How to identify the key problems to solve while building products- Part 1
The art of problem discovery in product management
This blogpost is part 1 in a 2-part series on problem discovery in product management.
- In part 1, we will try to work out a structure for discovering problems. For this, I will share my journey of discovering a key problem in new user activation at Grofers (now Blinkit)
- In part 2, we will leverage the structured approach and figure out key problems to solve in a Meesho (Lowest price e-commerce marketplace) case study
In 2019, at Grofers (now Blinkit), I served as a product manager focused on new user activation. After our OKR planning, we took the goal to increase the following 2 metrics massively (for the scale we were at):
- App Install to First Order Conversion
- First Order to Second Order Retention
Now, these are just metrics that we are taking goals on. As a budding product manager, I had been used to solving user problems that the seniors had identified. And starting from just a goal metric now was just baffling. Which problem do I need to solve now?
- Should I do competitive study to understand what other players are doing?
- Should I read secondary research articles on the needs of users?
- Should I do correlation analysis to identify opportunities from data?
Where do I begin?
Such is the field of product management that we need to deal with ambiguity almost on a daily basis. This blog post aims to address two prevalent challenges encountered by product managers during problem discovery:
- There is a lot of ambiguity in identifying the problems users are facing with our product; is there a structured way of approaching this?
- There could be too many problems (big and small) ; how do we decide which ones to solve?
As we walk through a couple of case studies from Grofers (now Blinkit) and Meesho, here are the key takeaways for PMs from this 2-part problem discovery series:
- Steps/ Tools that can help reduce the ambiguity to some extent
- How to gather evidences of problems to build conviction on key problems to solve
Continuing with new user activation problem at Grofers (now Blinkit)
Grofers (now Blinkit) around 2019 was positioned as a lowest price monthly grocery shopping platform; we used to benchmark ourselves with online and offline competitors on the total cost of the basket of the top selling 150 products (key value items) and ensured that we kept the lowest cost on this group in totality given that people buy these as a major part of their monthly grocery.
A practice that consistently aids me in navigating through the daily ambiguity is engaging with users directly to comprehend the challenges they encounter on the platform. Here again, in confusion, I went to users and spoke to
- about 20 users who were dropping before first order and
- about 20 users who were dropping after first orders and before second order in the next 30 days.
- I also spoke to users who came back to place their second order
Step 1: When in doubt, turn to users
Here are the learnings:
- A significant segment of users who did not place either first or second order felt the we did not have the lowest prices (this was a shocking revelation given our hardcore benchmarking)
- Other reasons were- my products are not available, difficult to use the app
- Interestingly, people who came back to order believed that Grofers had the lowest prices
Users perceive high prices, struggle to find desired products, and encounter usability challenges. Now we know the problems users are facing, what’s next?
We first tried to understand from user surveys on the extent of these problems:
- User feel grofers have high prices (70%+)
- People don’t find products they want (~20%)
- Usability challenges (<10%)
Our major problem was clear to us (and also confusing)
Ambiguity: Now what should we do?
- Create stronger messaging around lowest prices across the journey
- Improve the lowest price perception by showing slashed MRP/ Market price and then showing Grofers price
WRONG!!!
This is a common mistake most product managers make.
“Don’t rush to solutions without understanding the ‘why’ behind the problems.”
Step 2: Understand the ‘why’ through analysis, research, and hypotheses building
We built quite a few hypotheses around the problem. Here are a couple of them:
- Is this happening in certain locations?
- Are users comparing us with a different platform than we benchmark prices with? If so, where?
Step 3: Validating the hypothesis to strengthen the insight
We looked at internal data, and spoke to 30+ users to test our hypotheses. Here are the findings around the 2 hypotheses
- Is this happening in certain locations? : Invalidated.
The conversion numbers were similar across locations and the same came out in user interviews that people across location has perception problems on Grofers being the lowest price platform - Are users comparing us with a different platform than we benchmark prices with? If so, where?: Validated
Kirana benchmarking users have higher conversion on Grofers, but for the rest- supermarket and online shoppers- conversion was significantly lower and these users contributed to a meaningful chunk of active users on the platform.
We repeated step 2 to go into the why of these hypotheses and spoke to users in person to understand when they feel prices to be higher (20+ interviews).
Insight: Users’ price perception is driven not by the entire basket, but by specific key products; on these products Grofers may not always be lowest priced compared to the benchmark sources users had
e.g. the following specific products drove users’ product drove users price perception
- A certain brand atta in North India
- A certain brand sunflower oil in South India
- Toor Dal per Kg price and Sugar per Kg price across users
These users were benchmarking us on the price perception creating products (we called them Super KVI) from the places they bought (mostly regional supermarkets and specific online players)
And when we benchmarked again on these products and with the sources users mentioned, we found that : Some key perception driving products were not always lower priced compared to benchmarks (more than 40% cases were higher priced). Insight Validated with data
And now we know what key problem exists and why.
Step 4: Prioritising the refined key problem to solve
Key Problem to Solve: Users see that Grofers prices are higher on certain high price sensitivity products (super KVIs) and hence they do not place their first order or come back
Next Steps: Deciding what to build to effectively solve the key problem (Product Solutioning)
Outcome: Experiments that improved discovery of these perception driving (super KVI) products after keeping their prices lower than competition benchmarks led to a massive increase in first order conversion and also improved subsequent retention.
So now we have a basic structured approach to problem discovery and how to move from ambiguity (just a goal metric) to a specific user problem:
This is the end of part 1 of this 2-part series on problem discovery.
In part 2, we will go through Meesho’s Lower price over convenience problem and derive more learnings around how to adapt problem discovery technique in different type of problem statement.
