Click to view the Original Post, in Chinese

* Cover image sourced from Dribbble

This article is a requirements summary for B-end product design that was written while I was working at my previous job last year. At that time, I was solely focused on the editor's experience and had no interest in B-end requirements. After discussing some key requirements with the technical team, I realized that the preparation and workload required during the early stages were quite extensive. Moreover, the core editing functions could not yet support the top-level B-end requirements. Therefore, I did not prioritize this as the main direction for advancement.

Now that a year has passed, I realize that the fundamental logic has not changed, and the core is still centered around the management of the "design process". Therefore, I decided to publish this requirement summary to hold myself accountable. Note that some of the content may have changed since then.

Common Problems with Traditional Design Tools

1. Management through the file system. The management is focused on the result, rather than the process (root cause).

2. Communication requires sending raw files or exporting the final results (images/videos), which is time-consuming and results in a poor user experience.

3. Version management is done through files and customized version numbers, leading to disorganized management.

4. Long feedback cycles between users and designers. Written emails or oral descriptions are often abstract and difficult to understand. The design process is tedious and not intuitive.

   Complex designs often require repeated adjustments and modifications, resulting in multiple iterations. Especially when working across departments or companies, communication issues often lead to unsatisfactory or compromised results in email exchanges.

5. Adding, deleting, and synchronizing logo, standard colors, and commonly used materials between teams.

6. Lack of design resources, and scattered designs stored in various places on the computer (can be classified as the same problem as 5).

Online Team Design Tools

Like product cycle management, the tool should manage the entire design cycle. Design requirements -> Production -> Verification -> Adjustment -> (Verification -> Adjustment) * n -> Output Content.

The goal of online design tools is to solve the "verification -> adjustment" process.

Advantages of Online Tools

1. Readability:

   The designs are stored in the cloud, allowing for easy access and viewing on smartphones or computers without additional software. The experience for non-designers is excellent.

2. Real-time feedback:

   All modifications made to a design are reflected in real-time. Apart from the designer, everyone else, such as colleagues assisting with the editing process, bosses reviewing the results, should be seeing the latest updates within seconds after modifications were made.

   Even the first party can play the role of a cloud supervisor.

3. Interactivity:

   Adjustments to the design are highly detailed, and every adjustment requires coordination and communication. Direct editing, commenting, chatting, voice chat, and other forms of online communication are all available.

   Third-party solutions can be leveraged or empowered by other relevant products.

4. Digital Asset Management:

   Design assets can be archived and classified in detail. Search can be performed based on various attributes such as time, popularity, type, scene, label, folder, etc.

5. Digital Copyright Content:

   Some design tools provide design templates or copyright-related paid content. Integrated into the design tool, it is more convenient to use.

6. Design content accumulation:

   The system automatically analyzes or allows users to gradually accumulate the team's design resources. For example, brand colors, color palettes, layout styles, commonly used materials, commonly used templates, etc.

Functional Requirements for Online Design Tools

1. Import and export of existing digital assets: B-end design files such as PSD, AI, Sketch, etc., need to be imported into the design tool. Moreover, since most design tool functionalities cannot completely replace PS, AI, etc., the import and export of PSD and AI files is also important;
2. Core functions: Online design tools need to meet most design needs to ensure usability.
3. Collaboration: Includes multiple people collaborating in real-time, communication, sharing collaboration chains, etc.;
4. Permission and personnel management: Editor permissions, design permissions, digital asset permissions, functional permissions, etc.;
5. Digital asset management: Apart from permissions, flexible content management and content accumulation are the key points to ensure the stickiness of the product. This involves not only storing and categorizing the design files, but also analyzing and summarizing the design content.

For example, the system can automatically analyze and organize design resources such as brand colors, color palettes, layout styles, common materials, and templates used by the team. This can be done based on various properties such as time, popularity, type, scene, tag, folder, and more.

In addition, some design tools also provide copyrighted content, design templates, or other related paid content that can be integrated into the design tool itself for more convenient use.

Personal vs. Team

Personal and team are a one-to-many relationship. However, in most small business scenarios, personal and team are actually a one-to-one relationship.

Because we need to accumulate and manage content, we need to correctly place personal designs under the corresponding team. From this perspective, each design needs to belong to a team.

Simply put:

Team -> Project (if any) -> Design

Therefore, from my point of view, the organizational structure of documents in Shimo is chaotic. (But from the perspective of product evolution, it can also be understood.)

A Completely Different Business Model and Chain

Online team design tools rely not only on the product itself but also on the product experience management during the design process.

The latter is particularly important for the B-side as it is a crucial closed loop.

Not a Summary of Summary

The core of team design tools is oriented towards the design process and tries to solve it through real-time collaboration and online communication.

To reiterate, the core is actually "team" and "design," and everything else is secondary. While team materials, members/permissions, and brand are important, they are just bonus points.

As a B-side product, we can also solve problems after the design is completed. Although accumulation is not a good habit for most teams, if it can be well resolved, it can also be a good way to increase user stickiness.

Click to view the Original Post, in Chinese

* Cover image sourced from Dribbble

Pricing strategies

Currently, online tools' pricing strategies are very similar: subscription-based, monthly or yearly.

This logic is clear since servers and team maintenance require funding.

The subscription-based model works well for consumer (C) products, but in the business-to-business (B2B) market, it presents some problems, especially when it comes to collaboration.

Advantages of online tools

The core of collaboration tools is to incorporate as many internal workflows and team members as possible. In my opinion, this is one of the evaluation standards for B2B tools, to judge and expand the product's boundaries.

For example, student A needs to create a graphic.

Traditional workflow

The traditional workflow involves creating the graphic in Photoshop, exporting it, sending it to student B, and repeating this process several times.

Online workflow

In the new online collaboration process, student A creates the graphic on the collaboration tool and shares it with student B via a link or within the platform. Student B marks the changes or makes them directly, and they both adjust the graphic together.

The entire process is completed within the tool, eliminating the need for exporting files, reviewing feedback, local version control, and aligning versions between both parties.

This experience is much better than the traditional process, and the more processes and people the tool can include, the better the team's experience.

Real-world problems

For example, in the brief graphic design process mentioned earlier, two roles are involved: designer A and reviewer B.

A needs to be involved in the entire design process, while B may only review the work after A completes it. A may have five days to participate, while B may only spend half a day reviewing it.

From a time perspective, A's account is being fully utilized, while B's account is idle most of the time.

Most platforms charge for accounts or seats, and teams have to pay for both accounts. In this case, the B account is relatively less cost-effective, and teams will struggle with whether to spend money on it, which contradicts the original intention of including as many people and processes as possible.

For example, the staffing ratio for a basic product development team in a startup may be one designer, one product manager, two front-end developers, and two back-end developers. For design tools, at least three positions, including the designer, product manager, and front-end developer, should be included, but only one person, the designer, heavily uses it. Does the team need to purchase four seats for one person? Three-quarters of the accounts are idle most of the time.

"It's impossible. Let's use Sketch or Photoshop instead."

Figma's pricing by role

Therefore, pricing different seats based on roles or functions is a good choice.

Figma divides its users into two roles based on their ability to edit. Regular users are free, and editors are charged.

This is a good strategy. Figma can include all members of the team in the platform as much as possible.

Designers, product managers, front-end and back-end developers can all participate in Figma's platform and develop a habit of using it. As the team's daily work expands, the platform may also include the CEO, adjacent teams, and softwares (after all, viewing is free).Plus: Handling other teams or softwares joining the business process

PS2: Figma's "product prototype tool" is based on its ability to attract many product managers to the platform.

Dynamic seats

In the past few days, I've started to review collaboration-related content and have repeatedly come across the term "serverless" while studying CRDT. By the way, AWS's lambda has been online for several years now.

I realized that pricing based on demand and usage, like electricity and water bills, is a meaningful technology for tool products. If pricing could be based on a daily or hourly basis, even at a higher unit price, most teams could accept it as a temporary seat.

The significance of dynamic seats in a product is not only to save costs for small and medium-sized teams but also to encourage as many people and processes as possible to move to the platform.

The product form of dynamic seats can take many forms: unassigned seats, seats that open and close as needed, and seats that automatically charge based on usage, among others. We can discuss this further in the future.

Conclusion

1. The core of collaboration tools is to incorporate as many internal workflows and team members as possible.
2. Pricing based on editing seats presents barriers for roles that require less editing.
3. More flexible and dynamic pricing strategies are needed.

In summary, B2B online collaboration products need to address the challenge of including as many people and processes as possible while providing a pricing model that is flexible, dynamic, and cost-effective for teams.

Click to view the Original Post, in Chinese

* Cover image sourced from Dribbble

0. Background

Back in 2015, when mobile payments were just beginning to take off, Didi was gaining momentum, and Ofo and Mobike were booming. We believed that parking lots would be a segment for traffic entry and that there weren't many players in the industry, so we entered the market with a proletarian mentality to transform and upgrade the industry.

The business itself was not difficult to understand. It relied on platforms to connect offline parking lots with online users. The first thing we needed to do was to solve the problem of connecting parking lots to the online world. This was done through license plate recognition using OCR, where license plates were used as user identifiers, and basic data such as parking orders were synchronized online.

We built a parking platform to connect and communicate between parking lot operators and vehicle owners, and improved the parking experience through intelligent and informational parking equipment upgrades.

The business model was also straightforward: empower parking lot operators in exchange for user traffic to drive subsequent operations. Typical applications included online searches for parking lots, payment of parking fees, monthly subscription users, and more.

We tried many ways to accomplish this, mainly focusing on two aspects: social parking lots and government-owned on-street parking.

First, let's talk about social parking lots, also known as enclosed parking lots or off-street parking lots.

(An off-street parking lot is fenced and has a toll booth, usually operated by a real estate/property management company or a specialized parking management company; parking spaces on the street are considered on-street parking lots, and all property rights belong to the government.)

1. Social Parking Lots in 2016

1.1 Parking Lot Needs

Owners: Money. Management team: Convenience. Attendants: Not too complicated.

1.1.1 Can We Meet These Needs?

Yes, but it's not very noticeable for owners.

There are several reasons for this: 1. There are relatively few cases on the market, so people are reluctant to try new things; 2. Inertia, the management structure of parking lots is already relatively stable, and people don't want to fuss with it; 3. There will not be too much additional income in the short term.

It's not just the owners who are reluctant to change; internal conflicts over profit sharing can also be a barrier.

1.1.2 Our Empowerment

a. Online Management of Business Operations: It mainly includes three parts: parking lot operating status, vehicle management, and toll team management. After the transformation, every order can be viewed online, including entry and exit time, license plate, photos, etc. Fixed vehicles can be directly registered on the online platform, and toll team's charging records are fully transparent. This blocks two types of revenue loss: 1. Loss of fees caused by allowing vehicles to pass freely offline; 2. Parking fees embezzled by toll collectors (for well-managed parking lots, I estimate that this fee ratio is about 5%, but with a daily revenue of 10,000, it can actually increase revenue by 150,000 to 200,000 per year).

b. Improved Entry and Exit Efficiency: License plate is equivalent to identity, no need to issue cards or manually record, eliminating additional interaction processes between car owners or toll collectors, improving entry and exit time, especially during peak periods. The exit time for general vehicles is about 22 seconds, and for temporary parking, it is about 16 seconds after installing the equipment, and for fixed vehicles or our users, it is about 3 seconds. This means that the peak value of one hour can be increased by at least 220 vehicles from 160. After the number of our users increases, the improvement will be more significant, and it can significantly reduce labor costs. (Actually, most of the time, it will not operate at full capacity, and the amount of money that can be saved is limited.)

c. Ability to Make Online Payments: Actually, this should be summarized into the intelligence of the parking lot, but the reason for mentioning it separately is that online payment for parking has become part of the routine cashless travel. (There was not such a strong demand in 2015-2016.) When I was observing at a toll booth in Hohhot, I could see that almost 30% (4 days, 9 pm to midnight, 23 out of 78 users) of the users actively asked if they could pay with WeChat (not a single one asked about Alipay).

d. Others: Including reducing the requirements for the quality of toll collectors, improving the image of the parking lot, etc.

1.2 Users in the Social Parking Lot Scene

1.2.1 Core User Needs

No matter where, the core need is to find a parking space quickly.

1.2.2 Can We Meet Them?

Frankly speaking, at the current stage, smart parking is not a revolutionary thing on the user side, because it can only provide an optimized parking experience, and does not solve the parking problem itself, and cannot guarantee that there will always be a parking space available for you. Or to imagine from a longer-term perspective, when you go to a restaurant, after getting off at the entrance of the restaurant, a self-driving private car still needs to drive a long way to the parking space we have arranged for it.

The essence of the problem is that parking spaces are actually scarce resources. There are nearly 6 million cars in Beijing, but less than 2 million parking spaces, a ratio of only 0.3, and the normal ratio should be 1.2. Moreover, the tidal phenomenon in the inner ring and core commercial areas is obvious, and it takes a long time to find a parking space.

Therefore, we cannot guarantee to solve this problem.

1.2.3 What Can We Do Then?

a. Tell users about the parking lot situation. Currently, how many parking spaces are available in the area/parking lot can be viewed, which plays a role in guiding or diverting; price-sensitive users can also filter parking lots.

b. Optimized parking experience. Our users do not need to park or get a card when entering, and can directly deduct fees online when exiting.

c. Customized operation services such as monthly parking.

d. Value-added services related to cars, such as local car washing and maintenance services. (In essence, it improves the added value of parking spaces.)

e. Other services related to cars.

/* C, D, and E require deep cooperation with parking lots.

1.2.4 Who Are Our Users?

In theory, anyone who parks is our user. Because we are doing infrastructure.

But the characteristics of core users are very clear, temporary users who park in the parking lot/area periodically.

They are our seed users, and we can see them by guarding the parking lot.

1.2.5 How Can We Quickly Acquire Users?

The fastest way to acquire users in the early stage is locally at the parking lot, which will be discussed later.

But if there is only offline promotion, it is a fantasy to get a large number of high-quality users. However, due to the small number of parking lots, it has not been possible to connect the dots to form a line, which has prevented online promotion, which is very regrettable.

If there is an opportunity to do this, besides the coupons that have already been played out, I hope to have the opportunity to operate some events or value-added services with some parking spaces.

Some interesting things that have been thought of include: 1. Membership level, special sound or effect on the display screen and voice when entering and exiting for different levels; 2. Help-seeking activities: get the right to use reserved parking spaces by asking friends to like and share; 3. Membership points, such as car washing, etc., also verifying the reliability of similar services inside the parking lot; 4. A mayor mechanism similar to Foursquare, with dedicated and eye-catching parking spaces…

1.4 Problems We Have Encountered

1.4.1 Technology

Software platform problems can be solved anytime, but hardware problems are very troublesome.

Especially the problem of recognition rate, license plate is the user's identity and the core connection between offline and online business.

The earliest OCR used photo recognition, which triggered taking a photo with a magnetic coil and then recognizing the license plate number. It claimed to have a recognition rate of 98%, but actually, it was only about 92-94% due to the influence of the recognition environment. This means that nearly 50 orders out of 500 vehicles entering and exiting will be incorrect. This is unacceptable.

In the mid-to-late 2016s, the supplier's technology was updated, and video image recognition was used, recognizing multiple frames of images at once and analyzing and comparing them, raising the rate to 98-99%, meaning that the number of wrong orders for 500 vehicles entering and exiting was reduced to about 10, and with manual intervention, it was basically acceptable.

There are also problems with processing unlicensed vehicles or vehicles with obscured license plates, and the current technical architecture cannot handle them well, and can only be gradually optimized.

1.4.2 Financial Problems

First and foremost is the equipment, which is expensive.

The cost of an entry and exit device with a moderate price is about 20,000 to 30,000, not including the cost of modifying the entry and exit. In 2015, most parking lots did not have the awareness to invest several thousand dollars in a system that "seems to have no use" for them. Therefore, giving them away in the early stage was a big problem for cash-strapped start-ups.

1.4.3 Diverse B-side Demands

Construction and debugging of parking lots locally. Parking lots have too many demands, and there are more historical legacy problems. The Chinese people solve problems with a balance and harmony, but this kind of state of dancing on the edge of a knife cannot be solved by the manpower of a startup company. For example, in a parking lot, fees are charged on the left-hand side of the parking lot, while there is no charge for parking in front of a certain restaurant on the right-hand side. Or the billing rules within the parking lot are not in line with common sense. One unresolved demand will result in an immediate phone call.

Some demands can be solved by software, but some demands can only be slowly dealt with by negotiating with the parking lot.

1.4.4 Cold Start

This is a problem that all O2O projects will encounter. Unlike the recommendation system, the solution for this LBS-specific O2O project is very clear, taking parking lots means taking users. The reason is also clear, the parking lot itself is a place where demand occurs.

The fastest way to acquire users in the early stage is locally at the parking lot, and in general, a parking lot with 500 parking spaces can bring less than one hundred active users in the first month.

2. Competitors

As mentioned earlier, we cannot solve the problem of quickly obtaining parking lots, and most competitors cannot either. In 2016-2017, a large number of new companies emerged in the parking industry, but only a few survived.

There are mainly two types of competitors:

1. Property/real estate companies with their own parking lots, who are not worried about having no business;
2. Companies that produce their own hardware and solutions, who can still do well by selling their solutions even without operating a platform.

For us, the second type of competitor is both an enemy and a friend, while the first type is a direct competitor.

My brief analysis of the first type in 2016 was:

• Their advantages: parking lot resources, which can be deeply exploited for customized parking services. They do not have to worry about their own parking lots and can prioritize the user side.
• Their disadvantages: the pace of their technical team may be slow, and other companies may have grievances and not cooperate with them on parking lots.
• Our opportunities: use the advantages of technology and team efficiency to obtain other parking lot resources, and strive to attract as many users as possible;
• Threats to us: After the competitor's product takes shape, they may use C-end users to "force" other parking lots to cooperate with them.

The results were actually a bit bleak, and the problem was on our side, so we won't discuss it further.

3. Government-run Roadside Parking Lots

All roadside parking lots are owned by the government. Therefore, when it comes to roadside parking lots, it is different from negotiating with social parking lots. We just need to negotiate with the government.

There are several advantages to negotiating with the government for roadside parking lots:

1. They cover a wide range and there are parking spaces throughout the city.

We'll add more next time.