A deep dive into Adobe’s Gen Studio

Scaling content generation for Performance Marketing

Adobe has recently released the first versions of Gen Studio—a tool designed to automate and streamline high-volume workflows for performance marketing. Adobe’s key positioning for the launch emphasizes the platform’s ability to generate tens of thousands of images quickly, representing a significant leap forward in how businesses approach content creation and management. This creates a serious competitive advantage for performance marketers possessing the scale and resources required to take advantage of this platform.

Adobe Gen Studio comes bundled with a version of Adobe Express that provides basic asset editing functionality, with connectivity into a variety of applications including Workfront and Experience Manager. Gen Studio also has Content Analytics bundled in the core as well.

Specifically, the base includes:*

• Users: Customer may use up to 10 Power Users and 20 Collaborator Users.
• Brands: Customer may define, store, and use up to 5 Brands.
• Storage: Customer may use up to 2 terabytes of Storage.
• Generative Actions (final rendered ads): Customer may use up to 60,000 Generative Actions per year.
• Insights (Analytics): Customer may use up to 25 million Rows of Data per year in connection with the Insights functionality.
• Adobe Express: Customer (1) via its Power Users, may access and use certain features and functionalities of Adobe Express as made available within the Express Embed SDK accessible from GenStudio for Performance Marketing and (2) via its Power Users and Collaborator Users, may access and use certain features and functionalities of the Adobe Express application directly.

*Shamelessly stolen from the Adobe product documentation 

In addition to generating ads using Sensei and Firefly, the tool automates metadata and keyword generation, performs brand validation, and offers remediation suggestions for “non-conformity to brand guidelines” — read: “Flag hallucinations before they go live.” Here’s some of the key highlights that I’ve learned:

This isn’t for everyone:

Adobe is the first to emphasize that, being an AI-first application, if you don’t have your workflows and asset management in place, this tool will not be useful for you. This is currently geared towards very high volume consumer package goods, high-volume consumer software, and other large scale organizations like travel and tourism.

You’ll need upfront work and planning:

This is not an application where you simply push a button and it spits out beautiful ads. There is significant upfront work and with planning and building the foundations required in order to take full advantage of Gen Studio. Practitioners need to specify products, brand guidelines, upload imagery and sample text, and define audience personas. I estimate that the upfront planning (not the app, but the planning before the app) in a typical organization would take several weeks (preferably before you sign the subscription contract) and then creating the guidelines within Gen Studio would probably take a day or so to build.

When the guidelines are created and uploaded and approved, Adobe generates a fine tuned model of your brand. This personalization ensures that the generated content aligns closely with the organization’s visual identity, reducing the likelihood of off-brand outputs. 

Generate the grunt work

Once the guidelines are in place and the fine-tuning has completed, that’s when the magic starts. You can create, iterate, and finally generate production ads that follow your brand guidelines. You can save formats as templates for re-use. Once you’ve settled on a style and template, you can generate production ads.

Currently limited to email, Meta (Facebook), and banner ads —with other large platforms coming soon— Gen Studio will develop and iterate a fantastic volume of images and text based on those brand guidelines. This relieves the grunt work that comes from creating a massive amount of value of formats.

I’m of two minds here. One, I’m very excited because in the past, content creation has stymied personalized communications. As a demand generation consultant, I’ve worked with clients to design audience segmentations and highly-targeted personalized journeys, however, because creative teams were so under-resourced, we could never get the fuel needed to power a good personalized demand generation program. Adobe Gen Studio solves this by taking that grunt work of hundreds or thousands of minor iterations and streamlines that creation using AI.

But the question comes, how would a junior designer or copywriter gain experience if this type of work is taken away? Where will they cut their teeth in the ad agency or corporate world? That’s a topic for a much deeper article on the impacts of GenAI within the creative space.

Cost of generation

Adobe provides 60,000 “Generative Actions” in a default subscription. Each ad or email typically costs five Generative Actions credits, for a total of 12,000 generative events per year. Subscribers can purchase more through their rep. I asked Russ Biehn, our demonstrator, what constitutes a Generative Action: if you are iterating or testing an idea, will you pay for those? Russ confirmed that you only pay for the Generative Actions used in an ad that is approved and published.

Automated metadata and keywords

As Adobe Firefly generates the imagery, Adobe has coupled an image-to-text generator model that will create keywords and tags in the meta-information. This is one of the standout features of Gen Studio. This data will become invaluable during the analysis areas of the tools, detailed below.

Streamlining approval processes

A vast library of generated content needs a significant resource pool to review and approve that content. To mitigate these challenges, Gen Studio incorporates built-in tools designed to streamline the approval process. The brand validation tool appears to employ a Generative Adversarial Network (GAN) model trained to identify inconsistencies from a company’s fine-tuned brand imagery, style, and voice. This model flags potential issues and offers remediation suggestions, enhancing efficiency without replacing human oversight entirely.

The adversarial model acts as an automated critical eye, identifying potential issues and flagging them in a separate area. This is a pretty neat feature and is a really smart inclusion into the product, because it acts as the brand police, but on a really high volume.

Suggestions for deviations from brand guidelines

The images that pass are pretty much good to go. But if content fails to meet the set criteria, Gen Studio  provides remediation suggestions. Specifically, it will call out basic non-conformity categories and provides actionable insights to improve your content to bring it back within brand standards. 

Content performance insights

The keywords and metadata that are created upon asset generation power the content Performance Insights tool, a powerful addition to the Gen Studio suite. Essentially, performance insights will cluster similar attributes and measure their effectiveness, often uncovering unexpected patterns like lighting, color or focal distance. You can split the analytics attributes in any level of granularity with this view. 

Getting real: Deploying Adobe Gen Studio for a brand

Now that you know the basic components, here’s how you would develop a brand within Gen Studio for performance advertising:

• Define your brand guidelines, voice and audience
• Upload assets to the system to fine-tune to your brand standards
• Create, iterate ad templates
• Generate production ads for email, Meta, or banner ads for review
• Approve or refine ads within the  brand validation screens
• Publish and analyze performance based on keywords, tags, and attributes
• Refine and profit. 

You can do this for multiple brands, so each brand (think of Unilever’s or Pepsico’s family of brands, for example) would have it’s own workspace.

Implementing Adobe Gen Studio

Adobe has publicly stated their commitment to ensuring a quick implementation process of Gen Studio with value realization within approximately eight weeks. This rapid timeline means that companies can see adoption and start seeing ROI quickly. Again, if you know you’re implementing Gen Studio, it would behoove you to get your business unit owners in a room to work through strategy weeks before signing that paper.

End-user training

Self-serve content has been available since December 2024. Certification is expected to roll out sometime at the end of 2025 once users have deeper familiarity with the tool.

Data privacy

At release, Gen Studio is powered by Azure API to OpenAI, with legal agreements that ensure that customer data is not shared or trained on, something I interpret as being analogous to how healthcare providers ensure privacy with third parties using a HIPAA Business Associate Agreement (BAA). Firefly is known to be trained on the Adobe Stock catalog, ensuring copyright provenance.

Future-forward thoughts on the product and it’s impact on creative professionals

Adobe promises further integration with their Experience Cloud applications and workflow managers in coming months. I’ve also heard talk of integrating third-party resources as well as third-party GenAI models, including bring-your-own self-hosted models are being discussed, but nothing official yet. The latter would be compelling for companies that have security mandates in place that restrict content sharing, but the product infrastructure supporting something like this would be significant.

Emerging roles and workflow implications for designers and marketers

The introduction of GenStudio also necessitates a reevaluation of traditional marketing roles and workflows. Brand designers, for instance, will need to consider inputs, technology, and endpoints more holistically. This shift mirrors broader trends in the industry, where digital transformation is driving new skill requirements and job descriptions. I think we’re going to see a whole new category of roles geared to managing performance branding assets, but that’s for another discussion.

Case study: Omnicom’s success with Adobe Firefly

A recent example of successful high-volume content generation is Omnicom’s Miranda brand image campaign. Using Adobe Firefly, the agency generated 540,000 images from 200 prompts for 200 subjects in 12 art styles—a staggering total of 1.5 million images in just five days. This case study underscores the potential of generative AI in performance marketing and highlights the need for robust safeguards and significant resources to review and approve such volumes of content effectively. This is a perfect example of high-volume work that could be managed by Adobe Gen Studio.

That asset tsunami you created? Workflow stressor par excellence

The sheer volume of content that Gen Studio can produce will have disruptive implications for existing business processes. Traditional workflows may need to be rethought entirely, given the exponential increase in creative output. This disruption is particularly pronounced in areas like approval and review, which require additional resources to manage effectively.

Generating tens of thousands of images quickly is great if you need them, but that tsunami of generated content needs storage, moderation, classification, and permissions. Generating tens of thousands of images will quickly swell your Digital Asset Management (DAM) system, and may overage trigger fees.

Adobe is very forthright in their warnings —if you don’t have your marketing house in order, if you don’t have disciplined workflows, templatized assets and campaigns, if you don’t have a good asset management system— you’re in for a house of pain through the sheer volume your organization will have to contend with.

Adobe’s own experience serves as a case study of the volume stressors. The company had to enlist outside agencies and studios to assist with reviewing and approving content due to the scale of creation. This necessity highlights the operational challenges that come with high-volume content generation, even for a tech giant like Adobe.

Adobe’s Gen Studio: niche product, huge implications for the industry

Adobe Gen Studio may redefine content creation and management processes in the same way that Ol' Grandaddy Photoshop did. While Gen Studio is cleverly positioned at launch to take full advantage of its strengths by focusing on Performance Marketing, I think we’ll see a sea change in design language and branding roles in the not-so-distant future.

As a high-volume brand management suite, its automated metadata generation, brand validation, remediation suggestions, and future integration plans make it a compelling tool for enterprises looking to streamline their performance marketing workflows. It surely will require additional resources and have a learning curve, but the benefits of high-volume personalized, brand-compliant ads are massive. 

I cut my teeth on a Mac II, Photoshop 1.07 and Illustrator 3.3. I still remember the announcement videos for the Adobe Portable Document Format as my colleagues realized the implications of PDFs for the design and printing industries. Adobe may have wandered around a bit, but take notice: Adobe owns the creative space, and their recent moves will have a fundamental impact on design and marketing.