Shulou(Shulou.com)06/02 Report--

What is AI's business model?

Http://blog.sina.com.cn/s/blog_cfa68e330102zelf.html

Author: Dan Nguyen-Huu, Source: Forbes

Compilation: Zhang Feiyi, known as T-guest

As the news rolls around, AI is indeed affecting every aspect of people's lives and work, but many people don't notice one thing: what should be the right business model for AI?

It is generally believed that the business model of AI startups is the same as that of cloud / SaaS.

Like AI, the original business model of cloud software startups confused customers and investors. For example, cloud computing pioneer Salesforce avoided the upfront payment of traditional software and the resources and costs required for local deployment. Now the Salesforce model has become one of the benchmarks for cloud / SaaS companies, and investors can even name common SaaS investment metrics, such as CAC and LTV, in their sleep.

But it is not easy to transplant this cloud business model to nascent AI startups. This is mainly because the foundations of cloud and AI technologies are different: AI is driven by data, large amounts of raw computing power, and algorithms that are difficult for the public to understand. For customers, it is actually much more complex than cloud computing startups, so the technology must also be sold in different ways.

Slower software deployment? More local integration? Sometimes that's what AI does. So how can we break the status quo and persuade customers to buy new AI technology? In the past two years, three emerging AI business models have come into play. They have different characteristics because they may be more effective than others in some scenarios.

AI business model 1: add-on

The pattern of this type of AI solution is very similar to SaaS's products, and the business models are almost interchangeable. These AI solutions will be seamlessly attached to other systems, such as CRM products or ERP systems. AI will access the data that flows through these systems and drive business improvements over time.

Many AI startups follow this model: Chorus AI and Gong are attached to Salesforce products and use AI to optimize customers' sales practices, while customer support software Solvy is attached to Zendesk or ServiceCloud services and can automatically reply to work orders. And Sift Science uses machine learning to reduce customer fraud, such as misuse of payment or false content.

Because this business model is similar to the SaaS model, it is easy for outsiders to evaluate it. Its strategy is to "insert" or "embed", starting with a value-added function and gradually maturing into a platform. These solutions are deployed quickly, just like cloud software, so the sales cycle is fast and ROI is easy to calculate. By quickly gaining a large number of customers, the AI solution will quickly build a data moat, so it will become smarter and faster.

But this speed and ease can also bring drawbacks. Like cloud software, these AI solutions are easy to steal and replace. If these solutions do not break their inherent image and pattern of beautiful, easy-to-use features, it is easy for users to deactivate or replace them when they experience budget cuts.

The business model itself may be the most familiar to us, but it doesn't necessarily make it the best.

AI business model 2: process enhancement

Although it works in the process, in the second type of AI business model, the new AI product does not change the existing workflow at all; it only improves the efficiency of the current workflow by integrating AI. However, these are deep manual integration, require a lot of implementation work, and of course the reward is a greatly improved process.

Such companies and service products include Ayasdi, IBM Watson and H2O.AI. These solutions span different industry verticals and help customers improve the operation of their core business. Taking IBM Watson as an example, it made its debut in the TV program Jeopardy and used natural language to answer questions, which attracted wide attention of the international community. It can analyze big data's model in real time and give its own insights, and it can even manage elevators in buildings by sending data back to computers through complex sensors. (although there is a lot of controversy around it. )

This AI business model is different from the current popular cloud model. Its disadvantages are obvious: intensive deployment and long sales cycle. At the same time, low turnover means that every transaction must be a big deal to ensure the sales of startups' products. But the advantages of this model are also considerable, and once implemented, these solutions are very sticky and can be well up-sold. Like the potential of AI, the ROI of this model may be unlimited.

AI business model 3: make machines independent

In the third AI business model, AI technology changes the entire workflow by introducing intelligence to help customers complete their business processes in a better way. This kind of AI "has" end-to-end experience, requires little human help, and the algorithm will run automatically.

Such companies include self-driving car services and drone company Kespry. The latter's drones can be used to collect data from construction, mining or insurance companies. After a storm, for example, Kespry drones can assess roof damage, so customers no longer have to install equipment on their roofs. Because the data is sent directly to the cloud and analyzed using AI computer vision, insurance companies can estimate the claim data almost immediately.

Because this model involves hardware maintenance, its advantages and disadvantages are different from the pure cloud model. In this model, hardware is the cost center and a commodity that startups must operate and store. AI software in drones (or vehicles) is a distinctive IP and a source of revenue for such companies. These AI startups sell software subscription packages to companies that rent hardware; these subscription packages can be expanded over time to get more work done.

In a word.

More viable AI business models may emerge in the future. The era of AI has come, and investors are still in a herd mentality. However, it will take time for them to find the "right" formula for AI's success.

What does this mean for AI startups? First of all, a business model that allows the enterprise to grow effectively is necessary, and then provide meaningful influence and value to customers and investors. If corporate business does not conform to familiar models, it may mean that they need to defend their models more firmly and persuade investors and customers more patiently, sometimes, entrepreneurs' intuition is more likely to find the right path for themselves than so-called "analysis".

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.