Picture this: You’re a brand new product manager.
You came from 12+ years in Customer Service. You have a Computer Engineering degree, but you’ve been managing a Social Media Team while building SRE (Site Reliability Engineering Team) for customer support transformation.
You know customers. You live and breathe the customer journey. You can tell you about zero-moment-of-truth (ZMOT), intent, decision-making, purchase, download, install, support contact, refund – the whole d*mn thing.
But metrics? The right KPIs for a product? That’s… new territory.
The Impostor Syndrome is Real
Everyone expects you to know this stuff. You’re a PM now. PMs are data-driven. PMs know which metrics matter.
You open analytics tools. They look familiar—similar to the customer service dashboards you’ve used.
But what should you measure?
And more importantly: how do you calculate it?
—
Enter Julius.ai
I learned about Julius.ai through one of those “Top 10 AI Tools You Need in 2024” posts on X.
You know the ones. Usually clickbait. Sometimes surprisingly useful.
Julius.ai is basically a data analysis AI. You can feed it data, ask questions, and it helps you think through metrics, visualizations, and analysis.
But here’s how I actually used it: As a sparring partner for metric design.
—
The Ritual
Every quarter, for a week or two, I’d sit down with Julius.
Not to have it build my dashboards. Not to have it tell me what to measure.
But to exchange ideas.
Typical conversation:
Me: “I need to measure app health. We have Android and iOS apps. What should I track?”
Julius: [Suggests a bunch of metrics]
Me: “How do I measure the funnel from app discovery to subscription? And what about the conversion rate between platforms?”
Julius:[Explains calculation methods]
Me: “What’s the difference between tracking active users vs. logins? If someone downloads but never subscribes, how do I categorize that?”
Julius: [Offers ideas]
I wasn’t asking the right questions. I was throwing ideas at the wall to see what stuck.
—
What Julius Actually Did
Julius gave me ideas on how to measure things.
Not perfect frameworks. Not the “right” metrics. Just… possibilities.
“Oh, you could measure it this way. Or that way. Here’s how other apps track conversion.
The real learning happened when I took those ideas and filtered them through:
- What actually matters for my product
- What’s realistic to track with my resources
- What stakeholders and the business actually care about
Julius was a brainstorming partner, not an oracle.
—
The Breakthrough Moment
After a couple of months of this ritual, something clicked.
I’d drafted a dashboard. I had clear hypotheses about what to measure and why. I could articulate the relationship between metrics and business goals.
I could present it to stakeholders and get decision.
—
What Made This Work
- I had a specific learning goal. Understand how to design metrics for mobile apps.
- I brought context. I already knew the customer journey, just needed to translate it to product KPIs.
- I treated output as input. Everything Julius suggested went through my filter + stakeholder validation.
—
For New PMs Figuring Out Metrics
- Start with questions, not tools
- What does success look like for your product?
- What user behaviors indicate health vs. decline?
- What predicts the outcomes you care about?
- Leverage your background
- I came from Customer Service → I understood customer journeys.
- Engineers → what data is being collected, where it is stored, what is being measured
- Marketers → funnel conversion
- Designers → engagement quality
Your background is a perspective others don’t have!
- Use AI after you’ve already started thinking
AI helps structure thinking you’ve already begun. It can’t do the strategic thinking for you.
—
The Takeaway
Julius helped me bridge the gap between my “Customer Service” background that understands users” and “Product Manager who can quantify that understanding.”
There’s probably an AI tool out there that can help you, too!
But start with your problem, not the tool.
For me, it was: “What are the right product KPIs? And how do I calculate them?”
Once I knew what I was trying to solve, finding the right tool was easy. Well, reading “Top XXX Tool” posts on X or Threads sometimes help.
—
Next in series: Midjourney: 9 Months of Fun and Ethical Debates
—
About the Author
Former Customer Service Leader turned Product Manager, stumbling through the AI revolution with 12+ years of customer obsession and a newly-discovered love for metrics that make sense. Currently using AI tools strategically instead of desperately.
Leave a comment