← Sticker Wall AGENT · DETAIL

ImagineAI Marketing Agency - Custom Agent with Ad Spend Commission, Monthly Income 50k RMB

Workflow: Every day, input brand positioning, industry keywords, and budget, and the agent automatically generates social media te

AGENT

Key Fields

FIELD STAMPS

IndustryMarketing / Advertising

RegionChina

ScaleSME

ChannelOnline

🔧 Workflow

Every day, input brand positioning, industry keywords, and budget, and the agent automatically generates social media text, images, and videos, while drafting target audiences and bids. Based on real-time ROI data, it automatically adjusts keywords, swaps materials, and halts low-performing campaigns, outputting weekly review reports for human judgment on optimization directions. Humans only need to spend 30 minutes a day reviewing agent output and ad campaigns, with the remaining time dedicated to expanding new clients.

🛠 Setup Requirements

Requires mastery of low-code agent orchestration platforms like Coze and Dify, calling generative capabilities such as GPT-4o, Jimeng, and Tongyi Wanxiang, and integrating open APIs of Ocean Engine or Kuaishou Maglev Engine. An MVP can be built in about 2-3 weeks, with no deep algorithm background required. Familiarity with ad platform data reports and callback interfaces is sufficient; those with Xiaohongshu/Douyin operation experience can get started faster.

🧰 Toolchain

🔧 Coze
🔧 Dify
🔧 GPT-4o
🔧 Jimeng
🔧 Tongyi Wanxiang
🔧 Ocean Engine Open Platform API
🔧 Maglev Engine API

💰 Revenue

① SME ad spend commission (main income): Brands pay a 5% - 10% service fee based on ad spend. A single client spends 50k - 100k RMB per month × servicing 5 clients = monthly income of 20k - 50k RMB (estimated range 12.5k - 50k RMB); Douyin feed CPM has risen to 75-120 RMB (media estimate, independently unverified), the proportion of income is not public (merchant self-reported caliber, independently unverified, data as of 2026); ② Performance-based client expansion: Expanding to 10 SME clients and sharing performance-based revenue can push monthly income past 80k RMB, actual client count unverified, share also unstated; ③ Batch material production on a per-use fee: Single professional marketing content production cost suppressed to 50-200 RMB (media estimate, independently unverified), content production cost/piece reduced by 81% (merchant self-reported data), charging clients per piece or package, quotes not public, order volume untraceable, proportion unknown; ④ Opportunity item - Multi-vertical industry agent template reuse: One person manages multiple accounts, exact number of accounts manageable not publicly disclosed, proportion of total revenue also not given.

💸 Cost

Tool subscriptions are about 200-500 RMB per month, various platform API fees based on call volume are about 500-2000 RMB per month, total cost is below 3000 RMB. If self-building servers to deploy the Dify version of the agent, cloud server fees will increase by about 200 RMB per month.

⏱ Time Investment

Initial setup investment is 2-3 hours per day for three weeks, later operations and maintenance take about 1 hour per day to monitor delivery and handle customer feedback, and half a day per week is spent reviewing and optimizing agent strategies.

🚀 Getting Started

First, select a niche industry such as catering or local life, use Coze to build two skills for content generation and delivery strategy, and connect to a real client account to run through the closed loop; the first order can use a performance-based payment model to lower the customer decision threshold, and after generating data, replicate to adjacent industries such as beauty, education, and e-commerce.

🔑 Keys to Success

✅ Trust mechanism of commission based on performance rather than monthly fees
✅ Accumulating industry-exclusive material libraries and delivery knowledge bases
✅ Humans control strategic direction while agents handle execution and iteration
✅ Multi-platform advertising interface redundancy to prevent single-platform policy changes

⚠️ 风险

⚠️ Platform advertising interface policy changes may cause delivery capabilities to fail, requiring multi-platform redundancy
⚠️ Customer industry material copyright or platform content review rule changes create ban risks, requiring built-in compliance verification
⚠️ AI-generated materials suffer from severe homogenization; without manual creative intervention, ad click-through rates may decay rapidly

📌 Real Cases

📌 A digital marketing practitioner managing online promotion for 5 brands introduced an AI system at the end of 2024. A single person's output became equivalent to 3 previous full-time marketing specialists. By 2026, through agency delivery commissions, they reached a monthly income of 50k RMB while simultaneously serving multiple enterprises to achieve compound growth.
📌 A local life service provider used a marketing agent built on Coze for a hotpot restaurant to generate Douyin text/images and group-buying delivery plans, increasing ad spend from 20k RMB to 80k RMB per month. Based on an 8% commission, monthly income was 6.4k RMB, with a customer repurchase rate reaching 70%.
📌 An independent developer used Dify + Jimeng to generate Facebook ad creatives for 3 cross-border e-commerce brands, with each creative set automatically producing 50 variants, increasing ad ROI by 1.8x, and achieving a monthly income of about 30k RMB via ad spend commissions.

Sources