The Latest PPC Upgrades, Reworks and Features

The PPC landscape is rapidly developing, even more so with AI becoming so prevalent across so many channels and platforms. Upgrades, reworks and newly implemented features are being rolled out constantly to keep up with the times, and below you will find some insights into some of the more notable changes throughout the month of March. 

Google Demand Gen Upgrades

As of March, lookalike signals no longer act like strict targeting signals and instead function as optimisation signals. These signals act as a guide for Google’s AI to find users, meaning that campaigns can now expand beyond selected similarity thresholds. 

While this means that precise audience control is no longer available, the alternative is a gained wider potential for reach and algorithmic expansion. Moving forward, overall performance will be more dependent on quality of conversion data, creative strength and bidding strategies. 

Lookalike segments now have shifted to ‘suggestion mode’ in Demand Gen campaigns, however it is possible to opt out through the manual action of a form request. To put it into simpler terms, Google is slowly pushing advertisers towards automation-first targeting, with manual audience segmentation becoming secondary. 

Additionally, Google has introduced an enforcement for minimum daily budgets within Demand Gen campaigns, meaning that small-budget testing will become more difficult to do. However, this implementation means that campaigns can more effectively leave the ‘learning phase’ earlier. 

OpenAI Begins Rollout of Sponsored Ads in ChatGPT Responses

For US ChatGPT users, sponsored ads have been appearing more frequently within responses received through the AI model, but according to those early adopters of the ad system, it’s unclear whether these ads are achieving meaningful business outcomes at present or not. OpenAI has confirmed their plan to show ads to all free and “Go” tier users (low-cost plan), starting with the US, but this is expected to be rolling out in the UK before the end of 2026 after some more fine-tuning. 

There seems to be two primary gaps within the current infrastructure: 

1. No self-serve platform

In order to buy an ad, there is a required manual back-and-forth with representatives at OpenAI, instead of an automated online system like with Google and Meta. This limits the scale of ads and slows down the process.  

2. Minimal reporting

Data from performance is delivered weekly via basic CSV files, which offers very little real-time insights or actionable metrics. This lack of dashboards and tools means that measuring ROI is difficult, and it is unclear at present if there is any noticeable positive change following the implementation of these ads. 

The primary goal appears to be the monetisation of free users who choose not to pay for the subscription service, and OpenAI has explicitly framed ads as a way to fund wider access to ChatGPT while keeping paid tiers cleaner. Early results have shown that the ad pilot reportedly hit $100M annualised revenue in approximately six weeks. 

OpenAI has disclosed that sponsored ads only appear when relevant searches are made and have no influence on the responses generated. These ads are also dismissible and are driven by user feedback. 

Microsoft Testing Almost Transparent “Sponsored” Labels in SERPs

Towards the end of March, Microsoft began testing an almost-transparent sponsored label on advertisements in SERPs to measure the impact of users having a preference between paid and organic results.  

As a result of this, advertisers may see higher click-through rates as users may click on ads accidentally or casually. With less visual distinction between paid and organic results in SERPs, copy, relevance, and landing page experience will become far more vital for boosting conversions from users who clicked without realising. 

However, there is a chance that users will begin to lose trust in brands/advertisers due to feeling misled and could trigger scrutiny from regulators for not clearly disclosing when a result is an advertisement. 

There is also the additional issue of finding it difficult when measuring statistics between organic and paid results due to not having a clear view of user intent and complicating attribution. 

Key Points for March 2026:

• Advertisers are moving away from manual keyword bidding. 

• With AI becoming more prevalent, quality of data is a priority. 

• Active management of “AI evaluation errors” will help to prevent campaign drift. 

For more information on all things PPC, please get in touch with the team today.