Lumora monitors brand visibility in the AI era and seeks financing

AI is Becoming the New Face of Brands

When users start relying on AI for brand recommendations, product comparisons, and service selections, the competitive landscape for brands is also changing. In the past, brands could determine whether they were noticed by users through SEO rankings, social media presence, public opinion data, and advertising conversions. However, in the context of AI search and large - model Q&A, users often directly receive the results aggregated and filtered by AI. For brand owners, the new question becomes: Does AI know this brand, does it understand it correctly, and will it recommend the brand when users have a purchase intention?

Lumora enters the scene in this context. It positions itself as an AI brand visibility monitoring platform for brand owners and marketing agencies. By simulating real - user questions, it detects the mention, understanding, recommendation, and citation of brands in the responses of mainstream large models such as Doubao, DeepSeek, Qianwen, and Wenxin, and generates an AI cognitive audit report for diagnosis and action.

Currently, Lumora is up and running. The product is in the paid testing phase with an application - based and invitation - based system. The team is validating the product's value and commercial demand through real - brand reports. The project plans to launch a Pre - Seed/Angel round of financing, aiming to raise 3 - 5 million yuan and offering approximately 10% of the shares. Specific terms can be further negotiated based on the investor's resources, transaction structure, and follow - up cooperation methods.

The funds from this round will mainly be used for early - stage commercial validation and product stabilization. Lumora plans to invest about 60% of the funds in customer acquisition and paid - brand validation, to obtain real - industry samples and verify the payment willingness of different customer groups. The remaining funds will be used for API calls, model concurrency, subsequent development, sample library construction, small - team expansion, introduction of technical advisors or partners, as well as investments in trademarks, compliance, and basic brand promotion. The founder hopes that this round of funds will help the product establish a reproducible delivery process and build a preliminary industry benchmark database.

According to the 57th "Statistical Report on the Development of the Internet in China" by CNNIC, as of December 2025, the number of Chinese users of generative artificial intelligence has reached 602 million, with a penetration rate of 42.8%. Gartner has also predicted that by 2026, the search volume of traditional search engines will decline by 25%. These changes do not mean the immediate disappearance of traditional searches, but high - intent questions are shifting: the core in the SEO era was "users find me", while in the era of AI Q&A, it has further become "will AI recommend me?"

The most common measurement method in the current GEO industry is the "mention rate", that is, whether a brand appears in AI responses. However, Lumora believes that the mention rate can only answer "has AI mentioned you", and cannot cover the process from being mentioned to being understood, recommended, and converted. Just because a brand is mentioned by AI does not mean it is in the right context, nor does it mean it will be recommended when users make real purchase decisions.

Therefore, Lumora does not simply view GEO as "SEO in the AI era", but breaks it down into the brand's qualification for recommendation in AI responses. Its core methodology includes a three - level AI cognition and a two - dimensional scoring system: the former determines whether AI "remembers you, trusts you, and is willing to recommend you", while the latter determines whether the brand is noticed in the entire market or recommended in the key scenarios of target customers.

In the future, Lumora hopes to develop the "Lumora Index" to measure the real GEO level of brands in AI recommendation scenarios. Compared with the single mention rate, this index pays more attention to whether the brand is correctly understood, recommended in high - intent scenarios, and has room for continuous retesting and optimization. The team also summarizes Lumora's long - term goal as "becoming the Google Analytics in the era of AI marketing".

Establishing a Reproducible Monitoring System from Question Generation to Diagnostic Algorithms

In the product process, Lumora generates more than 50 real - user questions around a brand, and multiple mainstream large models answer them concurrently. The system then analyzes the responses, counts the mention, ranking, sentiment, citation source, and failure scenarios of the brand and its competitors, and then outputs a report.

Different from dashboards that only show scores or rankings, Lumora tries to structure the report into three layers: "measurement, diagnosis, prescription". The measurement layer answers whether the brand is noticed by AI; the diagnosis layer analyzes why competitors are recommended, in which scenarios the brand is absent, and whether there are cognitive biases in AI; the prescription layer provides content briefs, suggesting what content assets the brand should supplement, which information sources should be corrected, and how to conduct the next round of retesting.

Lumora's technical capabilities are first reflected in question generation. AI brand visibility monitoring is not about asking dozens of random questions. Whether the questions are close to real - user scenarios and can cover cognition, comparison, decision - making, and risks directly determines the value of the report. Lumora organizes questions according to different cognitive entry points and user intentions, distinguishing whether users are generally learning about the brand, comparing multiple brands, or making pre - purchase decisions.

The second aspect is the label dimensions and analysis algorithms. Each question is labeled with different cognitive levels, trigger intensities, and visibility types, used to determine whether the brand is "passively mentioned" or can spontaneously enter AI recommendation results when users do not explicitly name it. The system also breaks down indicators such as decision - making penetration, cognitive entry gap, competitor substitution gap, and trigger sensitivity to determine whether the problem lies in insufficient content assets, unclear positioning, competitor occupation, or lack of high - quality information sources that can be cited by AI.

In a desensitized test report, Lumora breaks down the reasons for low scores into dimensions such as content gap, positioning deviation, competitor relationship, and risk resistance, avoiding simply providing a single score. The founder believes that while the concept and pages of GEO monitoring can be imitated, truly integrating question generation, label systems, scoring logic, failure attribution, and content prescriptions requires continuous optimization and calibration with real samples. Especially in the stage of rapid changes in AI models, Lumora's current single - person company form can more quickly incorporate new judgments into products and reports.

Entering the Market through Third - Party Audits to Serve Brands and Agencies

Lumora's current business model starts with AI visibility audit reports. Its pricing strategy is divided into three categories: single - time audits for direct corporate customers, periodic retesting for continuous optimization needs, and multi - brand monitoring quotas for marketing agencies and proxy companies. The idea is to first lower the decision - making threshold through single - time reports and then establish long - term subscription relationships through retesting.

In terms of target customers, Lumora mainly targets brand marketing departments, public relations and public opinion teams, digital marketing teams, brand consulting companies, and SEO/GEO agencies. For brand owners, it provides a new external perspective; for agencies, Lumora can serve as a third - party measurement tool before and after delivering GEO strategies to clients.

Currently, Lumora has tested more than 10 real brands, including new energy vehicles, liquor, consumer electronics, and local services. The team is opening paid testing through an application - based and invitation - based system. According to the team, two agencies have confirmed their cooperation intentions after viewing the test reports. One local education and training institution in Chengdu has reached a service - exchange pilot with Lumora, exchanging course resources for AI brand visibility monitoring services to verify the visibility and content optimization needs of local service brands. Another cross - border e - commerce agency is promoting paid cooperation.

The Lumora team was initiated by the independent founder, Yinhe. Yinhe has a background in UI/UX design, having worked in Tencent's user experience design team and later being involved in digital transformation projects of traditional enterprises and industrial artificial intelligence startup projects for about 14 years in total. His advantage lies not only in visual design but also in the information architecture of complex B - end systems, business process decomposition, and user understanding. For Lumora, the key is to transform complex, fluctuating, and noisy AI responses into a report structure that brand owners can understand, judge, and execute.

In the early stage of the project, a single - founder + AI collaborative development model is adopted, and content strategy and conversion consultants are introduced to supplement experience in content ecosystem, commercialization, and customer conversion. In the future, Lumora plans to introduce technical advisors or partners to expand the product scale, enhance system stability, and support the construction of multi - brand and multi - industry sample libraries.

In terms of competitive strategy, Lumora emphasizes that it adheres to third - party monitoring and does not enter the content distribution and agency operation fields. The founder believes that if the monitoring tool also undertakes the KPIs of content production and distribution, the report conclusions are likely to serve the delivery needs rather than objective diagnosis. Lumora focuses on improving the accuracy of monitoring, the reliability of suggestions, and the credibility of retesting, and continuously calibrates the system with real samples.

Next, Lumora plans to complete the validation of about 50 paid brands within 12 months and gradually build industry benchmark libraries for 5 categories. For a product that is already launched but still in the early - stage commercial validation phase, Lumora needs to verify not only the viability of AI visibility monitoring but also whether brands are willing to pay continuously for "how AI recommends me".