User-Focused Product Development Model

There are thousands of technology companies working on building products, but only a few create products that delight their users. Generally, startup founders or product leaders start with a beautiful product vision, but what eventually gets released is not as pretty or delightful (see Figure 1). The main reason is that building products requires many people who think differently to work together, resulting in a compromise. 


Figure 1 - Vision vs. Reality (image courtesy of Christine Wang)


How do you build the culture and process that delivers products that delight users? The answer is simple: solve user problems and don’t compromise on the product vision. However, executing this simple idea is hard. Let's explore how we can make it a bit easier throughout the rest of this article.

The organization needs a clear mission and an understanding of how solving user problems will deliver on that mission. Customers don’t buy your products because they like your mission; they buy them because they solve a problem for them. Therefore, everything an organization builds should be framed in terms of solving user problems (see Figure 2).


Figure 2 - Focus on user problems


The five-step process to build an organization around user problems provides diverse teams with the singular focus necessary to create delightful products. I have developed this process after contributing to the creation of industry-defining products such as Public WiFi services, fitness bracelets, noise suppression algorithms, remote device management systems, shopping apps using image recognition, sensitivity analysis with Natural Language Processing, and cryptocurrency solutions. This broad experience has enabled me to create a product development model that is versatile enough to be applicable in almost any scenario. 

In my recent role as Chief Product Officer at GlacierGrid, I was responsible for ten teams (hardware engineering, data science, product management, product design, front-end engineering, mobile apps development, backend engineering, dev ops, firmware, and provisioning) working from nine locations worldwide. Without a singular focus, coordinating efforts among team members takes a long time, and people lose sight of why we are building the product.

The most important part of the process is to pick the right user problem to solve for the right user-set and solve it better than all available solutions. After selecting the problem and the user-set, the five steps (the numbers below match the numbers in Figure 2) to build a user-delighting product development machine are:

1. Technology: Select the right technologies to solve the user problem.

2. User Experience (UX): The selected technologies must create a delightful UX for the entire experience, not just the user interface.

3. Analytics: Analyze the UX quantitatively and qualitatively to understand how it can be improved. 

4. Iterations: Use analytics to identify opportunities for UX improvements.

5. New User Problems: Solving user problems and focusing continuously on UX will reveal new problems your product can solve.


This five-step cycle can continue indefinitely and at every step you have to synthesize what you learned with what you know. If you are a mission-driven company, the problems you choose must align with the mission. Solving these problems should move you closer to realizing the mission, not just selling the mission to customers.

Let’s take a hypothetical example. A tech startup, Change the World (CTW), has a mission of making a positive climate impact. There are many ways to achieve this. CTW must find a niche within the big sustainability/climate-tech ecosystem where it can solve user problems that current solutions do not adequately address. Suppose CTW identifies an unaddressed user problem it can solve: HVAC (Heating, Ventilation, and Air Conditioning) optimization to reduce electricity bills (HVAC can contribute up to 40% of the total bill). Successfully reducing and shifting energy consumption for many customers will help CTW realize its mission of positive climate impact by reducing CO2 emissions from energy generation. So, the user problem CTW is solving is reducing electricity bills. Much work is needed to figure out how to do that. Through testing and market validation, CTW decides to build a new thermostat to control HVAC systems, focusing initially on single-story office buildings with Roof Top Units (RTUs) where users recognize high electricity bills as a problem. Single-story office buildings is a broad market. You have to figure out which customer segment has electricity bill as the highest pain point and target that segment to go to market. Once the user problem and the target customer segment are identified, following are the five steps:


1. Technology

In this case, think of technologies in four dimensions:

(i) Architecture: The chosen technology architecture will inform suitable technologies. Consider cloud operations, customer location, device functionality, data collection frequency, storage, communication protocols, APIs, audit logs, Radio Frequency, etc.

(ii) Software: Decide on software development technologies. For example, React Native vs. native iOS/Android for app development, database types, container management, testing and provisioning tools, etc.

(iii) Hardware: Build a thermostat with sufficient memory for Over The Air (OTA) updates, microcontroller, schematic, sensors, controls, display, power source, how does the thermostat talk to other thermostats within the same building, remote sensors, etc.

(iv) Data Science (AI/ML) : Determine the types of Machine Learning (ML) and statistical models, simulators, data analysis tools, anomaly detection, etc. Can you find a way to change the cooling cycle of the HVAC to align with lower electricity bills? 

The technology is always evolving so you have to continually update technologies and invest in research and development (R&D) to create the best possible UX. 

2. User Experience (UX)

The chosen technologies enable the UX. The best way to save money on the electricity bill with HVAC is not to use it, but that wouldn’t provide a good UX. The challenge is balancing comfort and bill savings. Decide how much control to give building guests vs. administrators responsible for electricity bills. Address issues such as getting the correct temperature far from the thermostat sensor, detecting open front doors, and making many tactical decisions about mobile app and web dashboard interfaces, notifications, user journeys, energy savings calculations, and what should be automated vs. policy-driven or user-controlled. Educate users on the impact of thermostat setpoints on electricity bills and the climate. Determine the Minimum Viable Product (MVP).

UX should give you users' mindshare. If you get mindshare, marketshare will follow. 

3. Analytics 

Your MVP is in the market; customers are paying for it and using it. Now, you need to understand how close to reality your idea of the MVP is. Analyze every step of user behavior, both quantitatively and qualitatively. Use audit logs, track user journeys, conduct user interviews, and gather feedback from users and salespeople. This is a continuous process.  This step provides you data to make better decisions and to develop a product strategy, grounded in reality, to evolve the MVP to a commercial product that serves multiple market segments. 

You have to develop a system of feedback loops with all departments (marketing, sales, operations, etc.) so that this process of understanding the users becomes ingrained in the organization. 

4. Iterations 

With analysis, you know what needs improvement and in what priority. A common mistake is to blindly follow the customer and build the features they ask for. You have to understand the customer's business to know why they are asking for features and how those features fit into the product strategy. What customers want and what customers are willing to pay for can be very different things. Use judgment to decide what to build and in what order, considering why people ask for different things, integrations with different workflow management software systems, and long term impact on the product. Test ideas with prototypes before building. This is an ongoing iterative process. For example, you might need to inform users how long it will take for the HVAC to reach the desired setpoint. End-to-end testing is necessary for any new feature.

Get comfortable with the idea that you can not make all customers happy. Think of the product having a core that all customers need and unique features that vary by customer segments. Something to think about is how many resources go to strengthening the core and how many to building unique features (see Figure 3). 


Figure 3: Strengthen the core 



5. New User Problems 

Continuously refining user problems helps you understand user behavior and build trust with users. This process helps identify new user problems. For example, you might discover that customers are interested in knowing about the air quality in the building, leading to a new cycle of problem-solving.

You have to be consistently growing. Lack of growth leads to stagnation and that leads to decline. 




Connecting the Product to the Mission 


While building the product, you must figure out a way to measure progress towards your mission. Product usage should automatically contribute to the mission. The idea is to quantify and make this visible. For CTW mission, measure total kwh saved and shifted to times when the grid is not loaded and electricity is cheaper. Convert the kwh shifted to CO2 emissions saved and share this with the organization regularly.


It is a lot of fun to build products that delight users. 

Popular posts from this blog

Obituary: Charles T. Munger

Systems Thinking as taught by Ackoff

Marry the best who will have you and other wisdom from Munger and Buffett