The writer:

Wang Wei, CEO of Drink Technology, an undergraduate and graduate student at Tsinghua University and a Ph.D. from Yale University, was responsible for product optimization in the Advertising Quality Department at Google Headquarters. Computer system optimization experts, published a number of top-level papers, with a number of international patents. Experience in big data technology and theory, with a particular focus on A/B testing techniques and practices.

As a practitioner and entrepreneur in the SaaS industry, talk about SaaS simply about your understanding. 

Let's start with SaaS. The (partial) product form of drinking technology is SaaS, software as a service. For customers, AppAdhoc A/B Testing is a software used for A/B testing. The software features test design, configuration, deployment, implementation, real-time analysis of test results and decisions.

That doesn't sound any different from what traditional software brings to customers. A traditional software is also used by customers to solve a problem, with a series of related functions. Excel, for example, is a form processing software that has a variety of table operation-related functions.

So what's the difference between SaaS and traditional software? In particular, what is new about software-as-a-service services?

In my opinion, the concept of service is an abstract concept developed by the Internet. Unlike online cultural words like "Zhao Zhitian", I think service is a very important and far-reaching abstract concept.

The business form of Internet enterprises has been innovating and developing, and it is difficult to abstract it with the traditional concept of "software products". From early communications to media, e-commerce, to entertainment games, and later social, video, immersive content and so on. These products seem to provide users with a more "full" experience, more "new" and more "Internet" than traditional software.

Where does this experience come from? I think what makes Internet products unique to "non-Internet" products is that there are two points:

1) The deep interaction between the product and the user is a core part of the Internet business.

Internet products can be directly collected to obtain the user's specific behavior and detailed data, according to these real-time user feedback, Internet enterprises can dynamically adjust and update the product, to provide users with a better experience. This particular work is covered by the concept of "Internet Product Operations". With operations, there is a fundamental difference between Internet products and traditional software products.

2) The Internet product itself update iteration is very fast, faster than the traditional software more than an order of magnitude, can quickly deliver the new version to the user's hands.

Unlike traditional software engineering with long lead times, Internet product updates can be real-time, agile, and on-demand.

If serious problems are found, thermal updates can be fixed in a timely manner. If there is a new requirement, the new feature can be developed and put online and pushed to the user as soon as possible. Coupled with the grayscaness release and A/B testing support for each product iteration update, it reduces the risk of quick launch and ensures that the new version always delivers the best (and better) user experience.

Product operations and product iterations have become the daily work of Internet companies. This day-to-day work clearly revolves around the needs and experiences of the user. Imagine the two ends of the Internet, one end is the user, the other end is the enterprise, the enterprise through the operation and iteration to continuously discover and meet the real-time needs of the user.

Businesses are always saying "What(else) can I do for you?" and users are always expressing "I need you to do this for me" by using the product, and the enterprise responds immediately. Such a scenario is indeed a business form of "service". At this level, I think service is a really great abstraction.

SaaS, software as a service, delivers products that are still software, but more importantly, software products are made into a service.By upgrading the secondary industry to a tertiary industry, we can give full play to the capabilities of professionals (customer success, product managers, operations managers, technology research and development) so that they can serve users more efficiently, bring more direct value to users, and ultimately bring higher revenue to SaaS enterprises.

I think it's important to emphasize that the unique business form and working methods of Internet services encourage SaaS business professionals to "innovate" as much as To C Internet people. As we all know, innovation is the only way to produce high-premium products and reap exponential growth.

In contrast to Google Drive and Microsoft Office, Google Drive determines which new features/removed from the next iteration will be based on the user's usage data, and then pushes the new version to a small group of users in a grayscause release, using the A/B test to determine whether the new version of the user experience is better and ultimately where Drive will go.

That's something you can't do with a new version of Office that's been released in more than two years. Office's project management team must have a vision to determine all the features users need over the next 2 years, and must anticipate how all the details in Office software should be designed to meet the user experience.

A bad feature, Office users may have to endure 2 years; This may be an important reason microsoft is investing deeply in Office 365 -- upgrading Office to a new era of online services.

Of course, in addition to the content of products and services, SaaS and traditional software are different in business models.

The SaaS model emphasizes flexible sourcing, ready-to-use, on-demand.This lightweight form is not new, it is very similar to subscription business. My personal view is that the difference between this business model is based on the service specificity of SaaS.

Because SaaS can be rolled out to users with minimal availability (MVP) and iterations are constantly updated, SaaS's research and development costs can be spread relatively evenly across the timeline.In this way, the process of charging from the customer can also be more evenly distributed over the timeline. While the revenue from customers is relatively small each month (traditional enterprise software is much more expensive), the cost of developing a new version each month is also small. This allows subscription installment fees to support business research and development.

On the other hand, SaaS enterprises continue to provide "services" to their customers, which are "new value" that occurs every month, so there is a reason to charge customers on an ongoing fee.

Therefore, the essence of SaaS is that it is an Internet service, and innovation in business models may be secondary.

It has been noted that, unlike To C's Internet products, To B's SaaS products are more complex and specialized, with many differences in research and development and delivery. That's true, but To B's SaaS and To C products are Internet services. I think at this level, SaaS is completely connected to the unique operational and iterative mechanisms of traditional software. And this level of traits, perhaps SaaS represents the future of the software industry trends of the most essential characteristics.

From IaaS to PaaS to SaaS, more and more "traditional" products are becoming Internet services. I believe this presents a bright and clear future for users.

Other articles by the author:

Shooters, farmers, billiards, to explore the data science in the Three Bodies

AB testing: Data-driven product optimization

Wang Wei: AB Testing - Growth Hacking Must Kill