Anamap Blog
Product Data Interview: Luisa Bider
Interviews
Updated 2026-11-16

This is part of our Product Data Interviews series, where we ask product managers across the industry the same set of questions about how they use data, what slows them down, and what helps them make better product decisions.

Luisa Bider is Product Lead at Flourish, the data visualization platform by Canva. She co-owns the product roadmap with engineering and design leads, covering AI features, enterprise security, the editor experience, and visualization templates and features.
In this interview, Luisa explains how data challenged the assumption that free usage limits were the main barrier to AI Assistant adoption. She also shares why validating numbers takes more time than analyzing them, how shared definitions support self-serve data access, and why a finding still needs focus and ownership to become a product change.
Connect with Luisa on LinkedIn.
1. What kind of product decisions are you personally responsible for at Flourish?
I co-own the product roadmap with the other leads, mainly engineering and design. That covers the platform, from AI features to enterprise security and the editor experience, and also new visualization templates and features. We have way more ideas than capacity, so a lot of the job is deciding what not to build. Roughly 80% of our capacity goes to big strategic bets and 20% to everyday stuff like bug fixes and small features. Ideas come from everywhere, so I spend a lot of time triaging and shaping them with the team.
2. When you’re evaluating whether something is working (or worth building), what signals matter most to you?
First, how much something comes up in support and with customers. If it keeps coming up, it matters. Usage data comes second, and I use it to put those asks in context. For example, we tend to prioritize work on our most-used visualization types over rarely used ones, even when the idea for a rare one is great.
3. Walk me through a recent product decision you made that involved data. What did the process actually look like?
Free users get a limited number of messages a day with our AI Assistant, and we kept hearing the limit felt too restrictive. Several people on the team wanted to loosen it, from not counting clarifying questions to making it unlimited for a while. The assumption was that the limit was the main thing holding back usage.
Before debating solutions, I asked our data scientist three things. How many people hit the limit? Do they upgrade when they do? Does the limit explain how much people use the Assistant?
A meaningful share of free users hit the limit, and those users were more likely to upgrade. So the limit was doing useful work, and it wasn’t the biggest lever. The bigger opportunity was helping more people discover and try the Assistant in the first place.
So we’re prioritizing making it easier to find and try, and we’re planning to A/B test it to check it makes a real difference before committing further.
4. How confident do you generally feel in the data available to you when making product decisions? What tends to increase or reduce that confidence?
It depends on what I’m trying to learn. For usage inside Flourish, like what people create and which chart types they use, I’m quite confident. For a quick query on a table I understand I’ll just run it myself.
What I’m keen to learn more about is how Flourish is perceived and found, and the detailed journeys between products and parts of Flourish. I’d also love more visibility into where people embed their visualizations, since that’s where most of our work gets seen.
It gets tricky when data lives in two places and the numbers don’t add up. That’s when I get nervous and ask our data scientist. Same goes for the growing number of internal AI tools that produce sensible-looking numbers. They’re quick, but it’s hard to know whether to trust them, so I get their findings validated before acting on them.
5. What’s the most frustrating or time consuming part of getting the insights you need to make a decision?
It’s not the analysis, I love that part. It’s working out whether the numbers can be trusted: tracing definitions, hunting for historical quirks, and finding who knew how a query was built. Different people also end up with slightly different numbers for the same question because they used different sources or queries, and then we have to work out why before we can talk about what to do.
6. How self-serve is data access for product managers at your company today?
Pretty self-serve, and I’m really happy about that. I can write and run queries myself, helped by AI tools and our BI tool’s integration, so I’m rarely blocked. It works best if you already know the schema, the pitfalls, and which metrics to trust. Harder questions, like methodology or causal inference, still need a data scientist.
7. What's the hardest thing about turning data into action rather than just more dashboards or reports?
Capacity and focus. Data rarely settles what to do, because we have far more ideas than we can build, and more dashboards don’t help with that. What helps is agreeing on the decision the data is meant to inform before anyone pulls it. It’s also tempting to keep investigating when the honest answer is “we can’t fully explain this, here are the gaps.” And getting from a finding to a prioritized, owned ticket takes more effort than the analysis.
8. Are there product metrics or definitions that people at the company regularly interpret differently?
Yes, a couple. “Activated” is one. In practice it just means a user created a visualization, but people sometimes assume it’s a more refined set of actions (and maybe it should be). Retention is another. Some people report a percentage and others the number of days a user was active, so the numbers can’t be compared.
9. What’s a surprising or overlooked source of product insight that you think more teams should pay attention to?
Customer-facing teams. Sales and success often hear what customers need long before it shows up in metrics.
10. What advice would you give another PM trying to make better product decisions with data?
Get your team aligned on definitions before you dig in, and document everything: how a metric is defined, how a query was built, and what you don’t know yet.