In my last role as a product leader for Amazon’s Alexa, Excel was my “best friend” when analyzing customer feedback/anecdotes. It was common for my team and I to wrangle a few thousand anecdotes of customers having poor experiences and annotate in Excel. Next to each anecdote, we’d manually categorize each issue, create Pivot tables of our categorization, look at the stacked ranks, and use those to inform the roadmap.
If it surprises you that this manual, low-tech process was my go-to method, given the tools at Amazon’s disposal, you’re not alone! I too was pretty surprised that this was the best we had. Given how common this workflow is in product and voice-of-the-customer teams, I wanted to share a few of the pain points I experienced.
1. It’s unnecessary manual
We’ll start with the most obvious. We’d spend hundreds of hours manually categorizing anecdotes. While reading raw anecdotes is super useful, you quickly realize patterns in your categorization that a computer can solve. Excel has basic formula automation, but this was only halfway effective, since it relied on brittle exact-match keyword detection. Setting formulas up became a Sisyphean task.
2. The dreaded _v2_03_04_correct_edits.csv file jumble
We’ve all been there. A team passes around revision after revision to the point where you don’t know what’s ground truth. We routinely had to redo annotations to ensure we were working with the correct analysis.
3. Excel files have a short shelf life
I routinely saw valuable analysis disappear when a colleague left the team. We couldn’t go back and drill into the initial analysis and ended up reinventing the categorization methodology over and over. We encouraged everyone to store their analysis in central cloud repositories, but the road to hell is paved with good intentions.
4. Tracking progress over time became a task of its own
Finally, Excel analysis like this is point-in-time. We’d run these monthly, analyzing last month’s feedback, building up a collection of these. Unfortunately, each had a slightly different methodology, or files got lost, and analyzing how categories of feedback were trending QoQ or YoY became another tedious task, since in a data driven organization like Amazon, providing accurate information about these trends was a basic expectation in business reviews.
It’s not all bad news!
A lot of these pain points were the inspiration behind Unwrap.ai, a company my colleague Ryan Millner and I founded to solve these problems and get you out of “Excel hell”. Unwrap dramatically speeds up the process of integrating all your sources of feedback, analyzing it to understand what the top actionable patterns are, and using those insights to build your roadmap.
Are you ready to leave Excel hell? Book a demo here.
Ready to see what we can do for you?
We'll be in touch within 24 hours.
Oops! Something went wrong while submitting the form.