CakeGrowth is a one-stop affiliate marketing platform focusing on AI applications, connecting AI product parties (advertisers) and traffic owners, providing promotion links, data tracking and commission settlement, helping AI tools acquire users and promoters to obtain sustainable commission income.

CakeGrowth

Core parameters and statistics of CakeGrowth

CakeGrowth is not a consumer-oriented AI application, but a one-stop affiliate marketing platform focusing on the AI industry. Its core value is to "accurately match the customer acquisition needs of the AI product side with the monetization needs of traffic owners in a vertical market." The core contradiction it solves is: AI tools lack low-cost, controllable promotion channels, while content creators lack sustainable, highly matching AI monetization targets.

Projects	Public Information
Platform positioning	The first one-stop affiliate marketing platform focusing on AI applications
Operating entity	Hefei Xinyan Technology Co., Ltd. (Hefei, Anhui)
Bilateral roles	Advertiser (AI product side) / Traffic owner (promoter)
Cooperation model	CPS (commission based on sales), CPA (commission based on behavior), CPS+CPA hybrid
Product scale	About 28 promoteable AI products in the alliance market (as of 2026-07)
Channel coverage	300+ partner websites 100+ vertical field KOL, WeChat/Zhihu/X/YouTube
Daily exposure volume	Official claim: 100 million+ daily exposure
Average monthly traffic pool	Official claim 20 million+
Attribution technology	Client-side JS + server-side S2S dual attribution
Support Platform	Web
International site	cakegrowth.com

Publicity Verification: The official website advocates "the first one-stop affiliate marketing platform focusing on AI applications" and displays marketing data such as 300+ partners and $50k+ revenue cases. The platform has indeed settled on well-known AI product developers such as Volcano Engine, SenseTime Seko, LiblibAI, and iFlytek. The product catalog covers a total of 28 tools in 8 AI subcategories, and the vertical positioning is credible. However, macro data such as daily exposure and average monthly traffic pool are self-reported marketing calibers of the platform and lack independent third-party verification. They should be regarded as a reference for intention rather than a contract commitment.

Core Difference: Compared with general alliance platforms (such as Commission Junction, LinkShare), CakeGrowth's differentiation is not in the technical architecture, but in the vertical aggregation of "only doing AI" - the AI ​​product side gets promoters who understand AI better, and the promoters are aimed at audiences who are more receptive to AI tools. The matching efficiency is theoretically higher than that of pan-platforms.

CakeGrowth’s users and market recognition

CakeGrowth's current market recognition is mainly reflected in the brand endorsement of the settled product parties and the coverage density on the channel side, rather than public GMV or profit data (the latter is not officially disclosed).

Product side: The alliance market has included 20+ AI product parties such as Volcano Engine (owned by ByteDance), SenseTime Seko, iFlytek LiblibAI, Atoms, MaShangfei, etc. Judging from the quality of settlement, there are both products from major manufacturers (Volcano Engine, iFlytek) and leading AI vertical track tools (LiblibAI, Xingliu), indicating that the platform has certain market appeal on the AI ​​supply side. However, presence does not mean active advertising, and the advertiser’s actual continuous advertising ratio and ROI data are not disclosed.

Traffic main side: The official website is marked with 300+ partners and 100+ vertical KOLs, covering WeChat official accounts, ZhihuX, YouTube and other channels. The commission example shows "income $50k+, traffic owner +120, commission $8k+", but this is a marketing representation, and the real average monthly income, withdrawal success rate, and settlement on-time rate of a single traffic owner are not disclosed.

Industry benchmarking reference: In the domestic Internet alliance marketing track, there have been platforms such as Yiqifa and Duomai serving the e-commerce and gaming industries. CakeGrowth chose the AI ​​vertical cut, and the differentiation on the product supply side is obvious. However, the user decision-making cycle of AI tools is usually longer than that of physical e-commerce - users need to understand the product, try it out, and confirm the value before paying. This will affect the conversion efficiency and revenue stability of traffic owners, and is easily ignored when evaluating the value of the platform.

Cost Advantages of CakeGrowth

CakeGrowth's cost structure is not a single dimension of "expensive or cheap", but depends on the role you play in the two-sided market and the cooperation model adopted.

Advertiser (AI product side) side:

• Explicit Cost: Pay commissions based on CPS/CPA, only incur fees when there are conversions, no upfront advertising fees. It is free for traffic owners and advertisers to join, and there is no annual fee or deposit (subject to official real-time rules).
• Hidden Cost: The commission ratio is usually 15%-70% of the order amount. For low-priced SaaS products (such as a monthly subscription of 30 yuan), the absolute low commission may cause the traffic owner to lack motivation for promotion; in addition, manpower needs to be invested to configure tracking links, monitor conversion data, and process settlement reconciliations.
• Opportunity Cost: If the product's own conversion rate is low (bad trial → paid funnel), even if the commission ratio is high, promoters will not be willing to promote it - the platform effect is highly dependent on the product's own conversion ability.

Traffic owner (promoter) side:

• Explicit Cost: Zero threshold for entry, no subscription fees or deposits.
• Hidden costs: Promotional material production, content output, fan trust consumption, etc. For the CPS model, the commission settlement cycle ranges from 1 month to 12 months, and the revenue collection cycle is longer than the content platform advertising share.
• Income Fluctuation: AI tools belong to the category of "high customer unit price but long decision-making chain". Compared with FMCG CPS, traffic owners may take longer to see stable income.

Enterprise/large-scale advertising main development path:

• Initial stage (budget within 100,000 yuan): directly use the platform CPS/CPA model and pay based on performance.
• Growth period (500,000-2 million budget): You can negotiate with the platform for exclusive commission ratios, exclusive promotion activities, and customized tracking needs.
• Scale (budget of more than 2 million): Enterprise-level services such as customized API docking, privatized data dashboards, and advanced anti-cheating audits may be required. This part of the capability has not been officially disclosed and requires business confirmation.

Cost comparison: affiliate marketing vs other customer acquisition channels

Customer acquisition channels	Cost structure	Applicable stage	Risk points
CakeGrowth Alliance	Pay by conversion (CPS/CPA), no prepayment	The product is online and there are paid conversions	Depends on promoter network density, long settlement cycle
Information flow advertising (Jumbo/Guangdiantong)	CPM/CPC prepaid, requiring continuous optimization	Brand exposure + effects are required	Traffic costs continue to rise, and the optimization threshold is high
KOL direct investment	Fixed price or CPx mix	New product cold start/brand endorsement	Single investment ROI fluctuates greatly and cannot be sustained
SEO/ASO organic volume	Early content + technical investment	Long-term stable customer acquisition	Effective cycle 3-6 months, risk of algorithm changes

Main functions of CakeGrowth

• AI Application Alliance Market: Gathers 28 AI products, displayed in 8 categories: office efficiency, text creation, creative design, image tools, video creation, audio tools, and programming and development of AI agents. Each product is marked with the commission mode (CPS/CPA/Mixed), commission amount/proportion, cookie validity period/commission cycle. Traffic owners can independently choose promotion targets based on commission attractiveness and product matching.

  • Expert View: Classification granularity directly determines matching efficiency. The eight categories cover the current mainstream tracks of AI tools, but they lack "education/academic" and "data analysis", two categories that have grown rapidly in recent years. You can pay attention to the pace of their category expansion in the future.

• Promotion link generation and attribution tracking: After the traffic owner selects the product, the exclusive promotion link is generated with one click, and the system tracks the user behavior link through the client JS code + server S2S dual attribution model.

  • JS side: Plant a cookie when the user clicks the promotion link to identify the promotion source.
  • S2S side: The advertiser's server returns real transaction/behavior data and cross-verifies with the JS side.

Expert View: Dual-track attribution is CakeGrowth’s relatively solid infrastructure. A single JS cookie can easily be cleared or blocked. Adding S2S postback can fix the leak and provide advertisers with anti-brushing verification methods. However, S2S relies on advertisers’ active connection and data integrity. For small and medium-sized AI product developers, the technical cost of backhaul access and the complexity of reconciliation need to be taken into consideration.

• Commission settlement and revenue dashboard: Supports two settlement modes: CPA (based on the number of effective users) and CPS (based on the percentage of the order amount). Traffic owners can view the clicks, conversions, estimated revenue, and settled revenue of each product. Advertisers can view advertising expenditure, conversion cost and ROI.

  • Commission cycle difference: The commission cycle of platform products covers four types: "1 month/3 months/12 months/permanent" - for example, Daydream AI provides "permanent recurring commission", the card is 3 months CPS, and the super resume WonderCV is 12 months CPS. Traffic owners need to plan their revenue expectations based on the commission cycle.

• Fully managed agency operation service: Provide advertisers with full-process agency operations including project configuration, release changes, and communication with traffic owners. Advertisers only need to pay attention to the data results, and the platform is responsible for connecting and expanding traffic channels.

  • Expert View: Agent operation is a double-edged sword - it can reduce the burden on small teams with tight manpower, but for product parties who want to finely control the delivery strategy, they need to weigh the granularity of "leaving it to the platform vs. controlling it themselves".

• Anti-cheating and traffic anti-counterfeiting: Built-in click quality detection mechanism to identify abnormal click patterns and false traffic. Under the performance-based payment model, anti-fraud capabilities directly determine whether advertisers are willing to continue advertising.

CakeGrowth’s model and version evolution

CakeGrowth is a SaaS platform product that does not follow the version number release rhythm of traditional software. The following evolution is based on the public function milestones on the official website, and there is no official version number statement.

• Early online version (~2025-12): Complete the construction of the alliance market, and launch basic tracking and commission settlement capabilities. The initial product supply is mainly small and medium-sized AI tools, and the bilateral matching algorithm and attribution model are in the early verification stage.

• Functional improvement period (2026-01 to 2026-03): Gradually introduce JS+S2S dual-track attribution, traffic anti-counterfeiting detection, and fully managed agent operation services. The number of products in the alliance market has grown to 20+. Online advertiser and traffic owner RBAC role management.

• Platform version/current status (~2026-04 to present): The platform product scale has reached approximately 28 models, covering 8 AI subcategories. The international website cakegrowth.com is online (for overseas promotion needs of overseas AI products). The official website claims that the number of partners has increased to 300+ and KOL covers 100+. Online help center and about page to improve platform endorsement information.

Version evolution characteristics: The iteration rhythm of CakeGrowth reflects the cold start process of the bilateral market - first do supply (pull AI products to settle in), then demand (pull promoters), and then consolidate bilateral stickiness through data and settlement systems. We are currently at a stage where the supply density is basically sufficient, but the demand side still needs to continue to expand. In the future, attention should be paid to the iteration frequency of its anti-cheating algorithm, the degree of settlement automation, and whether it has open API docking capabilities.

Technical advantages of CakeGrowth

CakeGrowth's technical barrier is not in the underlying AI model or high-performance computing, but in the "data link governance capabilities in affiliate marketing scenarios" - it requires the establishment of a credible tracking, attribution and settlement system in a two-sided market with asymmetric information.

Dual-track attribution mechanism (JS + S2S): This is the core technical capability of CakeGrowth. The client-side JS Cookie captures the click link of the promotion source, and the server-side S2S receives the advertiser's real conversion response. The two are cross-compared to form an auditable conversion evidence chain. Compared with the pure JS solution, the dual-track design can resist scenarios such as cookie clearing and cross-domain tracking loss; compared with the pure S2S solution, the JS side provides independent promotion source credentials, reducing advertisers' incentives to under-report.

Anti-cheating detection: The platform has a built-in click quality detection engine that identifies abnormal traffic by analyzing click frequency IP distribution, device fingerprints, conversion time windows and other dimensions. In the performance-based payment model, the accuracy of anti-cheating directly affects advertisers’ confidence in advertising and traffic owners’ commission fairness. The specific threshold of detection capabilities, manslaughter rate and appeal mechanism have not been officially disclosed. It is recommended that advertisers test the tolerance of brushing on a small scale in the early stages of launch.

Vertical Category Supply and Demand Matching: Unlike the full category distribution of general alliance platforms, CakeGrowth limits both supply and demand to the category of "AI tools", reducing bilateral information retrieval costs and making promotional content more likely to resonate among AI interested groups. The technical implementation is the matching of category target label system + promoter portrait, but the precision of its matching algorithm has not been officially disclosed.

Engineering Limitations: The current platform does not have a publicly open API, which means that advertisers and traffic owners mainly rely on the Web console for data operations, and cannot achieve automated reconciliation or batch management - this may cause operational efficiency bottlenecks for advertisers who manage a large number of promotion links. Whether the data of the international website (cakegrowth.com) and the domestic website are connected and whether the settlement is unified also needs to be confirmed with the business owner.

How to use CakeGrowth

The usage path of CakeGrowth is designed around bilateral roles. The entrances at both ends are unified but the operation processes are different.

Entrance control:

How to use	Suitable for roles	Operation entrance	Prerequisites
Web console	Traffic master (promoter)	dash.cakegrowth.cn registration/login	Have audience channels that can be promoted
Web console	Advertiser (product side)	dash.cakegrowth.cn registration/login	Have AI products that can be launched
Fully managed agency operations	Advertisers	Contact platform business activation	Confirm agency operation scope and SLA

Traffic main operation link:

1. Registration: Visit cakegrowth.cn, click "Traffic Master Registration", and complete account registration and channel information.
2. Product selection: Browse the alliance market, filter AI products by category, and check the commission model (CPS/CPA), commission amount, and settlement cycle.
3. Generate promotion link: After selecting a product, generate an exclusive promotion link with one click and support custom parameter tracking.
4. Content Promotion: Embed promotion links into public account articles, Zhihu Answers/Twitter posts, YouTube profiles and other channels.
5. Data Monitoring: Check the clicks, conversions, and estimated revenue of each product on the revenue dashboard.
6. Withdrawal: Apply for withdrawal after reaching the settlement threshold, and the settlement will be according to the platform rules.

Advertiser operation link:

1. Settlement configuration: Register an advertiser account and submit product information and promotional materials.
2. Set commission sharing rules: Configure commission mode (CPS/CPA), commission amount/ratio, cookie validity period or commission cycle.
3. Access Tracking: Deploy JS tracking code + configure S2S return interface (technical docking required).
4. Audit Traffic Master: Can manually review or automatically open promotion applications.
5. Monitoring and Analysis: View advertising expenditures, conversion cost ROI data, and evaluate promotion effects.
6. Settlement Confirmation: Confirm the settlement data after the end of the cycle and complete reconciliation and payment.

Notes on docking: S2S callback requires the advertiser's technical team to develop a callback interface on the server side. The docking cycle is usually 1-5 working days. It is recommended to confirm the return interface specifications, test context and exception handling process with the platform before settling in.

Product Pricing for CakeGrowth

CakeGrowth's charging model is a typical performance-based commission, with no front-end platform usage fees. The pricing focuses on the commission ratio and settlement terms.

Traffic master (C client/individual):

• Setting fee: 0. There are no security deposits or annual fees required.
• Source of income: Promotion conversion commission. The CPA model is usually ¥3-¥30/effective user, and the CPS model is usually 15%-70% of the order amount.
• Billing Period: Varies by product, ranging from 1 month to 12 months (there are also permanent recurring commissions, like Daydream AI).
• Minimum withdrawal threshold: Undisclosed, subject to the real-time rules of the platform.

Advertiser (Developer/Product Side):

• Setting fee: 0. No annual, monthly or listing fees.
• Main Spend: Commission paid at CPS/CPA. The commission ratio is set by the advertiser, and the platform usually recommends referring to the level of similar products.
• Additional Cost: If fully managed agent operation service is enabled, the cost needs to be confirmed with the business, and the official standard pricing has not been disclosed.
• Technical Service: S2S backhaul docking is free, but requires the technical team to invest in development and testing manpower.

Enterprise/scale advertisers:

• Enterprise-level services such as customized commission plans, exclusive promotion schedules, exclusive data dashboards, and advanced anti-fraud audits are subject to business negotiation, and the official standard price list has not been disclosed.
• The promotion of the international station (cakegrowth.com) covers overseas channels, and the currency, exchange rate and tax treatment of cross-border settlement need to be confirmed separately.

Reference for price comparison with competing products: Industry mainstream alliance platforms (such as Impact, PartnerStack, FirstPromoter) usually charge platform commissions (5%-20% commission) or SaaS subscription fees ($99-$999/month). CakeGrowth's zero-threshold strategy on the entry-end reduces advertisers' upfront costs, but please note that whether the platform collects intermediary fees for commission flows, platform service fees for different commission models, and other terms have not been fully disclosed on the public page. It is recommended to request a complete rate schedule before settling in.

Application scenarios of CakeGrowth

The value of CakeGrowth focuses on the two interlocking needs of "AI products need low-cost customer acquisition" and "content creators need high-quality monetization targets."

• Customer acquisition based on the performance of AI tools: AI product developers (especially small and medium-sized teams) share commissions through CPS/CPA, allowing promoters to become a sales team that pays based on performance. Compared with the CPM prepayment of information flow advertising, this model has less cash flow pressure and the conversion data can be audited. Typical applicable objects are AI SaaS products that have stable paid conversions and need to be scaled up.

• AI monetization track for content creators: Public account bloggers, Zhihu answer masters/Twitter creator YouTubers, etc., convert content traffic into commission income by promoting AI tools. AI tools have high content discussion value and topicality, and are naturally suitable for evaluation and tutorial content creation. The core income of traffic owners is not a single conversion commission, but long-term content reuse and commission accumulation.

• AI new product cold start and market education: Newly launched AI products lack brand recognition, and content is cultivated through a network of promoters - promoters lower the decision-making threshold of potential users in the form of personal test experiences, comparative reviews, tutorials, etc. During the cold start phase, it is recommended to give priority to the CPA model (pay per effective user) to control budget risks.

• Local promotion and global coverage of overseas AI products: Reach overseas promoters through the international website cakegrowth.com, covering overseas channels such as X and YouTube. For Chinese AI product developers, CakeGrowth provides the convenience of "one platform to manage domestic + overseas promotion", but the settlement currency and tax compliance of overseas promoters need to be confirmed with the platform.

• Cost reduction and efficiency improvement deduction:

  • New Media Operations (Content Team): Assume that 2 AI tool recommended articles are produced every week and a promotion link is attached. The traditional monetization method (such as receiving software advertising) costs about 1,000-5,000 yuan per article and requires business negotiation. After using affiliate marketing, the lifetime commission income of an article can be estimated to be hundreds to thousands of dollars, and passive income will continue to be generated (especially permanent commission products).
  • AI product side marketing/growth team: The customer acquisition cost of traditional information flow advertising is about 30-80 yuan/effective user (depending on the category). After the platform is settled, the CPA model may drop to 3-30 yuan/effective user - but it is limited by the promoter's coverage scale and content quality.

Applicable groups of CakeGrowth

CakeGrowth’s three core user roles and their adaptation scenarios are as follows:

• AI product owner/advertiser (growth team or founder): Owns an AI product and needs a stable and controllable customer acquisition channel. Suitable for products that have paid conversion data and can pay based on performance in CPA/CPS mode. Prerequisite: The product must complete at least one round of paid conversion verification, otherwise the promoter's conversion rate will be too low, resulting in insufficient promotion motivation.

• Traffic owners/promoters (content creators, self-media operators, community managers): Have a certain audience base (official accounts/Zhihu/Xiaohongshu/X/YouTube/WeChat groups, etc.) and hope to use AI tools as monetization targets. Recommended Threshold: It is recommended to have at least 1,000+ active fans or stable content production capabilities, otherwise commission income may be difficult to cover content creation costs.

• AI Product Business/Ecology Leader: Responsible for handling affiliate marketing strategies, commission negotiations, data docking and compliance audits for enterprise-level advertisers. Channel management and data analysis capabilities are required.

• Does not fit boundaries:

  • Pure brand exposure needs: What is needed is brand advertising (CPM/CPT), not pay-for-performance affiliate marketing. CakeGrowth's commission model is based on real conversions and is not suitable for delivery goals that only require brand exposure.
  • Individuals without promotion channels: After simple registration, they have no promotion ability and cannot generate commissions. The platform will not allocate traffic or promotion tasks to idle accounts.
  • Low conversion rate products: If the conversion rate of a product from trial to payment is less than 1%, the promoter's input-output ratio is too low, making it difficult to promote spontaneously.
  • High customer unit price and low frequency decision-making products: For example, enterprise-level AI solutions (annual payment of hundreds of thousands) have complex user decision-making chains and are not suitable for standardized promotion link tracking.

Summary and Outlook

CakeGrowth starts with "AI vertical alliance marketing", and its core competitive advantage lies in the dual-track design of AI product aggregation and attribution technology on the supply side. Compared with general alliance platforms, its matching accuracy in the AI ​​​​track is higher; compared with self-built promotion systems, it provides standardized tracking, settlement and anti-cheating infrastructure, reducing bilateral docking costs.

Current Limitations:

• Core operating data such as the platform’s real GMV, number of active promoters, and average revenue of traffic owners are not disclosed. The revenue cases displayed on the official website are marketing hints and should not be used as a basis for contract decision-making.
• The API interface is not open, and advertisers and traffic owners rely on the Web console for data operations, which limits efficiency in large-scale management scenarios.
• The promoter coverage depth and settlement maturity of the overseas station (cakegrowth.com) need to be verified.
• The specific thresholds, manslaughter rates and appeal mechanisms for anti-cheating detection have not been made public, and long-term credit protection requires continuous observation.

Follow-up observation points:

1. The growth rate of settled products and the coverage rate of top products: Whether it covers more first-line AI products (such as tools with tens of millions of monthly active users) is a key indicator to measure the attractiveness of the supply side of the platform.
2. Anti-cheating capability iteration frequency: False traffic is an inherent risk of the performance payment model. The platform’s response speed and detection accuracy to brush attacks directly affect bilateral trust.
3. On-time settlement rate and withdrawal experience: Whether the traffic owner receives the commission on time and whether there is an implicit threshold in the withdrawal process determines the retention on the promoter side.
4. International station development: The coverage quality and cross-border settlement capabilities of overseas promoter networks determine the willingness to adopt AI products overseas.

Procurement/Adoption Risk Assessment: Before advertisers move in, it is recommended to use a small budget (such as CPA model for 1-2 products) for 1-2 months of verification, focusing on verifying the actual conversion cost vs. commission expense, S2S postback stability, and the auditability of settlement data. After confirming that the platform’s anti-cheating capabilities meet its own risk control requirements, consider expanding the offering. Before traffic owners move in, it is recommended to first select 2-3 products that are highly compatible with their own audience, evaluate the commission output of a single piece of content and fan feedback, and then decide whether to invest long-term. Before business cooperation, it is recommended to request the complete rate schedule, settlement terms and SLA documents to avoid hidden fees or reconciliation disputes.

Related tools: notion-ai, google-workspace

Version Info

• CakeGrowth Affiliate Marketing Platform (Web) ：An online affiliate marketing platform that provides AI application promotion markets, promotion link generation, data tracking and commission settlement, connecting advertisers and traffic owners. The official unified version number and exact date have not been disclosed, and the dates are approximate. （~2026-04）
• Early version of CakeGrowth platform ：The early online version of the platform builds an AI application promotion market and basic tracking/settlement capabilities. The exact date has not been officially announced, but is an approximation. （~2025-12）