Continuous product research is an ongoing cycle of learning from users that runs from before you write code through launch and beyond, not a one-time project. At Sprig, we break it into five phases: discovery research, concept testing, usability testing, post-launch evaluation, and continuous UX measurement. Each phase answers a different question and feeds the next.
Research is not a one and done project. In my work as a product manager and now a CEO and founder, I have learned it is a continuous circle, and it is never finished. It starts on day one, before you have written a line of code or dreamed up a design prototype, and it continues through launch day and beyond. At Sprig, we have organized that ongoing process into the five phases below so any team can run it as a repeatable loop.
Key takeaways
- Research is continuous, not a phase you finish. It begins at customer discovery, before any code is written, and never truly stops.
- The five phases are discovery research, concept testing, usability testing, post-launch evaluation, and continuous UX measurement.
- Each phase answers a specific question and supports a specific decision, so you always know which method to reach for.
- User research finds the "why" that A/B testing alone cannot. Only about one in seven A/B tests produces a clear winner, so testing blindly wastes weeks or months.
- The cycle compounds. Post-launch and always-on measurement surface the next set of problems, which feeds the next round of discovery.
Phase 1: What is discovery research?
Discovery research, also called exploratory research, identifies pain points before they become a problem in the live product. If a feature already launched, it can reveal issues that stop users from taking a desired action. Done early, it shortens the product lifecycle by avoiding wasted work and reaching the most effective solution in earlier iterations.
Here is a real example. Sprig worked with a popular real-estate technology company that noticed onboarding drop-off was much higher than expected on its "Get a Quote" page. The team had three options to understand why:
- Run a variety of A/B and multivariate tests to adjust content on the page and remove fields.
- Take an educated guess based on past learnings and assumptions.
- Conduct user research.
The first two options would take a few weeks to several months and likely waste effort, because only one in seven A/B tests produces a clear winner. By running a few simple in-product surveys, the team went straight to the source and learned from users completing the "Get a Quote" page in real time. Two findings stood out: users were hesitant to give their phone number so early in the quote process, and many visitors were not planning to get a quote at all. They were just shopping around, in a completely different part of their journey. When the team removed the phone number field, the conversion rate increased 10% almost immediately.
Phase 2: What is concept testing and when do you use it?
Concept testing narrows several possible solutions down to the one worth building, before you invest significant engineering and design time. Fixing a funnel is only one use of discovery research. More often, discovery surfaces several concepts for one issue, and the goal is to choose the right one quickly and get it right before you build.
Say that while researching the "Get a Quote" experience above, the team learns that understanding a mortgage is confusing for users and stops them from moving forward in the home-buying process. The team develops a few ideas and lands on an interactive mortgage calculator. That might be the best option, but it will take real engineering and marketing resources to build and launch.
This is why you de-risk the project first: create several product mockups and test them with users before building. Unmoderated concept testing makes it easy to compare a few options and gauge the viability of each. When testing multiple prototypes, limit the options to two or three to lessen the cognitive load on test-takers.
Phase 3: How many participants do you need for usability testing?
Usability testing needs relatively few participants: a common guideline is at least 5, and up to about 50 for larger studies. With the most compelling mortgage-calculator concept selected, usability testing confirms the design actually works. Participants complete tasks using a prototype or a live app, which surfaces points of friction and chances to improve the experience.
Participants are asked to "think aloud" as they work, explaining their questions, hesitations, and challenges. For the mortgage calculator, tasks might include "Can you adjust your down payment?" or "Set your rate to 30-year fixed." The Nielsen Norman Group recommends testing 5 users, because that small group surfaces most of the usability problems you would find with many more. Usability testing is the simplest technique in this article, so do not overcomplicate it. The point is to confirm the design is functional and users can complete the intended actions.
Phase 4: How do you evaluate whether a launch worked?
Post-launch evaluation measures whether a change delivered the intended result by comparing outcomes against a baseline. User research is continuous, so the work does not end at launch. After shipping new features and flows, you measure satisfaction and compare the results with previous data to confirm the changes worked.
Back to the mortgage calculator: compare the metrics from the previous onboarding experience with the latest iteration. Run the same in-product survey before and after the calculator ships to see whether it improves the experience and drives the right behaviors. You might ask, "How confident are you in the results you received?" to gauge whether the calculator delivers on its promise of building buyer confidence. If it does not, open-ended responses will show why and point to next steps, and Sprig's Synthesize Agent can turn those responses into clear themes. This process is often iterative, with multiple rounds of research and solutioning.
Phase 5: What is continuous UX measurement?
Continuous UX measurement is always-on research that surfaces problems you did not know to look for. Not all research maps to a specific, identifiable business problem like poor onboarding conversion or a dip in engagement. As teams set up and scale research, continuously monitoring the experience reveals unknown issues that are not yet on the product team's radar.
This kind of research does not need to be long or complex. Adding simple in-product surveys on common pages to measure net promoter score (NPS) and customer satisfaction (CSAT) can lead to some of the most important "aha moments" in a business. The signal you are watching for is a shift in sentiment on a page you assumed was healthy.
Back to phase 1: the cycle begins again
The cycle then begins again. Insights from continuous measurement and post-launch evaluation keep uncovering new pain points, which feed a fresh round of discovery research. In business, and especially in tech, there are always new problems to solve, and they need to be solved quickly.
That is even more true in the era of agile development. Unlike the quarterly release schedules of a decade ago, teams now work in continuous release cycles and sometimes ship every few days. Companies that understand user research are in a far better position to keep up with that pace and keep their customers happy.
.png)
%20(1).png)