TalkingData Cui Xiaobo: Exploring market gaps based on comprehensive data analysis of mobile games

199IT News: TalkingData, which focuses on statistical analysis of mobile applications, recently released its mobile game data analysis tool TalkingData Game Analytics. TalkingData CEO Cui Xiaobo said that at present, there is a gap in the market for all-round data analysis and mining based on mobile games, which is also an important reason for TalkingData to enter this field.

(Data source for the total amount of mobile game users paying in China from 2009 to 2014: iResearch)

Currently in the mobile Internet sector, mobile games have a clearer profit model than mobile applications. Cui Xiaobo believes that the current competition in the mobile game sector is already very fierce, and the need for refined operations is becoming more and more obvious. More and more mobile game developers need to conduct data-based operations.

Regarding the prediction of the trend of mobile games in 2013, Leo Cui believes that this field is gradually becoming a red ocean. The cost of acquiring mobile game users is constantly rising, the retention rate is decreasing, the monetization ability is gradually weakening, and more and more manufacturers are entering the market, resulting in fierce competition.

On the other hand, Leo Cui said that the operation mode of mobile games has also changed in 2012, and a large number of operation platforms have emerged between CPs and channels. Mobile game operations show a strong trend of web games. This is because there are many small and medium-sized developers in the mobile game field, and most CPs have weak operation capabilities, so they will choose to supplement them through joint operations. Operation will become the key to competition.

(Data source for the trend of new players’ “next-day retention” in the second half of 2012: TalkingData)

Cui Xiaobo emphasized that compared with some mobile Internet data service companies on the market, TalkingData not only provides tools, but also provides overall solutions. "Game developers need to not only understand their own problems during the operation process, but more importantly, how to solve the problems."

At the same time as launching the TalkingData Game Analytics tool, TalkingData also released the "White Paper on Mobile Game Operation Data Analysis Indicators", which defines data indicators for the mobile game industry, such as user acquisition, user activity, retention & churn, and game revenue.

(Data source for players’ participation in each game in the second half of 2012: TalkingData)

Cui Xiaobo also pointed out that in the era of mobile big data, the data departments of many game companies rely on manual operations to extract data, which takes a long time to extract data and wastes opportunities. However, mobile games need real-time data to guide operations, which requires strong real-time data query capabilities. The ability to process big data largely determines the success or failure of game operations.

(In 2012, the top 25 App Store best-selling apps gradually showed a trend of online gaming. Data source: TalkingData)

When talking about his views on capital, Leo Cui said that the overall investment environment in 2012 was not very good. Investment was just the icing on the cake, while the enterprise's own model verification and hematopoietic ability were the foundation of survival. TalkingData is more cautious about capital.

It is reported that TalkingData has completed its Series A financing at the end of 2012, and the specific financing details will be announced in the near future. (199IT ralf)

The following is the interview content:

Q: As a comprehensive mobile Internet data service company, why does TalkingData launch data analysis tools specifically for mobile games?

A: In terms of the current market situation, the profit model of mobile games is clearer and better than that of mobile applications. Moreover, the current competition in the mobile game field is already very fierce, and the demand for refined operations is becoming more and more obvious. The previous operation model of "making a game and releasing it in major application markets is considered to have completed the task" can no longer adapt to the fierce competition. It is necessary to adjust the product and market strategy more finely. These are real needs, and they have not yet been met.

Q: How does TalkingData Game Analytics compare with game manufacturers’ own data platforms?

Answer: Compared with mobile games, the data platforms of traditional PC games and web games are more mature. Most traditional game manufacturers have their own set of data analysis methods and data systems; but the standards are not unified.

The data statistics and analysis of mobile games are very different from those of traditional games. For example, the instability of the data collection environment. For example, there are many time problems on mobile terminals, such as the problem of clocks. For example, entering the game right after turning on the computer and not setting the clock will make the data collected by the mobile terminal very strange. Such problems do not exist in the field of traditional games. Therefore, this involves issues such as data cleaning and calibration. The standardization of mobile game data is very critical.

If we use the automotive industry as an analogy, TalkingData can probably be considered as a company that specializes in tires and low-level professional stuff, and is also an indispensable part of the entire industry.

TalkingData's value to mobile game developers includes basic data cleaning and other work; data standardization, unified measurement; and real-time data monitoring and acquisition. Data cleaning has been mentioned before, which is the basis of data analysis and mining; in terms of data standardization, the "Mobile Game Operation Data Indicator White Paper" we launched earlier is to standardize mobile game analysis indicators. Only with standardized data indicators can there be comparison and analysis; in terms of real-time, it is very helpful for game developers to respond to problems quickly. Many traditional PC game manufacturers rely on manual operations in data extraction, and the data extraction cycle is very long, which delays the opportunity. However, the operation rhythm of traditional PC games is not as fast as mobile games, so it is acceptable. Mobile games need real-time data query capabilities.

As far as the current platform and services are concerned, small and medium-sized manufacturers are most suitable, and a complete set of solutions can be used directly. Moreover, TalkingData provides not only analysis tools, but also methodologies and consulting services. There are also some large manufacturers who may do their own analysis tools and actually have no professional methods and ideas for mobile game operations, but often choose TalkingData's consulting services because we provide professional analysis and mining methods and data consulting services to help customers create value based on data. In fact, TalkingData also provides more in-depth data mining services, TalkingData Insight, and TalkingData provides complete data service solutions.

Q: What is TalkingData's business model? How do customers use TalkingData Game Analytics services? Small and medium-sized customers are more sensitive to operating costs. Are they willing to pay for data services?

A: First of all, let’s make it clear that TalkingData Game Analytics is a paid service. This is the result of repeated arguments. The data collection and analysis of mobile games actually generates costs based on the activity of game players. The more active players there are, the more data is generated, and the greater the pressure on the data platform. Therefore, our charging method is actually to charge service fees based on the number of active users.

The cost of the entry-level version is very low, which can cover the usage needs of small and medium-sized users during the testing period, and is sufficient for the testing stage. After the game is officially released, after promotion and operation, as the number of active users increases, we also provide higher versions. Compared with self-built platforms, our service prices are much cheaper. In addition to the hardware costs that gradually increase with the increase in the number of active users, the corresponding manpower and time costs, added together, are far higher than the price of using TalkingData services. Moreover, with the increase in game revenue and compared with the cost of self-built, TalkingData Game Analytics is more worthwhile.

Q: How does TalkingData view the competition between domestic and foreign data products?

A: A recent trend is "Everything as a service". Any software or application should be presented as a service. Providing tools without corresponding services or just using tools to look at numbers is not very meaningful. In the past, we added a lot of functions in product iteration, but now we are starting to do subtraction. We are currently streamlining indicators and reports, strengthening methodology and consulting services, with the goal of creating value for customers based on data analysis.

At present, there is still a lack of analytical tools and service products specifically for mobile games at home and abroad. We are very optimistic about this market and have a large share. Currently, many CPs on the Top Grossing list of many App Stores are our customers.

Q: When collecting game data, will you adopt a volume exchange model like Chartboots?

A: TalkingData provides enterprise-level services. We follow the B2B model. We are currently able to support ourselves, which is a healthy model. In addition, as a data service provider, we must be in a neutral position to be more credible and valuable to customers. Chartboots' model is not suitable for the domestic market. For advertising-related services, we only provide DSP Demand Side services.

Q: What will be TalkingData’s key overseas markets in the future?

A: Data services are a "slow" company, and brand globalization is the only way forward. TalkingData's future globalization will first enter developed countries, with North America as the first stop.

Q: What is the survival status of mobile games in 2013?

A: Many media and bigwigs are optimistic about the mobile gaming sector. It is true that mobile gaming will have great development in 2013. However, the mobile gaming sector is gradually becoming a red ocean. The user acquisition cost of mobile games is increasing, the retention rate is decreasing, and the monetization ability is gradually weakening. Because more and more manufacturers are entering this field, the competition is too fierce.

In addition, the operation mode of mobile games has also changed in 2012. A large number of operation platforms have emerged between CPs and channels. Currently, mobile game operations show a strong trend of web games. This is because there are many small and medium-sized developers in the mobile game field, and most CPs have weak operation capabilities, so they will choose to supplement them through joint operations. Operation will become the key to competition. TalkingData hopes to help users perform refined operations and gain better opportunities in the competition.

Q: Jifeng praised TalkingData’s big data processing capabilities. Can you explain it in more detail?

A: Currently, the application recommendations in Jifeng Market are actually implemented by a set of intelligent recommendation engines provided by TalkingData. This engine helps Jifeng Market to better utilize long-tail positions and fill content in line with user interests and preferences. Before using this engine, the click-through rate of the recommended position of related applications was about 3%, which increased to 8% after using our recommendation engine. This is only part of the TalkingData Insight data service.

TalkingData's current product line is already very comprehensive. Compared with competitors, it can fully obtain data from inside and outside the application, and organically link these data for further mining. Provide greater value to our customers. Moreover, these services are standardized. We can provide big data processing capabilities to channel partners through standardized APIs. Through data sharing, partners' data can continuously train our machine learning engine to provide more accurate prediction, analysis, and recommendation results. This is a positive cycle.

Those who follow us should have discovered that our recent reports mentioned analytical data such as player interests and behavioral analysis, which are the results of processing using our models.

Q: What does TalkingData think about capital intervention? Why does it reject the entry of many capitals?

A: The overall investment environment is not very good. Investment is the icing on the cake. Our own model verification and hematopoietic ability are the basis for survival. We are more cautious in treating capital.

Q: What is TalkingData’s current financing situation? What is its revenue situation?

Answer: We closed our Series A financing at the end of 2012.

TalkingData's products have been released in Enterprise versions since their inception. Currently, the largest source of TalkingData's revenue is the Enterprise version of TalkingData Analytics. With the full development of mobile Internet, many large companies also have mobile Internet products and also need data analysis and mining services. Currently, some large corporate customers have purchased our Enterprise version and data consulting services.