Zion (No-Code Development Platform)
Zion, a drag-and-drop app builder, handles interaction design, business logic, and data models all through visualization for users, improving development efficiency by 10 times with everything automated from software design to delivery.
Jiheng Lu, a Cornell graduate in computer science, spent time at Facebook and Groupon as a software engineer/tech lead to “stand on the shoulders of giants” to understand how the best tech companies operate before answering the call of entrepreneurship.
A million dollar question, if anyone else also wondered when failing to debug code after hours of head scratching. No throwback needed however, Sequoia seems to agree.
Founder
Q: As a technical founder, how did you develop well-rounded skills as you transitioned from being a software engineer to an entrepreneur?
A: By reading as well as trial and error. I am in charge of commercialization, sales, and growth, so I try to ask questions from people who are the best in the field and bring them to share with the team whenever possible. I find it the most efficient way to grow in a new space. Mentorship from experts provides not only insight but also an introduction to quality learning materials.
Q: You used to be a software engineer/tech lead at Facebook and Groupon. Have these experiences inspired or helped your entrepreneurial pursuit? If given a chance to choose again, would you consider starting your company immediately upon graduation instead of spending some time at tech companies first?
A: I would not. Having 1 to 3 years of work experience is helpful for building a team and finding direction in the early stage. Obviously, some people do an amazing job without any work experience, but exposure to a very successful company widens your perspectives quickly. I only spent a year and a half at Facebook but it was super helpful. I knew that ultimately I’d start my own company but I went there to see how the best company in the world operates. The company was over 10 years old at the time so effectively I was standing on the shoulders of a giant, absorbing its experience accumulated from all those years as quickly as possible. It has been instrumental for my efforts in building team and culture. Top-tier tech companies like FAAMG (Facebook, Amazon, Apple, Microsoft, Google) have very strong cultures that you’ll only get a flavor of by being part of it as an insider. Having some experience at a great company, be it a startup, a unicorn, or a big company, sets the stage for you to take off with more aggressive goals.
Q: Is finding product market fit at a startup quite different from developing new features at a big tech company?
A: Yes. At a big company, when you push new development, your degree of freedom is relatively constrained, even if you are at a more bottom-up company like Facebook, because your jurisdiction falls within a defined niche of a big machine and is confined by direction from above. At a startup, however, your degree of freedom is enormous because there are so many things that you can experiment with. One of the biggest differences from being at a big company is that it is particularly important to know what you need to get done, where you are going, and what your focus should be. Meanwhile, due to the wide range of options, iteration needs to happen faster.
Q: What led you to pick the direction of no-code development for your startup?
A: Three reasons primarily — my vision, technological feasibility, and market demand.
First and foremost, it is something that I personally want to accomplish. My very first exposure to coding in 2000 was through a tool from Microsoft called FrontPage. It was like Dreamweaver of the early 2000s, like a no-code development tool with which you could create and publish a website visually using the drag and drop method. That introduction was instrumental in piquing my interest in programming and the Internet. Due to tech stack changes and various other reasons, Microsoft discontinued FrontPage and Dreamweaver declined over time, but the mission of visualizing coding to lower the bar for more people who do not have a technical background to get involved has sustained value and demand. We find it a meaningful endeavor.
Second, market research led us to conclude that it was good timing to get started from the tech perspective, as the increasing maturity of cloud services and cloud-native computing has enabled the automatic deployment of modern applications’ backend. As a result, the frontend and the backend get to be built in unity on a single platform through visualization, similar to FrontPage which handled interface, logic, and data all in one place with HTML, CSS, and JavaScript.
Third, the slowdown of growth in the to-C space has redirected demand from digital operations to more refined software. Nonetheless, the supply of software developers has not fundamentally increased, resulting in a gap between expectation and reality. We think this pain point can be resolved by a tool that can empower people without coding skills to build something with minimal training, with an estimated 25 times higher software delivery rate.
Product & Company
Q: How is no-code development different from low-code development?
A: By definition, low-code still involves coding, despite providing a development framework with a higher level of encapsulation. No-code adds another critical layer for visualization. Low-code development platforms are like fill-in-the-blank questions for users, which have room for a wide range of variations in answers that could go wrong in different directions, raising the cost of learning. The biggest value of visualization is that it concentrates the fill-in-the-blank into a multiple-choice question that allows users to pick it up without much training cost. This model had been validated on computers in the late 1990s and early 2000s when the earliest operating systems were more esoteric ones like MS-DOS and Linux that you can think of as a fill-in-the-blank type of interface. What popularized personal computers was the graphics-based operating system interface, which encapsulated many concepts in an operating system via visualization and hence reduced the cost of knowledge for more to participate. We hope to accomplish the same goal in software development.
Q: Do users need specific training to get familiarized with using your tool?
A: To some degree, yes. Our product is essentially a tool that can be thought of as a more efficient pen but users need to come up with ways to sketch the design of the software, which can be challenging for many given the average digital capabilities prevailing. Our training focuses on helping users sort out their specific needs for developing the software, which in turn facilitates using the tool.
Q: How much time and cost does it take to develop a program using Zion? How much are users saving, compared with the traditional development approach?
A: Developing a mini program with both the frontend and the backend with the traditional development approach takes approximately 20 man-days versus 2 man-days with Zion for someone with little coding experience, increasing delivery efficiency by 10 times. Bigger projects can realize an efficiency improvement of 7 to 8 times.
Q: What features enable Zion’s product to improve software development efficiency by 10 times for users?
A: By reducing labor cost, for the most part. Most of the time developing interactive applications is spent on more repetitive and manual procedures such as translating design and business logic into code, setting up the environment and automatic deployment process, as well as connecting the flow between frontend and backend. Zion removes miscellaneous technical details like deploying cloud services, publishing Mini Programs onto WeChat’s platform, and integrating code fixes at the frontend to backend APIs so that users can focus on core aspects of the software, such as business logic, interaction design, and data models. We have also abstracted the more complicated and technical concepts for users to pick up more easily.
Q: Why did you pick WeChat’s Mini Programs as the first platform to support users with?
A: A few considerations here.
After researching and talking to software outsourcing companies, we found that the vast majority of new projects in the Chinese market would choose to use WeChat’s Mini Programs because it’s better supported and more integrated than a website in the WeChat ecosystem and cheaper than an app that oftentimes charges high promotion fees.
Get Techcelerated’s stories in your inbox
Join Medium for free to get updates from this writer.
In addition, it reduces technological complexity for us as well. Zion’s system involves 3 layers, an interactive development environment, a self-developed compiler or framework, and automatic deployment. Given our relatively long tech chain and limited resources, we need to focus on a platform with the minimal cost and least complexity for development. WeChat added a level of encapsulation on top of apps for Mini Programs that handles many issues like multi-platform support and compatibility for us, making it easier to build an MVP.
Q: Speaking of visualization, is that why Zion’s interface looks more like a design tool?
A: Correct. Our interface looks more like those of design or prototype tools such as Figma and Framer. Because they have already been widely accepted by designers, product managers, and frontend developers, who are our target customers, we want to create a sense of familiarity for them and also reduce complexity through visualization. These prototype tools have very mature interaction design toolkits but are lacking connection to the backend data, which, if added, will constitute a complete application. Framer, for example, is taking a step ahead by launching a new website product called Sights, moving beyond pure prototyping. We are marching to the middle from the other side of the spectrum by benchmarking our interaction design against design tools that our users are familiar with.
Q: Some web development tools employ templates, which confine the degree of customization that can be achieved. How customized can a program developed using Zion be?
A: Our tool allows a degree of flexibility that almost approximates a program developed from the ground up, because of our atomic components. For example, all of the mini programs are essentially comprised of pictures, texts, videos, buttons, dropdown lists, and other elements arranged in a certain format and logic. Development documentation of apps, websites, and mini programs always contains the components used and supported. Zion encapsulates these smallest components for users’ interaction design. Similarly, business logic is broken down into action units and data models into smaller types. We have redefined the smallest unit for all of these 3 aspects, thus enabling a level of flexibility close to manual coding.
Q: What business scenarios can we apply Zion’s tool to?
A: Our goal is to serve general scenarios, similar to how Xcode and Android Studio can be used for the development of all interactive apps in iOS and Android instead of just one in particular. We can cover almost any use case, currently including but not limited to eCommerce, community apps, spreadsheets, and smart device vendors.
Market
Q: What kind of customers are you serving, mainly individuals and SMEs?
A: Our target customers at this stage are mostly software outsourcing companies. We equip them to better serve their clients, although there are many large corporate clients that we directly support as well, such as our investor Sequoia, Oppo, and others.
Q: Many cloud service providers, software companies, developer tools companies, and business process management companies might also have incentives to develop no-code or low-code development products. What differentiates Zion’s?
A: BPM companies focus on the digitalization of internal processes within a company. Zion can serve a company’s external need of building interactive applications or systems that face customers. In this sense, our offerings are highly complementary for our clients.
It’s not uncommon for big tech companies heavy on cloud services to have no-code or low-code development products, such as Tencent WeDa and Microsoft Power Apps. Compared with them, our biggest advantage is cloud neutrality, similar to how Snowflake has carved out a special market for itself as a cloud-neutral data company on top of the big data tools hosted within AWS, GCP, and Azure. Some clients do not want to be bounded by a particular cloud product. Cloud neutrality accommodates special demands that big tech companies cannot satisfy because of business interdependence.
Q: Enterprise tech has been on the rise in China in recent years. What opportunities and challenges do you see in this market?
A: China’s digital infrastructure is still at an early stage, compared with the U.S. It brings challenges since many SaaS products depend on a grounded infrastructure to realize digitalization. For example, Salesforce relies heavily on the majority of its employers’ recognition of digitalization. The value of many of its systems will be discounted without being operated properly.
At the same time, this growth stage creates many opportunities, as most industries are still a vacuum with unsatisfied demand. It won’t be a replicate of the U.S. market however, since scenarios, corporate cultures, and business processes are significantly different, which leads to unique opportunities for Chinese SaaS companies.
Q: Why is China behind the U.S. by so many years in SaaS and the PLG (product-led growth) model?
A: PLG is an inevitable trend when SaaS reaches a certain stage. SLG (sales-led growth) depends on sales and MLG (marketing-led growth) depends on marketing. PLG completely relies on the quality of the product as well as the efficiency of to-B content marketing. Building great products takes a long time yet the SaaS market is still relatively young in China, hence not surprising to be a few years behind the U.S.
Second, the supply of to-B marketing and software development talent is still lagging in China because a lot of talent and resources flooded to to-C opportunities a few years ago. It takes more time and effort to pick up the to-B side.
Q: Sales might not be the biggest strength for some technical founders. If that was also the case with you, did you take steps to intentionally enhance your sales skills, or does Zion adopt a PLG approach where the company’s growth depends mostly on the product itself?
A: We would love to stick with PLG but there are still challenges to be addressed on the product side so we are relying on SLG at the early stage. PLG is a lagging metric whose benefits can be enjoyed only after the product gets molded into good shape. The technical nature of our product determines that it’s going to take a while before we can complete the features. We are still at a preliminary stage of PLG and rely on traditional SLG for market entry and business expansion.
As for me, coming from a background in engineering and product, I think my sales talent is ok but I don’t enjoy it that much, so I put together a more experienced team with whom I could explore the market together. Also as previously mentioned, I try to gain insights from the best in marketing and business through the introduction of investors and friends, which has helped me improve tremendously.
Q: Which is the PLG company that you appreciate the most?
A: I greatly appreciate design tools like Webflow and Figma as well as Zapier which has achieved organic growth in SEO (search engine optimization) products through content marketing. Content marketing enables establishing a competitive edge and an ecosystem over time that can’t be achieved by short-term input. I’m trying to replicate the same model here in China but still a way to go.
Q: Any great Chinese PLG companies?
A: Lanhu is one. Some API products have done well. PingCap and other community-driven, open-source infrastructure products are making good progress. A difference in China is that content marketing won’t be able to gain organic traffic from Google SEO. WeChat’s redirection to a product’s page can get interrupted and affect conversion. There are some challenges with Baidu’s SEO, too.
Q: What changes do you hope to bring through Zion to the no-code development industry or software development at large?
A: Our vision is to become a one-stop shop for no-code development of interactive apps as a product of the era like FrontPage and Dreamweaver. The biggest compliment to a product is that it will be remembered as people think of the era. We hope to have that kind of impact on no-code development.
Q: What other platforms will Zion move on to support, in addition to WeChat Mini Programs?
A: We will support all kinds of interactive platforms. We will be kicking off beta testing of web apps and mobile apps this quarter and rolling out the development of cloud-native iOS and Android platforms in the second half of this year. In the future, any human-computer interaction platform could be our target to support, including Tesla’s upcoming app store announced to be launched this or next year as well as AR/VR platforms.
Q: Will you consider incorporating development tools for web3 at some point?
A: Very interesting question. Yes, we are in the middle of exploring the web3 market. It’s still at the seed stage but we are already seeing plenty of opportunities. Many technical founding teams have encountered pain points in UI and interaction since their strengths are in smart contracts, algorithms, and so forth. It might be that they don’t care about frontend design as much but the tech stack will look a little different in web3. We hope that our product will help bring more people into the new space to build more UI-friendly full stack web3 applications. We are working on it.
