Demystifying the Product Discovery Process
This content was inspired by my Lecture at Product Camp 2019.
Creditas, Brazil’s leading secured lending platform, has undergone a period of accelerated growth, which has been thrilling to experience.
And, since we were born digitally, technology and product are essential to advancing our business. To this end, the Product area applies the most modern processes to determine what to launch in the market.
Product Discovery
Product Discovery is a process, applied by Product, to establish what will be developed — at its core is rapid learning almost every day.
A good Product Discovery should prioritize the needs of the clients, value alignment with the main business owners/stakeholders, and follow the business’s guidelines.
The Opportunity Solution Tree
The opportunity solution tree, designed by Teresa Torres, is a powerful tool used by Creditas’ Product area. It is applied by Product Managers as much as technology and business areas.
We broadly establish an objective and/or an impact we want to generate in our business. Next, we list opportunities that give rise to the desired impact. Then, we design solutions for each opportunity. Finally, we define how we’re going to test each case.
The structure of Creditas’ Product and Technology team
At Creditas, we were inspired by the organizational model proposed by Spotify. The following people compose a Squad:
Product Manager
This is the person who leads the process of Product Discovery. On a daily basis, they study the market, carry out benchmarking, and conduct meetings that encompass:
- Team ceremonies: planning, refining, retrospectives, and the Product Discovery itself;
- Alignment with the business units and main stakeholders (business owners) involved;
- Recurring conversations with UX Designers to align solutions;
- Direct contact with Analytics to discuss business metrics.
It’s worth pointing out that a Product Manager at Creditas, generally, has a very analytical profile and possesses the autonomy to carry out analyses. We even have in-company Excel and SQL courses which enable everyone to explore metrics in the best way possible.
UX Designer
UX Designers are involved throughout the entire process of Product Discovery. Their participation spans from the study of the problem to the definition of the solution, while focusing on the best experience for the user. Part of their routine is to construct solutions alongside the team by communicating and aligning themselves with the Product Manager and development team, so they can address the interests of the user and the business simultaneously.
Tech Team Leader
A role which is essential for the technology team, since they are responsible for managing all the members of the development team (developers and Tech Lead I). Moreover, they are the guardian of the processes and metrics of the delivery stage. During Discovery, a Tech Team Leader usually estimates efforts and helps to align delivery estimates with Business Owners.
Tech Lead I
The role which represents technical leadership within the squad. They help the Product Manager understand the technical viability of a delineated solution, and they are generally the person that helps conduct the definition of software architecture.
Developer
In the Product Discovery process, the development team contributes technical insights which explore the viability of business solutions or even proposes solutions which could only come from those responsible for building the product.
Business Owner
Each of Creditas’ squads is composed of either one or various Business Owners. I work inside the Acquisition team, and to us, the Head of Performance Marketing is the person who fulfils this role. The Business Owner provides strategic business information and guidelines, as well as inputs that are valuable to Discovery, given their context and background working with the squad.
Analytics
The Analytics area at Creditas is one of the company’s differentials. I’d never before encountered a team with such a large impact on an organization, which they achieve through their knowledge of the business, expertise in complex analyses, and modeling. The common profile of someone working in this area is an ex-consultor for a strategic consultancy.
Analytics assists the Product Manager on a daily basis, whether it be with the calculation of each initiative’s impact, updating each squad’s principal metrics, or helping explore problems during the Product Discovery process.
The Auto Equity Lending Product
This is a simplified version of the process to solicit a loan on Creditas’ website, where a car will be pledged as collateral:
- Discovery: the stage that brings someone to our website;
- Lead: form that the user fills out to simulate their loan;
- Decisions: the stage where we analyze the person’s credit profile and asset to evaluate if they are within our policies;
- Offers: if the user is approved according to our decision drivers, they receive an offer through the site;
- Documents: the stage where the person submits some documents to complete the loan application;
- Contract: the last stage for approval; the contract is issued, signed by the client, and they receive their money.
Our Case: Putting Product Discovery into practice at Creditas
To demonstrate how we apply the opportunity solution tree, I’ll focus on the stage where a person registers themselves to perform a simulation:
A metric we monitor on a daily basis is the conversion of people who register by filling out the form for the simulation. Let’s say that we want to increase conversion by 5% in this stage; this is now the principal focus of the opportunity matrix. To meet this goal, our team thought up some opportunities and problems, depicted in the following slides.
1st Lesson: Create systems that allow quick hypothesis tests despite an increase in complexity of the system
Time to market is something we always keep in mind at Creditas. Therefore, we need to create systems that allow tests to be run in a timely and cost-effective manner which promote continuous learning. To this end, we’ll check out the first opportunity we mapped:
We focused on simplifying the steps the person takes to fill out the information of the collateral (in this case, a car). In the first version we asked a series of questions about the car’s year and model, which can vary from 3 to 9 fields. We aimed to simplify this process and substitute these fields with just one, which asks for the license plate of the car that is being used for the simulation.
The result of this experiment was a significant decrease in conversion:
2nd Lesson: Product Discovery is continuous
Given the result of the first experiment, we decided to create a new opportunity solution tree to analyze the conversion of the license plate field. We arrived at the following potential opportunities and problems:
3rd Lesson: Using analytical events tells us plenty about the user’s behavior
To support the opportunities we set up and analyze the impact of each one, we configured a series of analytical events using Snowplow to track the behaviour of users.
The team then did an in-depth analysis of the decrease in conversion using the results of the first experiment, and we found the following:
Since the largest decrease, in version 2, was in users who selected “I don’t know my license plate”, we reflected on how to conduct an experiment to better understand this behavior:
The second experiment was characterized by asking people if they knew their license plate or not. The results were as follows:
Lesson 4: At a company with accelerated growth, it is crucial to have a Platform mindset
At a company with accelerated growth, it’s essential to keep scalability in mind. When we develop products, we always think about building our deliverables in a way that can benefit other business units and products. This is what it means to have a Platform Mindset.
An example of this is the second problem we had with registration, in which we needed to recover clients who left in the middle of filling out the site’s registration form.
Symptoms of a bad Product Discovery
Despite being a often-discussed subject in the Product community, few talk about missteps. Of course I’ve had bad experiences, but they were important for me to learn about and evolve the Product Discovery process with my team.
Tools that we use in Product Discovery
These are some of the tools we use on a daily basis, not only to analyze data but also to realize experiments.
What about you? Tell us about the challenges you’ve had throughout the Product Discovery process!
If you want to see the slides in their entirety, check out this link.
Interested in working with us? We’re always looking for people passionate about technology to join our crew! You can check out our openings here.
