Karl returns to the Mapscaping podcast to discuss his latest venture, Tyche Insights – a platform aimed at building a global community of geospatial storytellers working with open data.
In this conversation, we explore the evolution from his previous company, Building Footprint USA (acquired by Lightbox), to this new mission of democratizing public data storytelling.
Karl walks us through the challenges and opportunities of open data, the importance of unbiased storytelling, and how geospatial professionals can apply their skills to analyze and share insights about their own communities. Karl shares his vision for creating something akin to Wikipedia, but for civic data stories – complete with style guides, editorial processes, and community collaboration.
Featured Links
Tyche Insights:
- Main website: https://tycheinsights.com
- Wiki platform: https://wiki.tycheinsights.com
- Example project: https://albanydatastories.com
Mentioned in Episode:
- USAFacts: https://usafacts.org
- QField Partner Program: https://qfield.org/partner
- Open Data Watch: (monitoring global open data policies)
In Conversation
From Building Footprint USA to Tyche Insights
Daniel: You were on the show back in October 2019, when Mapscaping had just started. Back then you were running Building Footprint USA. Could we start there, for context?
Karl: Back in 2019, Building Footprint USA had an idea: to curate building footprint data — gathering open data, manufacturing footprints using LiDAR and high-res ortho imagery — and not just creating the footprints, but connecting them to attribution. What’s going on inside the building if it’s commercial or residential, the demographics of the people inside, what the property is worth. When we first came on Mapscaping we were just crafting our first products for the US and Canada, figuring out how insurance, telecommunications, and financial services companies could get value from that higher-precision data.
Daniel: I remember wondering whether the Microsoft and Google building footprint datasets would impact your business. Would you start the same company again today?
Karl: There are amazing intellectual achievements going on — Microsoft building footprints, companies like Ecopia in Ontario doing a great job of feature extraction from imagery and LiDAR. If you were starting a company today, the play is really connecting that footprint data to attribution: the addresses, the sub-addresses, the businesses inside, the purpose of a building. That’s the interesting business problem to solve. I sold Building Footprint USA to Lightbox and spent about three years there, but as often happens, you find other interesting challenges. Right now one of the most interesting is that everywhere in the world there’s all sorts of public data — some open, most locked up. How can we as citizens unlock that data and get value from it?
Is Open Data Actually Getting More Open?
Daniel: You’ve worked with open data for some time. Are we on the right trajectory — is it getting more open and accessible, or have we hit a ceiling?
Karl: The trajectory is good. There’s an organization, Open Data Watch, that surveys the whole world and ranks countries on their open data policy. And then there’s the boots-on-the-ground view: tens of thousands of people taking public data — sometimes open, sometimes hard-earned and extracted from governments — and creating new value. Esri has done a lot to champion open data for geospatial. But there’s a ton of data still locked up, and no country can pretend it’s doing an A-quality job of pervasively making public data available.
Daniel: Do you draw a distinction between open and accessible? Sometimes we point at something and say “yes, it’s open” — but look at all the hoops you have to jump through, or the archaic file format.
Karl: What we should aspire to is local and national governments seeing the value of making information available regularly, in ready-to-consume formats, and periodically updating it. When data goes into the public domain, good things happen — for your GDP, for transparency, for accountability. But that’s aspirational. In reality you hit ownership issues — “this is my data, why do you want it?” — licensing barriers like “you can’t create commercial work off this,” or data that’s available but four years old and essentially useless. There are many ways open data breaks down.
The Vision: Citizen-Led Data Storytelling
Daniel: So what value will Tyche Insights create on top of this open data? Where’s your niche?
Karl: We’re strong believers that open data made available to citizens anywhere — to analyze, storytell from, create maps and graphs, and publish — is a good thing. Any public information that gets contextualized helps inform citizens. Beyond that, we hope for a couple of things. One, we believe citizen-led data storytelling creates an environment where people with different points of view can talk about issues civilly. Two, in the United States, local government is roughly $1.5 to $2 trillion of spending — and information that informs policy, spending, budgets, and property development can make government more efficient, decrease spending, and make communities more resilient.
Daniel: You said something important there — unbiased. How do you make sure these stories are unbiased, so we don’t just end up weaponizing data against each other?
Karl: Telling unbiased stories is something you need examples of, and a style guide. Think about Wikipedia — a huge chunk of its style guide coaches people on how to write encyclopedia articles in an unbiased fashion: words to steer away from, ways to avoid interjecting your own point of view, how to cite sources. When we started doing practice storytelling for Albany, New York, where I live, my co-founder Keith would point out, very precisely, “that’s an overloaded word,” or “that phrase suggests a point of view.” It was hard — your reaction is “no, no, this is so well written” — and then you realize, yeah, I’m really interjecting personal bias. So you give people a style guide, you go through editorial control, and you show examples of what unbiased storytelling looks like.
Why Geospatial Professionals Are Well Suited to This
Daniel: Who’s your target audience — geospatial professionals, journalists, concerned citizens?
Karl: Almost certainly all of the above. On the geospatial professional side — a personal anecdote. I’ve lived in Albany for 30 years and had a great career in data and analytics, and I spent zero time looking at my own city’s data until about a year ago. Then I started playing with it and found it fascinating. Three things stand out. One, the skills a geospatial professional uses in their day job lift right into analyzing their own community’s data — if you do remote sensing at work, you can use it to analyze suburban development in your metro area. Two, geospatial-adjacent skills — sifting through data, transforming it, getting value from it — all apply; I got really excited analyzing my first city budget, which has nothing to do with geospatial. Three, geospatial professionals are really good systems thinkers, trained to think about cause and effect — so they’re well positioned to connect demographics, property development, crime, and business growth.
Daniel: In my career I’ve had almost nothing to do with storytelling — it’s always been close to the machine room: fix this, transform that, plug that in. Someone like me would come with the technical skills but no clue how to tell an unbiased story. A template to follow would be really helpful.
Karl: Absolutely. We want people to look at stories that already exist and say, “I like that — I just want to plug in my city’s data.” For Albany we have two dozen stories, on everything from how much solar is being built to what the property tax base looks like, crime, and how to optimize traffic calming measures with a limited budget. We also want to coach people on simple geospatial techniques — how to create a heat map from crime data in QGIS — written up so someone can learn the tool and run their own analysis.
Data Storytelling as a Team Sport
Karl: When Keith and I first contemplated this, we had a very linear, solo view: we’ll help one person tell a story end to end. What we realized is that data storytelling is more interesting and more fun as a team sport. In Albany we pulled in a few other people — one is an expert in nuclear data who loves data and applies that skill set to the city’s data. When it’s a team sport, it’s not all on you. We’re thinking about group projects where everyone analyzes the same thing for their own community — say, crime data — following a rough template, supporting each other, even comparing results. It breaks data storytelling out of the “guy in his mom’s basement” image and treats it as something catalyzed by a community.
Daniel: It would be nice to have accountability — someone saying “come on, stay on track.” And it feels like there’s an appetite for this; more and more we see open-source intelligence quoted by legacy media. Having someone put guardrails around it would help.
Karl: This gets back to who actually reads or uses this. We think of both readers and users. A reader might say, “I just want to read everything written about my community to be smarter,” or follow a topic — infill housing, traffic calming. A user could be a journalist, academic, or non-profit. With Tyche Insights, when you write a story you contribute it to the community under a Creative Commons license — anyone can use the data and your contextualization for free. A journalist who doesn’t have time to analyze the data themselves can grab a story that helps inform their readers. The types of users are fairly limitless: journalists, academics, non-profits, corporations, politicians, policy developers.
Daniel: I can see this being a great way for someone to build their name. If your story gets picked up by a journalist, that’s a stamp of confidence — and maybe that person gets a leg up. A lot of people who love making maps would like to do it for a living.
Karl: If you gave me a job doing data storytelling all day long, I’m in. We really believe there are people who love taking data and turning it into value that informs public discussion and holds government accountable. If those end up being people developing a personal brand for data storytelling — fantastic.
AI, Reproducibility, and Going Global
Daniel: What challenges do you see, or have you faced already?
Karl: You can’t get through any technology discussion without AI these days. We don’t want people to just make a public information request, unlock some data, and throw AI at it to “write me a story.” That misses the point — and civic AI engagement just isn’t good right now, no matter the model. So one challenge is figuring out the right place for AI and whether to put guardrails on it. We’re heavily modeling the community off Wikipedia, but it’s different enough that we’ll have to figure things out along the way — deliberately not being prescriptive, working hand in hand with the community on questions like how a storyteller should cite sources or disclose how much AI they used.
Karl: The analogy we use is water. All the data stories are free for anyone, in perpetuity — but we also want to run a business on top. You can drink from the fountain for free, but we’ll find ways to bottle the water: additional metadata, APIs. And AI companies may come along wanting a great dataset in its totality to ingest. If we create a corpus of data plus contextualization, covering both the good and the bad of how local governments work, that can give much more meaningful answers when someone draws on it for civic engagement guidance.
Daniel: What about reproducible research — will you push that? When I looked at albanydatastories.com it didn’t look reproducible to me.
Karl: Good question. There are things we don’t do on that site that we will do as part of the Tyche community — sharing more about the methods, the process, and any caveats or shortcomings. If someone notices the data is a little thin, or biased, or old, we want them to be able to say so. And we want them to share the process and even the tools — if someone wrote Python code to do the transformations, share it. It helps with transparency, and what better way to spread these methods than “here’s the story, here’s the recipe I used, here’s the kitchen utensils — grab them, get the data for your own city, and try it.” That replicability across cities also validates the work.
Daniel: Are you focusing on a particular geography first, or is it global?
Karl: Nothing prevents this from being global. We’re starting in the United States for practical purposes, but a few things need figuring out. The legal side — access to public data differs by country. Supporting people — there are countries where telling a data story, telling a little truth to power, could have consequences that don’t exist in New Zealand or the US, so how do we let people stay anonymous or give them protection? And getting out of an English-first-world mindset to support Spanish, Tagalog, and a hundred other languages. To me that’s where the community helps figure things out.
How to Get Involved, and What Success Looks Like
Daniel: If someone listening wants to be involved — almost a founding member of this community — where can they go?
Karl: There are two main resources: tycheinsights.com and wiki.tycheinsights.com. The wiki is a fairly standard wiki used for all the storytelling and publishing, as well as the supportive activities. On both sites there’s a link — a Calendly link — so if you want to talk to any of us, you can book 15 or 30 minutes. We recognize this is new and people may want to talk through their ideas. We want to engage, help people brainstorm, and answer their questions.
Daniel: Final question — what does success look like for this venture?
Karl: One of our inspirations is USAFacts — usafacts.org — a non-profit founded by Steve Ballmer that takes public data at the national level and does storytelling widely recognized for being unbiased and fact-heavy. Success is taking that model of unbiased data storytelling with public data and making it available globally and at local levels — unlocking data, helping with transparency and accountability, having data-driven conversations. And there’s a personal corollary: nine months ago I didn’t know much about the city I’d lived in for 30 years, and now I know a heck of a lot more. I can engage in conversations, teach people, and learn from others about how my city works. We hope there’s a “raise all boats” value that comes from public data storytelling.
