ecosystem. Unlike opt-in features that allow for a phased transition, this upgrade is being pushed to accounts globally.
The Mechanism of Change
The new models are designed to be more nuanced in their decision-making process. Previously, bidding algorithms often operated within relatively rigid constraints. The updated system is engineered to analyze a wider array of real-time signals—ranging from user intent and historical behavior to contextual variables—to determine not just if a click will convert, but how much that conversion is worth to the advertiser’s bottom line.
Immediate Impacts on Campaign Dynamics
Advertisers may notice shifts in how quickly their budgets are consumed and how closely their campaigns adhere to established CPA (Cost Per Acquisition) or ROAS (Return on Ad Spend) targets.
One primary area of friction occurs when legacy constraints, such as manual Max CPC limits, conflict with the new model’s predictive capabilities. For example, if a campaign has a $50 CPA target and a $5 manual CPC cap, the system may be prevented from bidding $6 on a high-intent user who is statistically likely to convert within the CPA target. In the new environment, such caps can artificially throttle performance, preventing the algorithm from securing high-value conversions that would have otherwise satisfied the campaign’s broader efficiency goals.
Supporting Context & Metrics: Data-Driven Performance
Microsoft has provided internal data that highlights the potential efficacy of these new models. According to global Search and Audience Network data from August 2026, advertisers utilizing "Maximize Conversion Value" or "Target ROAS" strategies experienced a 10% increase in conversion value per click while maintaining similar ROAS levels.
Strategy-Specific Gains
The data also underscores the benefits of shifting to more advanced bidding strategies:
- From Target CPA to Target ROAS: Search advertisers who migrated their strategies saw an 8% increase in total conversion value at a comparable ROAS.
- From Enhanced CPC to Maximize Conversions: Advertisers who transitioned from manual or semi-automated bidding to "Maximize Conversions" with a set CPA target reported an 18% lift in total conversions, again while maintaining efficiency parity.
These figures suggest that while the transition requires a change in management mindset, the platform is effectively rewarding advertisers who lean into automated, value-based bidding.
The "Cold Start" Problem: Conversion Thresholds
A recurring concern for smaller advertisers is the "data starvation" issue—the minimum amount of conversion history required to trigger the efficacy of these models.

Microsoft’s official stance is that once an account reaches a threshold of at least five total conversions, the system is sufficiently "warmed up" to begin utilizing conversion-based bidding. However, this is a baseline, not an ideal. For those struggling to hit this initial milestone, the company maintains that Enhanced CPC remains a viable interim strategy.
Navah Hopkins on Low-Volume Accounts
Microsoft Advertising Liaison Navah Hopkins has emphasized that while five conversions is the technical minimum, account nuances matter. Addressing the community on LinkedIn, Hopkins noted:
"Every account is different so a blanket recommendation still needs to be put to the test… the system is much more responsive now!"
She advises a 30-day evaluation period for any significant changes. For performance assessment, Microsoft recommends aiming for 30 conversions in a 30-day window at the campaign or portfolio level to ensure statistical significance. If a campaign fails to hit this, the advertiser may be forced to wait longer to determine if the model is functioning optimally.
Navigating the Transition: Best Practices for Advertisers
To avoid "performance whiplash," Microsoft suggests a disciplined approach to modifying targets.
The 15% Rule
When adjusting CPA or ROAS targets, major fluctuations can trigger "learning mode," potentially leading to volatile performance. Microsoft recommends:
- Incremental Adjustments: Change targets in 15% increments.
- Cycle-Based Reviews: After each adjustment, allow at least one full conversion cycle to pass before evaluating the impact.
- Complex Changes: If you are switching bid strategies or adding entirely new conversion goals, allow for one to two full conversion cycles before making further optimizations.
The Importance of Value Accuracy
Perhaps the most critical takeaway for high-growth advertisers is the need for accurate data input. The new models are heavily reliant on the "value" assigned to conversions. If a business tracks all sales as having the same value—or worse, uses an arbitrary, flat-value estimate—the model cannot distinguish between a low-margin sale and a high-margin order.
Advertisers are encouraged to use the Conversions API or offline conversion imports to ensure that the value reported in the system aligns with the actual revenue generated in their CRM. This is particularly vital for lead generation, where not every lead is created equal. As one expert query to Microsoft highlighted, if a company sends a preliminary estimate of a lead’s value and updates it later, the bidding model becomes significantly more precise.
Future Outlook: Looking Ahead to 2027
The current rollout is merely the first phase of an aggressive development cycle. With more bidding updates planned for 2027, the platform is clearly positioning itself to compete more aggressively in the AI-driven advertising space.
Strategic Recommendations for the Coming Year
- Audit Your Tracking: Before the next phase of updates, ensure your conversion tracking is as granular as possible. If you aren’t passing actual revenue data, you are handicapping the AI.
- Resist "Micro-Management": If your campaigns are currently meeting their goals, consider leaving them in a steady state for a few weeks to allow the new models to calibrate without interference. The goal is to let the algorithm find its "new normal."
- Focus on Business Outcomes, Not Just Clicks: As the bidding models prioritize "value," align your advertising goals with your business’s actual profitability metrics.
Conclusion
Microsoft Advertising’s decade-defining upgrade signals that the era of manual bidding oversight is rapidly coming to a close. By moving toward a more predictive, value-centric model, Microsoft is empowering advertisers to capture more revenue—provided they are willing to provide the necessary data inputs and exercise the patience required for algorithmic learning. As we look toward 2027, the advertisers who succeed will be those who treat their campaign data not just as a set of numbers, but as the fuel for a more sophisticated, AI-driven machine.