We believe teams can harness the power of AI in customer research by following a few principles.
These are the principles CustomerIQ is built on.
This is our practice.
Unstructured data, especially audio, video, and text, is notoriously hard for customer research teams to manage and analyze. It takes a long time to read through, pull out relevant snippets, and organize those into something meaningful.
We call this the "unstructured data problem." This problem historically led teams to collecting only what they deemed most important, exactly when they needed.
In the old way, customer research is done manually, transactionally, and in silos.
Whole teams scour product reviews, interview transcripts, support tickets, and bullet points from notes to map out themes and pinpoint customer pains and desires.
If there are no answers to be found, teams take to surveys, customer interviews, and in-app feedback to fill research gaps.
Due to a lack of organization and communication, teams duplicate research, or worse, exhaust customers from repeated discovery on the same topics.
Eventually pains and desires are communicated to solution teams like engineers and designers in the form of slide presentations or PDF reports.
This type of research is done transactionally, it's not continuous. After all, it takes a ton of time and effort. Time and effort not spent doing things like building and marketing.
The problem is: deep knowledge of the customer is built over time, not through one-off transactions, and manual processes limit the breadth and depth of knowledge we can build.
A better way would be to continuously discover customer needs and make that discovery available to every team, at any time.
Thankfully, that's what we have in CustomerIQ.
Now, with CustomerIQ, we can automate the synthesis of unstructured data. This doesn't just save us time, it also opens up a world of possibility around where we find insights and how we socialize them with our team.
In the new way, product, marketing, and sales teams can use CustomerIQ to analyze product reviews, interview transcripts, support tickets, and notes, but also mine other existing stores of customer intelligence: support calls, sales discovery calls, competitor reviews, deep research reports, and more. Automatically and in minutes rather than days.
We can ask more open-ended questions to uncover rich insights we wouldn't have otherwise found.
We can shift from collecting only small amounts of unstructured data, to collecting all of it.
The problem shifts from, "How do we analyze this?" to "What do we have to be analyzed?"
We shift from thinking about research projects transactionally, to continuously: building deep knowledge for us to leverage in developing and marketing solutions customers love.
This paradigm shift is the result of what we've built with CustomerIQ: a place to centralize, analyze, and socialize customer feedback with blazing speed using AI.
With this new technology, we're able to think about customer research and discovery in new ways.
Because of this, we've outlined a number of principles we think teams should follow to get the most out of AI in their customer research. We call this our practice:
Let's dive in.
With our stockpile of feedback in place and organized, we're ready to move onto analysis. Depending on what you're researching, you're going to want to filter for specific customer/user segments, classifications of insights, or attributes before running an analysis.
For example, if you're trying to build a new solution, you likely want to analyze customer problems (so you can solve them). It probably doesn't make sense to include comments about competitors or little snippets of praise about existing features. In this case, you should classify insights in view by those describing problems vs preferences vs requests vs others. You can classify by whatever makes sense. The point is to classify, then search or discover.