In 2026, the advertising industry is experiencing a quiet earthquake. Not a revolutionary upheaval that overturns everything overnight, but more like the frog-in-boiling-water scenario—you notice that competitors' quotes are 30% lower, their delivery timelines are twice as fast, and the finished quality looks about the same, if not better.
1. Three Key Pain Points Facing Traditional Ad Agencies
First: Rising labor costs
A complete creative team (art director + designer + copywriter) plus project management and post-production can run 150,000–300,000 RMB per month. Senior designers in first-tier cities routinely earn over 20,000 RMB per month, and the retention rate for top talent is below 60%.
Second: Increasingly demanding client delivery timelines
"Show me the first draft by tomorrow morning" has gone from an unreasonable demand to an industry norm. This is especially true during major e-commerce promotions—the main image design demand during the Double 11 preheat period is typically 3–5 times the usual volume, but team headcount does not temporarily increase.
Third: Homogenized competition driving down profit margins
When all ad agencies use the same creative methodology and templates, it becomes increasingly difficult to establish a differentiated advantage. Clients have begun comparing quotes on a per-unit basis rather than project-based packages, continuously squeezing profit margins.
2. Entry Points for AIGC Tools in the Ad Workflow
Pitch stage: Rapid mood board generation
Traditional approach: Designers spend half a day collecting reference images from Pinterest and Behance and assembling them into a PPT. AI-accelerated approach: Use Midjourney v6/GPT-IMAGE2 to input keywords and directly generate style-consistent visual direction drafts—you can produce 4–6 sets of different mood boards for the client to choose from within 30 minutes.
Design execution stage: Batch first-draft production
Highly standardized assets like e-commerce main images, posters, and banners can have 80% of their foundational work completed by AI—generating compositions and color schemes. The designer's role shifts from "drawing from scratch" to "review + refinement + brand tone control."
Video ad stage: Animation acceleration
The demand to convert static posters into short videos is growing (clients say, "This image is great—can you turn it into a 15-second Douyin video?"). Traditionally, a motion designer would need to rebuild it from scratch, increasing costs 3–5x. Now, tools like LTX-2.3, Seedance 2.0, and Kling 3.0 can directly animate static images, adding camera movement trajectories and element animations.
3. Recommended Transition Path
Phase 1: Tool trial period (1–2 months)
Select 1–2 designers with high technical receptivity as "AI pioneers" and have them experiment with using MJ v6 for inspiration exploration and first-draft generation in their daily work. Do not demand immediate results—focus on building intuitive familiarity with the tools.
Phase 2: Process integration period (3–4 months)
Incorporate AI tools into standard SOPs. For example: the pitch stage must include at least 2 sets of AI-generated visual direction drafts as reference; e-commerce main image design adopts a hybrid model of "AI-generated first draft + designer refinement."
Phase 3: Full integration period (6+ months)
Build the company's prompt library and style template assets. Extract the excellent design styles from past successful cases into reusable prompt parameter combinations, forming a competitive moat—this is the core capability that competitors cannot copy.
4. Pitfalls to Watch Out For
Pitfall one: Using AI to completely replace designers
AI's current capability boundary is at the execution layer (generating assets)—creative strategy and brand tone control still require human designers. Blindly cutting senior designers will cause pitch quality to plummet.
Pitfall two: Over-relying on a single tool
MJ v6 excels at photorealistic imagery but still has limitations in typography and text rendering; Stable Diffusion offers strong controllability but has a steep learning curve. It is recommended to build a "primary + backup" tool matrix and switch flexibly based on project needs.
5. Team Training and Knowledge Management
It is recommended to hold a 30-minute internal sharing session every week, where designers who have achieved good results with AI tools demonstrate their workflows. At the same time, build the company's prompt knowledge base—categorize and archive proven effective prompts by style, so new team members can get up to speed quickly after joining.
