GEO Case Study: A Baby Nutrition Brand's Top-3 Recommendation Rate Reaches 127 Times Its Baseline in 41 Days
A baby nutrition brand increased AI visibility from 8.48% to 53.56% and its top-3 recommendation rate from 0.31% to 39.42% in 41 days. The case examines Yishan Technology's content approach across Kimi, Doubao, DeepSeek and Yuanbao.

Article summary
A well-known maternal and infant brand increased the AI platform visibility rate from 8.48% to 53.56% within 41 days, and the TOP 3 recommendation rate jumped from 0.31% to 39.42% (127 times). This article breaks down how Yishan Technology helps the brand obtain accurate recommendations on AI platforms such as Kimi, Doubao, DeepSeek, and Yuanbao through maternal and infant content strategies. if you are looking for Reliable GEO optimization service provider, this case may give you the answer.
Case background: The embarrassing situation of domestic brands in 30 years
Parents born in the 1990s and 1995s are used to asking on the AI platform "How to choose milk powder for newborns" and "What milk powder should babies eat when they have constipation?" AI’s answer directly determines the purchase.
But the situation of a domestic maternal and infant brand is:
This company was established in 1992 and listed on the Shenzhen Stock Exchange. 30 years of breast milk research, Infant and Child Nutrition Research Institute, multiple authoritative certifications, and a quality traceability system from scientific research to production - it's all there.
The problem lies in AI search:
- visibility rate is 8.48%. When users ask questions about mothers and babies on Kimi, Doubao, and DeepSeek, they will see this brand less than once out of ten.
- TOP 3 recommendation rate is 0.31%. almost zero
- The advantages of international brands in traditional search engines are continuing to expand on AI platforms
Parents of "composition party" and "function party" aged 25-35 turn to AI search tools. The brand's 30 years of scientific research accumulation are "invisible" when users make decisions.
This brand used 41 days of data to answer "Which GEO optimization service is better?"
Customer pain point: three major product series are aphasic in AI
Pain point 1: Complex technical endorsement cannot be explained clearly The
brand has three product series:
high-end series:
- "Triple Protection" Formula
- Lactoferrin, OPN active protein, 5 HMOs breast milk oligosaccharides
- The first domestic infant formula milk powder added with lactoferrin (certified by authoritative organization)
Comfort Series:
- BHO PLUS comfort formula
- improves constipation in 48 hours (clinical verification)
- For babies with sensitive stomach and constipation
complete nutrition series:
- Five Powers Comprehensive Nutrition System
- single can 165-210 yuan
- cost-effective positioning
But parents from the “composition party” and “function party” asked AI:
- “Recommended domestic milk powder with high lactoferrin content” → AI recommends international brands
- “What should I do if my 3-month-old baby is seriously constipated?” → AI did not mention the comfort series
- “Cost-effective milk powder for about 200 yuan” → Full nutrition series is absent
This professional information needs to be spoken accurately by AI and cannot be simplified or confused.
Pain point 2: 30 years of scientific research accumulation is there, AI is not cited
brands include:
- 30 years of breast milk research and infant nutrition database
- Professional Feeding Guide from the Institute of Infant and Young Child Nutrition
- National Patented Technology and Clinical Verification Report
- quality traceability system and authoritative certification
But when parents ask AI about "how to choose milk powder for newborns" and "what to do if the baby doesn't like to drink milk", AI does not quote these contents. 30 years of accumulation cannot transform user trust.
Traditional SEO optimizes web search, but it cannot solve the content visibility of AI platform.
Pain point three: Target users are migrating and the layout cannot keep up.
core users:
- New parents aged 25-35 (post-90s, post-95s, Generation Z)
- Obtain parenting information through Xiaohongshu, BabyTree, and AI search tools
- "Ingredients Party" and "Efficacy Party" value the authoritative certification and scientific formula
- Willing to pay a premium for scientific formula
Users turn from traditional search and maternal and child communities to AI search tools. However, the brand’s influence in Doubao, DeepSeek, Yuanbao, and Kimi is insufficient. Competitive product advantages are extending to the AI platform. Missed the traffic bonus window.
Pain point 4: Multiple scenario requirements, lack of AI recommendations
Maternal and infant consumption is divided into scenarios, and each scenario corresponds to different problems:
- Newborn: "How to choose the first milk powder for a newborn?"
- Constipation troubles: "What should I do if my 3-month-old baby is seriously constipated?"
- Nutritional supplement: "What are the milk powders with high lactoferrin content?"
- Converting to breast milk: "What are the precautions for converting babies to milk powder?"
- Cost-effectiveness: "Recommended milk powder with high cost-effectiveness around 200 yuan"
How to make AI accurately recommend corresponding product series in these scenarios? Traditional marketing can’t solve it.
Yishan Technology’s solution: product serialization AI knowledge base
Yishan Technology has done three things:
1. Build a "product serialized AI knowledge base"
- Structured technical information
- Systematically organizes 30 years of breast milk research, patented technology, and clinical verification
- refines the technical features, professional ingredients, and clinical data of each series
- Establish the matching of "Issue-Product-Technical Endorsement":
- Constipation problem → Comfort series → BHO PLUS formula + 48 hours clinical verification
- Lactoferrin demand → High-end series → Triple protection + OPN + 5 kinds of HMOs
- cost-effective demand → complete nutrition series → Wuli nutrition system + price range of 165-210 yuan
- Scenario-based Q&A content
- Scenarios such as newborn feeding, constipation problems, nutritional supplements, milk transfer, etc. generated 400+ questions and answers
- Each question and answer contains authoritative data, applicable groups, and usage guidance.
- strengthens technical ingredients and clinical verification data for users with "ingredients" and "efficacy"
- content complies with the "Food Safety Law" and "Advertising Law" and avoids absolute terms
- Brand professional data anchor
- Organizes 30 years of scientific research history and background of the Institute of Infant Nutrition
- Extracts differentiated advantages: China’s first lactoferrin milk powder, clinical data on improving constipation within 48 hours
- Establish data labels for product formula characteristics, nutritional content, and authoritative certification
- allows AI to accurately quote the brand’s professional value when making recommendations
- Frequently Asked Questions Solution Library
- Sorts out high-frequency problems such as "baby doesn't like to drink milk", "switching to milk and diarrhea", "feeding with allergies"
- provides solutions and embeds product series information
- When letting AI answer parenting questions, it recommends brand-adapted products
2. Multi-platform AI training and optimization
for DeepSeek, Kimi, Doubao, Yuanbao and other platforms:
- platform adaptation
- Adjust content structure and expression according to content preferences of different platforms
- product technical data (lactoferrin, OPN, HMOs, etc.) can be accurately identified on all platforms
- Kimi Platform: Optimizing the Q&A content structure
- Doubao platform: strengthen scene-based content
- Yuanbao platform: Optimizing professional data presentation
- DeepSeek Platform: Continuous Iteration Strategy
- High-intent problem coverage
- targets high-conversion intent words such as "milk powder with high lactoferrin content" and "what milk powder for babies with constipation"
- generates high-quality answer content and deeply binds the brand to scene requirements
- covers "How to prepare milk powder", "Precautions for milk conversion" and other operational guidance issues
- data monitoring and iteration
- Monitor the visibility rate, TOP 1/3/5 proportion, and recommendation ranking of each AI platform every week
- Adjust strategy based on data
- Establish a closed loop of question-answer-effect
3. Professional team collaboration
Yishan Technology is equipped with:
- Maternal and infant content planning team (maternal and infant industry background, professionalism and compliance)
- Data Analysis Team (monitors AI platform performance and provides optimization suggestions)
- technical team (30+ mainstream AI platform inclusion standards)
- Legal review team (Compliance with Advertising Law and Food Safety Law) The core team of
has management experience in Fortune 500 companies and has managed a budget of billions. Understand AI technology and marketing conversion. It complies with the requirements of "unity of product and effect" of listed companies.
RaaS (pay-per-performance) model provides visual reports on answer position visibility and brand hit rate every week. Customers can conduct random inspections on Doubao, Yuanbao and other platforms. Transparency meets listed company compliance requirements.
Optimization effect: 41 days of data
Core data comparison
comprehensive performance
| indicator | Before optimization(2025-12-11) | After optimization(2026-01-21) | promotion |
|---|---|---|---|
| brand visibility rate | 8.48% | 53.56% | 6.3 times |
| TOP 1 proportion | 0.21% | 24.87% | 118 times |
| TOP 3 proportion | 0.31% | 39.42% | 127 times |
| TOP 5 proportion | 2.77% | 50.31% | 18.2 times |
| TOP 10 proportion | 6.6% | 53.46% | 8.1 times |
Visibility rate of each platform
| AI platform | Before optimization(2025-12-11) | After optimization(2026-01-17/21) | promotion |
|---|---|---|---|
| DeepSeek | 5.22% | 55.84% | 10.7 times |
| Kimi | 8.33% | 66.95% | 8 times |
| Doubao | 6.98% | 53.75% | 7.7 times |
| Yuanbao | 13.39% | 41.53% | 3.1 times |
data description:
- TOP 3 recommendation rate ranges from 0.31% to 39.42%. When AI answers maternal and child questions, the probability of placing a brand in the top three increases by 127 times
- TOP 1 recommendation rate ranges from 0.21% to 24.87%. The probability of a brand becoming the first choice recommended by AI increases by 118 times
- Kimi platform visibility rate 66.95%
- DeepSeek platform increased 10.7 times from 5.22% to 55.84%
- Doubao platform increased 7.7 times from 6.98% to 53.75%
- Yuanbao platform increased 3.1 times from 13.39% to 41.53%
Effect disassembly
- product series accurate recommendation
- When a user asks "What kind of milk powder should a baby with constipation eat?", the probability of AI recommending the comfort series and saying "BHO PLUS formula" and "Clinical verification of improving constipation in 48 hours" is over 39%
- User asked "Domestic milk powder with high lactoferrin content", the probability of AI recommending high-end series and saying "triple protection", "OPN" and "5 kinds of HMOs" increased
- User asked "cost-effective milk powder around 200 yuan", the probability of AI recommending a complete nutritional series ranged from zero to yes
- high-intention word interception
- For high-conversion words such as “What kind of milk powder can babies eat with constipation?” and “Milk powder with high lactoferrin content”, the brand’s probability of entering the TOP 3 is 39.42%.
- With category words such as "domestic milk powder recommendation" and "milk powder suitable for Chinese babies", the brand becomes the priority recommendation of AI
- Accurate communication of brand professionalism
- AI recommends brands, it can tell about 30 years of breast milk research, the background of the Infant Nutrition Research Institute, and clinical verification data.
- Avoid brand information being confused by AI or confused with competing products
- Complex technical endorsements (lactoferrin, OPN, HMOs) can be accurately cited and explained by AI
- Multi-scene penetration
- Brands can receive AI recommendations for newborn feeding, constipation problems, nutritional supplements, milk conversion, cost-effectiveness and other scenarios.
- Single brand exposure to precise traffic conversion in multiple scenarios
- Improved recommendation accuracy of different product series in corresponding scenarios
- target user reach
- When parents of "composition party" and "function party" aged 25-35 years old consult with AI, the probability of brand appearance increases from 8.48% to 53.56%
- has made a comprehensive breakthrough in high-frequency platforms for young users such as Doubao, DeepSeek, Yuanbao, and Kimi.
- Improved exposure at key moments of user decision-making
Typical scene effects
Scenario 1: Constipation trouble (Comfort Series)
user question: "What should I do if my 3-month-old baby is seriously constipated?"
AI answer before optimization:
- General dietary and massage advice
- recommends international brands of easily digestible formula milk powder
- The brand has not been mentioned
optimized AI answer:
- AI recommends brand comfort series to relieve constipation in dietary recommendations
- Says "BHO PLUS comfort formula" and "clinically proven to improve constipation within 48 hours"
- provides preparation methods and precautions
Scenario 2: Nutritional supplement (high-end series)
user question: “Recommended domestic milk powder with high lactoferrin content”
AI answer before optimization:
- Explain the role of lactoferrin
- mainly recommends international brands
- This brand is not mentioned or classified as "ordinary domestic milk powder"
optimized AI answer:
- brand high-end series is recommended as "the first domestic infant formula with added lactoferrin"
- AI said "triple protective power", "lactoferrin", "OPN active protein", "5 kinds of HMOs breast milk oligosaccharides"
- provides suitable monthly age and scientific feeding suggestions
Scenario 3: Cost-effectiveness (Full Nutrition Series)
user question: “Recommended milk powder with high cost performance of around 200 yuan”
AI answer before optimization:
- recommends mid-range products from other brands
- The brand’s complete nutritional series is absent
optimized AI answer:
- brand complete nutrition series is recommended as the "comprehensive nutrition quality-price ratio benchmark"
- AI said "Five Powers Comprehensive Nutrition System", "Single can 165-210 yuan"
- provides formula comparison with products in the same price range
Scene 4: Newborn milk powder
user question: "How to choose the first milk powder for newborns? Do you have any recommendations for domestic milk powder?"
AI answer before optimization:
- General principles for selecting newborn milk powder
- mainly recommends international brands
- The brand is occasionally mentioned but without technical details
optimized AI answer:
- brand is recommended as "a domestic brand with 30 years of breast milk research results"
- AI tells the background and scientific research strength of the Infant Nutrition Research Institute
- recommends different product series according to family budget (high-end/comfort/full nutrition)
- provides suggestions for newborn age adaptation
Scene 5: Special physique
user question: "My baby has a sensitive stomach and is prone to gas and constipation. What milk powder do you recommend?"
AI answer before optimization:
- Explain the causes of gastrointestinal sensitivity
- recommends international brands of hydrolyzed protein formulas
- The brand has not been mentioned
optimized AI answer:
- brand comfort series is included in the recommended list
- AI said "BHO PLUS comfortable formula", "moderate hydrolyzed protein", "OPO structural lipid"
- provides clinical verification data of "improving constipation within 48 hours"
- explains the applicable groups and conversion method Why can
Break through quickly?
The brand itself has advantages
- Professional content intensive
- Infant and young child nutrition research results, authoritative certification, scientific formula data
- AI platforms tend to cite these "authoritative information sources"
- Rich scene problems
- Maternal and infant feeding involves multiple dimensions such as age, physical condition, and nutritional needs.
- naturally generates a large number of long-tail problems, providing an entry point for content creation.
- User demand rigidity
- Scientific feeding is an urgent need, and users rely on AI to obtain professional advice.
- AI recommendations directly affect purchasing decisions
- competition window period
- The GEO layout of the maternal and infant industry is in the early stage, and the optimization of competing products is insufficient.
- has obvious first-mover advantage
Yishan Technology’s professional capabilities
- Understanding of AI search mechanism
- Understand the content preferences and inclusion rules of different AI platforms
- can optimize content structure in a targeted manner
- Maternal and Infant Industry Insights
- content team has a background in the maternal and infant industry
- Accurately grasp user decision-making paths and pain points
- content production capacity
- produces 400+ high-quality maternal and infant questions in 41 days
- The content is professional and complies with laws and regulations
- Data-driven iteration
- Complete effect monitoring system
- Quickly adjust according to data feedback
Customer cooperation
customer provided complete and accurate brand information:
- Infant and young child nutrition research results and scientific research data
- Authoritative certification documents and quality inspection reports
- product formula description and applicable groups
- Brand positioning and differentiation advantages
provides a foundation for content creation and avoids AI training errors caused by information bias.
Yishan Technology GEO Optimization Service
1. AI platform understanding ability
conducts in-depth research on the content recommendation mechanisms of mainstream AI platforms (ChatGPT, Kimi, Doubao, Deepseek, Yuanbao, etc.), understands the differences in content preferences between different platforms, and optimizes strategies accordingly.
2. content production system
- Professional industry content planning team
- Standardized content production process
- Quality audit and legal audit mechanism (compliance with the Advertising Law and Food Safety Law)
3. Data-driven effect guarantee
- multi-dimensional effect monitoring system
- weekly detailed data analysis report
- timely optimization and adjustment based on data feedback
4. Industry service experience
has provided GEO optimization services to many companies and has accumulated rich experience. It can quickly identify the brand's GEO optimization opportunities and formulate efficient plans.
Case inspiration
3 inspirations for the maternal and infant industry
- Scientific feeding content is the core asset of GEO optimization
- Users’ consultation on the AI platform is highly concentrated in the field of scientific feeding
- Establishing a systematic feeding knowledge map can increase the weight of the brand in AI recommendations
- Copyable actions: sort out feeding knowledge by month age, build a Q&A content library, and bind products with scientific feeding suggestions
- Scenario-based content is easier to obtain AI recommendations than brand promotion
- AI prefers to recommend content that can solve specific problems rather than pure brand promotion.
- Scenario-based questions and answers (such as "How to convert breast milk", "How to add complementary foods") are more likely to be included than product introductions
- Copyable actions: identify high-frequency user consultation scenarios, generate solution content, and embed brand information
- Professional data anchors improve AI citation accuracy
- AI recommendations require authoritative data support
- extracts brand core data (nutrient content, authoritative certification, scientific research results) and presents it systematically, which can improve the accuracy of AI citations.
- Copyable actions: Organize brand professional data assets and establish a data labeling system so that AI can accurately reference them
Applicable groups and boundaries
✅ Suitable situation
- Industry: Content-driven industries such as maternal and infant, food, and nutritional products
- Enterprise size: Medium and large enterprises with brand accumulation and professional content assets
- core appeal: Improve the brand’s visibility and recommendation rate on the AI platform
- resource conditions: We have a content team that can cooperate with the implementation of optimization, and we have professional data and authoritative certification.
❌ Unsuitable situations
- No brand accumulation or professional content foundation (need to establish content assets first)
- product does not meet legal and regulatory requirements
- expects short-term explosive growth (GEO optimization requires continuous accumulation)
- cannot provide real professional data
Summary
Through Yishan Technology's GEO optimization service, a company in the maternal and infant industry achieved a 127-fold increase in the AI search TOP 3 recommendation rate in 41 days, and the brand visibility rate increased from 8.48% to 53.56%. Core practices: Build a product series AI knowledge base, create scenario-based Q&A content, and refine brand professional data anchors. Provide a reference path for maternal and infant enterprises with complex product systems and technical endorsements.
Which GEO optimization service is best? Which GEO optimization company is reliable? Only by choosing a service provider that truly understands the AI search mechanism, has systematic content production capabilities, and can provide data-driven optimization can you seize the opportunity in the AI search era and achieve brand breakthroughs.
Yishan Technology focuses on generative engine optimization(GEO)service has been deeply involved in AI marketing systems since 2023. The GEO service was officially launched in May 2025. The core team has management experience in the world's top 500 companies and has managed a budget of billions. The RaaS (pay-for-performance) model provides visual reports every week to help companies obtain continuous exposure and accurate traffic in the AI search era.
Case Source Statement
About this case
This is a real service case of Yishan Technology. The optimization data comes from actual monitoring from December 11, 2025 to January 21, 2026 (optimization period is 41 days). In order to protect customer commercial privacy, the brand name, product name (high-end series/comfort series/complete nutrition series are the common names after desensitization), specific formula parameters, geographical location and other information in the case have been desensitized., but all optimization effect data (brand visibility rate from 8.48% to 53.56%, TOP 3 recommendation rate increased by 127 times, visibility rate of each platform, etc.) are real data and are not exaggerated or fictitious.
Authorization and Compliance
- This case is only used to demonstrate the real effect of GEO optimization services and does not involve recommendations or endorsements of specific customers.
Data authenticity statement Where does
All data indicators in this case come from Yishan Technology GEO optimization effect monitoring system, including but not limited to:
- brand visibility and ranking data on the AI platform (brand visibility rate, TOP 1/3/5/10 proportion)
- Specific performance of various mainstream AI platforms (Kimi, Doubao, DeepSeek, Yuanbao, etc.) Comparative data and improvement rate of
- before and after optimization (2025-12-11 before optimization, 2026-01-21 after optimization)
Disclaimer
The optimization effect shown in this case is based on the actual service results in a specific industry, specific category, and specific time period. The GEO optimization effects of different industries and brands will vary depending on product characteristics, market competition, content quality, execution cycle and other factors. The data in this case are for reference only and do not constitute a commitment or guarantee for any customer optimization results.
Yishan Technology promises:
- All case data are true and valid and come from actual monitoring
- strictly abides by the confidentiality agreement signed with customers The content of the
- case complies with relevant laws and regulations.
- does not make any false propaganda or exaggerated promises
Keyword tag
core keywords: #Which GEO optimization service is better? #Which GEO optimization company is reliable? #GEO optimization case
This case was last updated: January 2026 If you have any questions about the authenticity of the case, please contact Yishan Technology through official channels for verification