As you start building your product, either from scratch or maturing an existing product, how do you chose the right problems to focus on?
At the end of the day, your customer will only use your product if you are solving a problem for them. At any given point in time, there are a lot of prevelant problems.
Greek philosopher Plato said, “Our need will be the real creator,” which got molded over the years into, “Necessity is the mother of invention.”
In 1998, Nick Swinmurn was visiting through shoe stores in San Francisco, only to find nothing. It was quite difficult to find the right size, color, style, and fit to suit his needs. After spending a couple of hours in the mall, Swinmurn went back home — tired, frustrated, and empty-handed. Nick later discovered there were no online stores that dealt specifically in footwear, and that gave him an idea. Nick quit his job and started working on a website called shoesite.com, which later transformed to the Zappos we know now. In 2009, Zappos was acquired by Amazon for $1.2 billion.
Nick started with a problem he was facing, validated it, and then tried to provide a solution. He did a more-than-decent job, and now he is a billionaire. Of course, this is an oversimplification of the process, but the idea is to start with identifying the problem you are trying to solve within the customer groups you have chosen.
There are two instances where you are looking to find problems:
1. When you are building something from scratch.
2. When you are adding to an existing product.
In both scenarios, you will want to identify your target customer’s pain points by assessing:
a. What your customers told you
b. Analysis of the present information which is a combination of googling and the information you have at hand
c. In addition to what consumers are telling you, you should also keep yourself aware of what some of the trends in the market are.
d. Your own personal conviction
Building something from scratch
When you are building something from scratch, you don’t start with any data from which you can take insights. At this point, in addition to what the customers told you, you will typically be relying on online tools and market data. Here, you are looking to analyze the common problems your user faces from the data available online. You don’t need to be technical. Many of these tools have a free version, which allows you to use basic functionalities. Some of these tools can be used to discover user patterns:
· Brand24. This is a good tool with a free version that lets you track analytics across certain keywords. For instance, if you are building a product providing recipes, you can enter the keywords “food recipe,” and it will provide you with an easy-to-understand dashboard, with data like how often the keyword is used and whether people are talking about it in a positive or negative manner. It helps you determine if there are enough people talking about it as well.
· Google keyword trend analysis. This tool helps you find out how many times a particular keyword is searched based on location. It helps give you an understanding of which keywords are important and the response to them over time.
· Amazon search. In case you are trying to look into a B2C product, there are multiple tools available that let you search for specific keywords used within Amazon. It will help you see how your potential competitors’ products are doing and what the likely demand is for your products as well. The associated searches will help you look at the common keywords people are using and will give you more insights on the problems they might be facing.
Identifying problems for an existing product
In this scenario, you have already launched a product and are looking to add features. Here, you will need to analyze the information you already have. The following are some places you may need to look at data and finalize the problem areas your users might be facing.
Net promoter score (NPS)
Essentially, NPS is calculated based on how likely someone is to recommend your product to someone else. You extract this information through surveys.
Typically, it is a scale from 0 to 10 asking how likely a customer is to recommend your product.
To calculate the NPS score, you basically subtract the percentage of detractors from the percentage of promoters. For instance, if 60% were promoters, and 20% were detractors, then your NPS is 40.
The way you collect the data is by having customers fill out survey questions. You can try to calculate an NPS for specific features, your brand, or for a product. You will need to be creative while getting the customers to fill out the survey. If you have funding, life is much simpler. There are many tools out there — such as SurveyMonkey — that will help you create a questionnaire and send it to their expansive set of users. However, if you don’t have money for this, you can send a short survey to your friends or post the survey on social media profiles (LinkedIn, Facebook, etc.) and offer some basic incentives. A good incentive I’ve used is mentioning the results will be shared with all the participants.
Note: I am not a big fan of surveys with a new source of users. Online surveys have a questionable intent around them. Either conduct surveys of your previously established user group, or post on social media and tap into your personal sources. Some additional tools that will help you with tracking NPS are Ramen and Wootric.
It is helpful to track NPS over time, as it gives you a high-level snapshot of whether your features are considered acceptable or not. This will give you a view of the overall satisfaction of your brand/product. Low NPS gives you a general direction of where to look and what to improve. If your NPS is at a high level, you need to dig further to really know what’s going on.
Analyze the behavior of your users, not only what they are saying.
When Rayman Raving Rabbids, a popular video game, was released in 2006, everyone was going nuts about it. Ubisoft thought that they had made another success. They had great ratings in all the gaming communities, at least a four out of five. However, upon further analysis, they realized that even though a lot of players were buying and playing the game, most were only playing the dance mini-game, a small feature.
Based on this analysis, Ubisoft released Just Dance in 2013 to expand on the idea, and it’s been a success to date. At more than $1 billion in sales since then., it’s their second biggest hit in company history.
If someone within Ubisoft had not analyzed user behavior, they would have been super happy with the success of the original game, and potentially wouldn’t have found significant success through the inception of ‘Just Dance’.
There are several ways to analyze your user behavior. One way is identifying beta testers and asking them about their usage pattern. I am a big fan of creating a visual flow to see what the users are doing at a given step, and when they drop off. You can create a simple flow that captures the user behavior from a high level. For example:
Instead of block diagrams, you can use the actual wireframes of the product. It really is up to you. A good practice is to capture the number of users on each block. This gives you a high-level overview of what might be working, and what might not be.
This is a great chart for any product manager who is responsible for metrics. However, if you look at specific demographic data based on age, you will realize that Facebook is actually losing its younger demographic quite rapidly. The percentage of Facebook users of ages thirteen through seventeen dropped from 71% in 2015 to 51% in 2018.
It will be important to not just look at overall metrics, but really look into what the numbers are actually saying.
Last, depending on the nature of the product, it’s important to analyze what your customers are saying about your brand on social media. If you are able to listen to your customers and incorporate those changes, it will build your brand loyalty.
For instance, a random Twitter user commented about an experience he was having with Tesla:
f you work on a more popular brand, you can collect qualitative feedback from social media apps. To collect that feedback, there are various free online tools that you can plug this information into that will give you a high-level sentiment analysis. Repustate and Brand24 are interesting sentiment analysis tools I have come across that have a freemium model for you to try. Simply collect the feedback and upload it into one of these tools, and they will provide you a decent classification of positive, negative, or neutral. You can then look into all the negative feedback tools and obtain deeper insights into the pain points of your customers.
A sample diagram of what sentiments across social media platforms may look like.
Once you have a deeper understanding of the problems, it helps to map the pain points along with the customer personas you had initially identified. There is a framework you can use called Jobs to Be Done (JTBD). I think that it is quite comprehensive and provides an interesting perspective on the way you should look at customers. At its core, the JTBD framework takes the view that people use products or services to get a job done so that they can make progress in their lives. These people can be consumers or employees of an organization.
Note: The JTBD framework can be used both when you are building something from scratch and when you are adding something to an existing product.
For instance, McDonald’s wanted innovation in their milkshake line. They had mounds of data and a very sophisticated profile of their ideal customers. They were running focus groups and would then improve the product based on this feedback, but they weren’t seeing improvements in revenue or profit.
Professor Christensen, a renowned American economist known for his approach on disruptive innovation, took a different approach. He and his team wanted to understand what causes someone to buy a milkshake. At first, they observed customer behavior in stores, which gave them a view of what was happening. They found that half the milkshakes were sold before 8:30 a.m. It was the only item the customers bought, and they drove off afterward.
Then, they asked, “Why are you buying a milkshake?” This equates to, “What job are you trying to get done?” People would struggle to answer. Christensen’s team would then follow up with “Step back a minute. Think about the last time you were in the same situation (needed to get the job done). What did you order?”
The responses really started uncovering something interesting.
“Well, I tried bagels, but boy, they are dry and tasteless. I have to put cream cheese on and try to steer the car with my knees.”
“I ordered a banana to do the job last Friday. I finished it quickly.”
“I ordered a Snickers bar, but I didn’t quite feel right after that.”
“When I go to McDonald’s, the milkshake is so thick it takes me about twenty-three minutes to suck it up that thin little straw. I don’t care what the ingredients are. All I know is when 10 o’clock comes, I’m still full.”
The milkshake providing customers something simple to eat on a long and boring drive to work.
Essentially, the JTBD framework provides you with a unique perspective on your customers and the problems they are trying to solve.
As a product manager, your task is to collect all the data in the format that you are most comfortable with so you can create inferences from it. Even with the JTBD format, there are many “right” ways to structure your approach. They ensure that the problem statement — or the “job” — should be written in a specific format, the surveys should be collected in a specific format and that inferences should be driven within a specific strict format. I personally like the perspective of revolving the problem statement around the “job” the customer is trying to get done. I typically use a hybrid approach between reading what the customer is saying on the internet and applying those pain points to “jobs” that the customer is trying to solve. Again, the idea is to do what you are most comfortable with.
Essentially, you need to make a best educated guess at a relevant problem statement.
Act on your own personal convictions
Reggie Brown, while inebriated one night at Stanford, was talking to his friend about random things. When the topic of girls came up, he just nonchalantly expressed, “I wish I could send disappearing pictures.” That was the start of the multibillion-dollar company Snapchat. Not the amount of data they had, or the amount of experience (which was almost nonexistent), and definitely not the number of certifications they had in product management.
These personal convictions — ideas lingering in the back of your head — are powerful, and are almost always the ones that lead to success. However, in a world of data, information overload, and processes, we tend to overlook this important source.
When you think about what problems to solve, you must understand the importance of your personal conviction and add it into the process.
At this stage, you have spoken with the customers and gone through a lot of data on the internet, and you can now start to formulate ideas of interesting problems to look at. As you start identifying different problem statements, you can tag them as “backed by data” and “personal conviction.” We will look at prioritization of these proble statements in the next blog.
 Klement, “What Is Jobs to be Done?”
 Nobel, “Clay Christensen’s Milkshake Marketing”
 Bobic, “Where Is Snapchat’s Reggie Brown?”