Adobe
in-house genAI production of marketing content at scale
Adobe runs its own marketing on Firefly and Generative Fill: social campaign production time cut by 33% and content volume doubled for Adobe's social media.
Key points
- Adobe produces its own marketing content with generative AI.
- Firefly and Generative Fill in Photoshop, industrialized through GenStudio.
- Social campaign production time cut by 33%, volume doubled.
- Evidence level B, live status confirmed.
Objective
Absorb the growing demand for social content and campaigns without inflating costs and timelines. Adobe uses its own generative tools to produce more assets, faster, while keeping brand consistency.
The deployment
Adobe runs its own marketing on Firefly and GenStudio. The marketing teams use Generative Fill in Photoshop and Firefly-generated images to produce the brand's social content. Adobe built GenStudio for Performance Marketing by dogfooding these tools: its teams produced tens of thousands of assets for a single campaign. Adobe Brand Studio also used Firefly for the visual identity of Adobe MAX 2024, generating hundreds of on-brand assets. The scope covers the creation, adaptation, and scaling of campaign content.
Results Proof B
Figures published by Adobe in its own official press release (33% less time, volume doubled) and internal use confirmed by a second Adobe release (Brand Studio, MAX 2024). Official quantified release from the subject brand, consistent sources.
How it works
Documented architectureThe stack in detail
- llm Adobe Firefly Adobe's generative image models, trained on licensed content, used to produce campaign and social visuals.
- llm Modeles Firefly personnalises Versions of Firefly trained on the brand's assets to keep visual consistency at scale.
- plateforme Adobe GenStudio Content supply chain that industrializes creation, multi-format adaptation, and validation of campaign assets.
- outil Adobe Photoshop (Generative Fill) Generative fill used to extend and adapt visuals to social formats (Instagram, TikTok).
How it runs, concretely
For ops teams-
1Campaign framing Marketing
Marketing defines the brief, the social formats, and the brand rules.
-
2Asset generation AI
Firefly generates images and variants; Generative Fill in Photoshop extends and adapts the visuals.
-
3Multi-format adaptation AI and marketing
The assets are adapted per platform (Instagram, TikTok) via GenStudio.
-
4Brand control Brand Studio
The teams validate on-brand consistency before publication.
-
5Publication and iteration Marketing
The content goes out on social channels; volumes and timelines are measured.
The brand guidelines and the customized Firefly models. Without a model trained on the brand assets, generated content drifts and requires more manual rework, which cancels out the time saving.
How your customers perceive this type of use
Sourced studiesUn ecart net separe les annonceurs des consommateurs : 77% des annonceurs voient l'IA positivement contre 38% des consommateurs (Yahoo/Publicis, 2024). Les mesures implicites confirment le rejet declare : en EEG, les pubs generees par IA produisent une activation memorielle plus faible que les pubs traditionnelles et sont decrites comme agacantes, ennuyeuses et confuses (NIQ, 2024). La disclosure a un effet ambivalent : elle augmente fortement la confiance quand elle est remarquee (Yahoo/Publicis), mais 27% des jeunes consommateurs disent faire moins confiance a une entreprise dont la pub est creee par IA (IAB, 2024).
Acceptance conditions
- Une disclosure visible : quand la mention IA est remarquee, la confiance globale envers l'entreprise augmente de 96% (Yahoo/Publicis 2024)
- Une qualite visuelle suffisante : les visuels IA de basse qualite augmentent l'effort cognitif et distraient du message (NIQ 2024)
Red lines
- Le contenu IA non declare puis identifie : 72% des consommateurs disent que l'IA rend l'authenticite difficile a etablir (Yahoo/Publicis 2024) et les marques utilisant des pubs IA sont plus souvent jugees inauthentiques ou non ethiques par les consommateurs que par les dirigeants (IAB 2024)
- Les mannequins et personnes generes par IA : 46% des consommateurs n'en veulent pas dans la publicite, l'inquietude premiere etant les standards de beaute irrealistes (Attest 2025)
Sources: Yahoo / Publicis Media (terrain Ebco) 2024 · IAB (avec Attest) 2024 · NIQ (NielsenIQ) 2024 · Attest 2025
How to replicate
Inference, not sourcedData prerequisites
- bank of brand assets to train a custom model
- formalized brand guidelines
- structured campaign briefs
Org prerequisites
- creative team or brand studio
- on-brand validation process
- tool to measure production time and volume
Possible stack
- generative image AI (Firefly or equivalent)
- editing tool with generative fill
- content supply chain / DAM
The plan, step by step
- Step 1Formalize the brand guidelines and gather the bank of reference assets.Deliverable: Asset corpus and documented on-brand rules
- Step 2Train or configure a customized generative model on these assets and equip the creators (generative fill in the editing tool).Deliverable: Custom model tested on real briefs
- Step 3Frame a first social production flow: formats, templates, on-brand validation step.Deliverable: Documented production workflow with a brand control point
- Step 4Produce at volume, adapt per platform, and compare production time and volume to the previous period.Deliverable: Published campaign + before/after measure of time and volume
First step: Train a generative model on the brand assets and frame a first measurable social production flow.
The same pattern, proven in other industries
BMW
advertising and experiential creation with generative AI
Coca-Cola
genAI video production of a global-scale brand campaign
Colgate-Palmolive
genAI video and marketing content production at scale
Sources
- S1 Adobe GenStudio Accelerates Enterprise Content Supply Chain with Customized Firefly Models Primary archive pending
- S2 Adobe Expands GenStudio Content Supply Chain Offering for Marketing and Creative Teams Primary archive pending
An error, newer info, a source?
This page lives on its accuracy. If a figure has moved, if the deployment has changed, or if you have a higher-quality source, tell us. Every sourced correction is verified before publication.