Skip to content

Google Marketing Live Keynote Review 2023

Google Marketing Live Keynote Review 2023

Ash Linney

June 7, 2023

Reading time: 8 minutes


There were over 20 product announcements:


  • The core themes were AI, automated creative, privacy and brand safety, and Performance Max.
  • You can now create campaigns using Conversational AI in the UI (much like ChatGPT).
  • Generative AI will be able to create ads and assets for your business.
  • There were 4 new features announced for PMax:
    • Expansion of insights and reporting to include a lookback window for analysis and asset improvement recommendations.
    • Experimentation in Performance Max campaigns e.g. AB Test different assets.
    • Expanding New Customer Acquisition for additional goals.
    • OptiScore and Prefills will be integrated into the campaign setup flow.
  • AI-powered assets that align to search queries will be launched.
  • Google Product Studio means product images can be created and edited (powered by AI).
  • Merchant Center Next will replace Merchant Center and remove blockers for easily creating Shopping feeds.
  • 2 new upper funnel campaign types have been added called Video Views and Demand Gen.


“Your marketing multiplied by Google AI” was one of the main lines in this year’s keynote. In other words; feed the machine.


This year’s GML feels like one of the more significant shake-ups of the Google Ads platform. Emphasis was placed on improving data quality and input to enable bid strategies to work harder. We’re big believers in data input so can really get behind these changes.


There were over 20 product announcements and other new releases announced at this year’s Google Marketing Live Keynote. And what was the main theme? You guessed it… AI. This includes everything from the way assets and creatives are generated, to the way you will interact with the platform itself.


Privacy-safe marketing and performance also came up. People want transparency and control, so Google is making this front and centre. They say this improves performance, though it’s also partly because they have to (they don’t necessarily want to do this). Chrome is still moving to 3rd party cookie deprecation next year and first party data is being streamlined in the platform to make it easier to manage between teams.


Finally, there has been a big push in helping advertisers better their creative capability in order to run more upper funnel campaigns. This means more branding capabilities for brands, and likely more dollars for Google.


Read on for our POV of what’s interesting and useful.

Artificial intelligence (AI)

Create campaigns using conversational AI



No doubt spurred on by the emergence of ChatGPT, you will soon be able to create Google Ads campaigns with a Google AI-driven chat in the Google Ads interface.


This new asset creation feature uses Google AI to assist you via a chat function. Starting with just a URL, you will be able to have a conversation with Google Ads to support you in creating campaign elements such as ads and assets.


A primary benefit here is you can work with Google AI to improve ad strength and ensure all ads are within character limit. It looks useful and could be helpful for in the moment ideation.


Launch date: July.


Generative AI for automatically created assets

Google will also use generative AI to complement the conversational AI above, drawing information from landing pages and other data in the account. AI generated headlines will be used to match the customer’s query. Our recommendation: use this for ideation, but definitely QA the output and don’t rely on this solely without some human input.


Launch date: Later in the year.


Google PMax adds generative AI to create text assets & images

Google calls out how expensive it is for brands to create quality assets for images and video. They want you to be able to easily create a proof of concept, and also adapt existing successful creative. They cite that currently 25% of YouTube video ads are automatically generated and foresee this growing across all ad types.


Image creation using generative AI will be included to help generate creatives for image and video-heavy campaigns. To get started, once again a URL is required and Google will then populate the fields for you, while you are able to give additional prompts. This is a nice way to get to proof of concept quickly, mock up creative tests and make minor tweaks to already successful creatives to iterate on them further, without the usual bills and long lead times that often come from making brand new creatives.  

Performance Max

Insights and reporting

Since its conception and launch, Performance Max has been heavily criticised for being a “black box” in terms of reporting capabilities and insights. This has led to many more experienced advertisers developing their own workarounds and custom solutions in order to better understand basic reporting, such as spend by channel.

Google is now updating the search term insights report, expanding it to include additional categories, asset insights, and above all the ability to download these reports! This should not only lead to a clearer understanding of performance across different audiences, but also means you can easily build out search terms that are performing well in PMax as new Exact Match keywords in your Search campaigns.

Two additional features have been added including a lookback window for analysis and recommendations for asset improvement. We think any additional reporting to Performance Max is useful, but we’d also like to see Google go even further e.g. spend splits by channel.



There is now the ability to run multiple Performance Max experiments. This will be useful to help A) configure the setup and B) run additional tests for insights. These are tests that run PMax against itself and are a welcome addition to aid in identifying the opportunities for incremental gains in this now highly adopted campaign type.

This should hopefully help you better understand crucial elements such as the effectiveness of creative and copy, as well as particular PMax settings such as final URL expansion.


Expanding New Customer Acquisition

More goal options that are powered by AI are being introduced to Performance Max campaigns. This includes expanding the New Customer Acquisition (NCA) feature to optimise towards higher lifetime value customers, as well as other goal settings for returning customers.

This is a nice enhancement to a feature that we’ve found to be useful in pushing spend towards new customers when this is a focus for our clients. There are however a number of considerations that should be taken into account here when reporting, such as client expectations, but also the additional assigned values being pulled into primary conversions.

Launch date: Later in 2023.


OptiScore & Campaign Prefills

As with conversational AI, Google is looking to streamline the setup of Performance Max and make the experience easier and more helpful. As part of this, the Optimization Score component will be integrated into the campaign creation flow.

Advertisers will also have access to “pre-fills” options and suggestions during campaign setup. This should be useful to make campaign creation more effective and efficient. Be aware however that OptiScore recommendations don’t always align with business goals and as such should be reviewed appropriately and dismissed where necessary.

Launch date: Later in 2023.


Google Ads will create AI-powered assets that target search queries

Automatically created assets will soon gain more relevance thanks to Google AI.

Instead of building assets pre-auction, with these new features Google AI will be able to generate assets based on the user intent of an individual query.

It will be very interesting to see what kind of assets are created here. We’ve seen some strange recommendations in RSA AI-generated headlines so we would encourage regular reviews of the kinds of assets created here on their launch.

Launch date: Later in 2023.


Google Product Studio brings AI-generated images to advertisers



This new feature will give retailers the ability to create high quality images in minutes directly from feeds. Google cites that offers with more than one image see a +76% increase in Impressions and a 32% increase in Clicks.

With Google’s new Product Studio tool, merchants will be able to manipulate product photos, as well as edit, enhance or sharpen them via Google AI. You can also create dynamic backgrounds – such as overlaying products on rich lifestyle backgrounds – or remove unappealing backgrounds. You will get 4 options to choose from and can add these directly to your GMC.

We think this could be a really interesting feature as creative execution can sometimes take time due to briefing requirements and limited time internally. Here, the advertiser can make the changes to any flagged products immediately themselves. Of course, there will need to be a sign-off process for the proposed images.

This should also help address product disapprovals at inconvenient times. For example, if an image overlay has been flagged late on a Friday, it could be quickly edited out on the spot, so no traffic is lost ahead of the weekend. It also means advertisers can easily run AB tests for image testing such as cutout vs lifestyle.

Launch date: Available in the UK and US later in the year.

Merchant Center Next

Merchant Center Next to replace Merchant Center

Google Merchant Center Next will soon replace Google Merchant Center.

This will be the new simplified place for product feeds and will use website data in the feed creation process to remove some of the harder technical work required for creating feeds. This means brands won’t need to upload products manually or create specific processes to push products into GMC. There will also be an improved UI as well as insights in the Performance Tab.

We like this one a lot because many brands have the product pages available to run on Google Shopping but don’t have the expertise to set this up. This should remove several barriers and therefore unlock some really valuable highly visual product reach for brands.

Launch date: 2024.

New Campaign Types

2 new campaign types have been added: Video Views and Demand Gen

New Video Views and Demand Gen campaigns should be used for enhancing YouTube views and demand generation for advertisers.

Video view campaigns will combine skippable in-stream ads, in-feed ads, and Shorts ads to get the most views for a video.

Demand Gen campaigns will show across YouTube Shorts, YouTube in-stream, YouTube in-feed, Discover, and Gmail to drive conversions. A new lookalike segment builder will also help advertisers expand their reach.

Combining channels and ad formats under a single campaign type is a common theme at Google. This can be helpful in some ways because it means budget will be allocated to the best performing segments or channels, but also be careful, because when targets are met, Google can go after low quality inventory to get closer to target.

Our take

Our Take

There were a lot of updates this year, some of which are very welcome. We particularly like the releases that give advertisers more insights and actionable data points to work with. We’re also fans of updates that really lean into first party data usage, which is key to powering future advertising. There are some areas however where we would encourage vigilance –  such as any AI-generated assets. As with all new releases, human monitoring will be key here, and we’d always urge caution with anything that could go live without review.


Customer Lifetime Value also came up. As mentioned in the introduction, data quality and what we send to the platform is super important to us so CLTV, margin, and any other offline conversions should continue to form a significant part of any strategy that is developed.


Owning first party data is hugely important. As the 3rd party cookie deprecates next year, we should leverage offline conversions, audience creation, etc. to future proof brands and their performance. Google is still working out how to manage audiences so it will be interesting to see the developments as the deadlines approach.


With the push for advertisers to do more upper funnel activity via improved creatives and Performance Max, you would hope that reporting and attribution would be more of a focus, but this didn’t get a significant look-in. Yes, advertisers have been listened to in some respects, such as the expansion of the insights tab, but in order to get the best out of Performance Max you will still need custom solutions.


The SERP is becoming highly integrated, interactive, and visual (e.g. Search Generative Experience). All of this clearly reflects Google’s ad direction at GML with helping brands create more visual content, and generate text a lot more efficiently. It’s early days, but making sure accounts are ready for these changes and agencies / clients have the relevant processes in place will be paramount.

Share your thoughts

We hope you enjoyed our take on this year’s GML.

Reach out to us