Lumora officially launches: Monitoring brand visibility in the AI era and seeking financing

AI is Becoming the New Brand Front

When users start to rely on AI for brand recommendations, product comparisons, and service selections, the competitive landscape for brands is also changing. In the past, brands could determine their visibility to users through SEO rankings, social media buzz, 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 market precisely against this backdrop. It positions itself as an AI brand visibility and recommendation mindset monitoring platform for brand owners and marketing agencies. By simulating real - user questions, it detects the mention, understanding, recommendation, and citation of the brand in the responses of mainstream large models such as Doubao, DeepSeek, Qianwen, and Wenxin, and generates an AI cognitive audit report that can be used for diagnosis and action.

Currently, Lumora has been officially launched. Users can independently complete registration, purchase, initiate monitoring, view and export reports, and observe the changes in the brand's AI visibility and recommendation mindset through continuous retesting. 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 subsequent cooperation methods.

The funds from this round will mainly be used for early - stage business validation and product stabilization. Among them, 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 form a retestable delivery process and precipitate 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 also predicted that by 2026, the search volume of traditional search engines will decline by 25%. This kind of change does not mean the immediate disappearance of traditional search, but high - intention questions are migrating: the core in the SEO era was "users can find me", while in the AI Q&A era, it has further become "is AI willing to recommend me?"

The common measurement method in the current GEO industry is the "mention rate", that is, whether the 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 placed in the correct scenario, nor does it mean it will be recommended when users make real purchase decisions.

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

In the future, Lumora hopes to precipitate 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 - intention scenarios, and has the space for continuous retesting and optimization. The team also summarizes Lumora's long - term goal as "becoming the Google Analytics in the AI marketing era".

Establishing a Retestable 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, first - recommendation rate, recommendation share, recognition accuracy, 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, and prescription". The measurement layer answers whether the brand is visible to 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.

Secondly, there are label dimensions and analysis algorithms. Each question is labeled with different cognitive levels, trigger intensities, and visibility types, 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 - penetration, cognitive - entry gap, competitor - substitution gap, and trigger sensitivity, to determine whether the problem lies in insufficient content assets, unclear positioning expression, 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 giving a score. The founder believes that the concept and pages of GEO monitoring can be imitated, but to truly integrate question generation, label system, scoring logic, failure attribution, and content prescription requires continuous optimization and calibration with real samples. Especially in the stage of rapid change of AI models, Lumora's current one - person company form can more quickly incorporate new judgments into the product and report.

Entering the Market as a Third - Party Auditor to Serve Brands and Agencies

Lumora's current main business model is continuous monitoring: the enterprise version establishes a monitoring baseline for a fixed brand or business line and conducts continuous retesting; the agency version has no limit on monitoring targets, which is suitable for agencies to flexibly allocate monitoring times among multiple client projects.

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 product has been officially opened for registration and purchase, and the team is further validating business needs through the continuous use of real brands. According to the team's disclosure, two agencies have confirmed their cooperation intentions after seeing the test reports. Among them, a local education and training institution in Chengdu has reached a service - exchange pilot with Lumora. The institution exchanges 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 project was initiated by 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 startups 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 ability. For Lumora, the key is to transform complex, volatile, 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 competition strategy, Lumora emphasizes that it adheres to third - party monitoring and does not enter the content distribution and agency operation tracks. The founder believes that if a 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 precipitate an industry benchmark database for 5 categories. For a product that has been launched but is still in the early - stage business validation phase, what Lumora needs to verify is not only the viability of AI visibility monitoring but also whether brands are willing to pay continuously for "how AI recommends me".