“Product discovery is something that’s incredibly important for product teams to be doing at any scale, any size, and any stage. And despite that, it’s one of these things that teams don’t do as frequently as they should”
- Sachin Rekhi, Founder & CEO Notejoy, ex. Product Head at LinkedIn 
In our conversations with product teams from 100+ seed-stage startups to unicorns, we discovered that most teams don’t have continuous product discovery processes. User feedback surveys, customer interviews, and usability testing are pervasive across these teams. But rarely done more frequently than once a month.
Teresa Torres, an internationally acclaimed product coach, makes a case for establishing a continuous product discovery process - “The best [product] teams recognize that digital products are never done. We [product teams] can always iterate and improve … if we are continuously making decisions about what to build, we need to stay continuously connected to our customers so that we can ensure that our product discovery decisions will work for our customers.” 
An easy way to shift from once-a-quarter user research studies to continuous product discovery is to set up continuous in-product feedback channels.
As companies scale, product teams delegate user research and only hear a filtered version of the customers’ voices. But hearing from the customers first-hand is the most important part.
Sachin Rekhi recommends setting up feedback rivers, which continuously and automatically collect in-product feedback at specific points in the product journey. Some examples are:
After a customer has been using a product for 14 days, they will get an NPS prompt within the product. The survey asks users how likely they are to recommend Notejoy to a friend or colleague and prompts them to give the product a rating and provide a reason for why they gave that score.
Anytime a user cancels an account, a feedback prompt asks users to explain why they are canceling.
Mature products like Google Search, Uber, Swiggy, and Netflix have such feedback rivers. Feedback prompts ask users about their experience immediately after they complete core product actions or at a certain frequency.
Feedback after every order on Swiggy
YouTube periodically asks for feedback on videos
Periodic NPS on TechCrunch
Such prompts that ask for feedback immediately after the user completes the action provide accurate and highly contextual data as the user is in the moment and remembers the experience clearly. This enables product and business teams to make better-informed decisions by continuously monitoring how users perceive the core experiences of the product.
The frequency of asking users for in-product feedback is a tradeoff between negatively impacting the user experience and collecting data to continuously improve the product. As a rule of thumb, the same user should not be asked for feedback more than once every two weeks.
Building such systems in-house limits the potential upside
Some startups have built feedback prompts in-house or use product engagement tools like Pendo or Clevertap. But will be missing out on many opportunities.
A lack of monitoring abilities over the collected feedback data limits product and business teams to track feedback trends in real-time. Querying, categorizing, and filtering the qualitative feedback is time-consuming. Much more if the data is in multiple languages. This reduces the frequency of checking the feedback data and leads to missing out on discovering customer needs or desires quickly.
The feedback data is disconnected from the analytics stack. If the collected feedback data is connected with quantitative (user events and actions) data, a lot more insights like “what negative feedback was given by users who churned” could be generated.
Taking immediate corrective action on customer feedback is impossible because it is not connected with the rest of the product stack. For example, the customer support team could have checked in users who had a terrible product experience. The engineering team could have fixed the bugs earlier. Product teams could have pre-empted the drop-offs in the engagement at specific flows. But most of this is not possible because the feedback data is not pushed to tools like Slack, Intercom, etc.
Cannot reconfigure feedback prompts easily because these prompts are hard coded and require a release for any changes. For example, just changing the question from a numeric to an emoji rating scale requires a new release. Similarly, it is hard to update the targeting or sampling logic.
Using Blitzllama to monitor feedback data and build workflows
You can get a full-fledged Google-esque in-product feedback stack using Blitzllama. Blitzllama takes < 30 mins to set up.
1. Collection: Launch highly targeted feedback prompts across platforms
Multi-platform surveys can be launched from a single dashboard. Specific user cohorts can be targeted right after they complete an event inside the app.
The in-product surveys can be edited in real-time. No app releases are required to publish surveys.
2. Monitoring: Create custom dashboards and conduct deep dives
Create and save custom charts on a single dashboard, so all teams can monitor and listen to the voices of the customers.
Conduct analysis and deep-dive on the collected feedback data.
3. Workflows: Automate and sync data across your product stack
Sync user and feedback data with Amplitude, Clevertap, and Segment. Configure alerts on Slack. Connect Blitzllama to any databases using Webhooks and APIs.