---
id: "workshop:thinking-with-data-leveraging-analytics-to-better-define-and-solve-problems"
type: "InnovateUS workshop"
title: "Thinking with Data: Leveraging Analytics to Better Define and Solve Problems"
url: "https://innovate-us.org/thinking-with-data-leveraging-analytics-to-better-define-and-solve-problems"
date: "2023-03-09"
duration: "90 minutes"
language: "English"
instructors: ["Richard Dunks"]
recording: "https://www.youtube.com/watch?v=_D2qL0P7rEg"
transcript: "automatic, from the recording"
source: "InnovateUS (innovate-us.org)"
licence: "CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/), as stated on innovate-us.org"
---

# Thinking with Data: Leveraging Analytics to Better Define and Solve Problems

InnovateUS workshop · 9 March 2023 · 90 minutes

Join [Richard Dunks](https://www.linkedin.com/in/richarddunks/), founder of Datapolitan, in this 90-minute virtual training where we will discuss the features of a data-driven mindset and key elements of a data-driven culture in the public sector. We will discuss the common barriers and strategies for helping you and your team move from just using data when it’s convenient to actually driving decisions with the critical information for success. You will walk away with a practical understanding of what a data-driven culture looks like and how you can better foster a data-driven mindset personally and in your organization. Recording passcode: aw6Zw3^R

## Learning goals

- Gain a practical understanding of what a data-driven culture looks like
- Understand how you can better foster a data-driven mindset personally and in your organization.

## Speakers

- **Richard Dunks**, Founder, Datapolitan

Recording: https://www.youtube.com/watch?v=_D2qL0P7rEg

## Transcript

Automatic transcript of the recording: names and terms can be misheard, and the times can be a few seconds off. Each time links to that moment of the video. The transcript alone: https://innovate-us.org/thinking-with-data-leveraging-analytics-to-better-define-and-solve-problems.txt (plain text) and https://innovate-us.org/thinking-with-data-leveraging-analytics-to-better-define-and-solve-problems.json (every segment with its start and end second).

**Richard Dunks**

[0:01](https://youtu.be/_D2qL0P7rEg?t=1) So with that, I'll give you kind of an overview of the agenda. First of all, we are welcome to set our ground rules, and I'll do some introductions. I know you've already done introductions in the chat, that's great. I'm going to ask you to expand on a little bit and introduce myself as well. I want to set the purpose, the intention here for what we're doing, and also kind of explore this question. What makes us data-driven? How do we think with data? This is an opportunity for me to share some of my thoughts and experiences. I'll be liberally lacing with case studies in there, and I'm happy to take anything that you all want to share about your work as well with us. I'll offer you a final thought, and then we'll move into our Q&A section. I do have a PDF of the slides, and I'll go ahead and put that here into the chat if you want to download a PDF of it, as well as the slides are online. You may have noticed I'm just doing this from the internet, and I'll post that in the chat as well. You've got the PDF, you have the slides in there as well.

[1:04](https://youtu.be/_D2qL0P7rEg?t=64) For those of you I know, like me, you want to go ahead and see what's coming and everything else. Perfectly fine. All right, so getting to know you. You've already made your introductions, I appreciate it. You've already let us know who you are, where you're coming from, and who you work for. I'm curious, and if you haven't done your introduction, please do so, but I'd like you to add one place you'd like to be right now on a relaxing getaway vacation, right? Where would you like to be? Maybe your California friends would like to go someplace dry. They have all those atmospheric rivers. New England friends, maybe somewhere warm. Somewhere on a beach. I've got some Ireland folks. I think someone had a... That's good. Southern New Mexico. Awesome. Italy. I have a friend who's going to Italy soon. I'm just so jealous. We were talking, I'm like, oh, go there, do this. I had all the advice for him. Awesome. Cool. We've got some global travelers.

[2:06](https://youtu.be/_D2qL0P7rEg?t=126) Wow. Cool. Thank you so much for that. Now we're going to be in a vacation mindset. That's a great way to start. That's wonderful. All right. So please keep those coming. It's great to see. I also know where everyone's coming from, hasn't introduced themself already. So for me, who I am, so Hannah, I'll give you a wonderful introduction. I really do appreciate it. My name is Richard Dunks. I'm currently in Dallas, Texas. I work for Data Policy LLC, a company that I own. I've been running for almost 10 years now. I would love to be in Thailand right now. It's a beautiful beach there. It's a lovely swing right on the sandy beach. Kusumay, it's a wonderful place to be. So that's my happy place there, if you haven't been. I highly recommend it. We're going to take a pause, though, for a second, because I want to test these Zoom reactions. So because we can't do breakout rooms and it's hard for everyone to talk, I want to try and experiment with using reactions in this. So where you can find these is if you go to your Zoom bar, and mine decided it wants to hide.

[3:13](https://youtu.be/_D2qL0P7rEg?t=193) Nice of it. Go to there. If you can find the reactions. And I want you to react to this idea, right? So if you like beaches, give me a thumbs up. If you don't like beaches, give me a thumbs down. Let's see if we can do this. Thumbs up. And if you have any questions, how to do reactions, there's actually several places you can do it. You can find them down at the bottom of your window. You may have to click the more and come up with reactions. And let's see. Oh, we have a thumbs down on beaches. Alyssa, you want to share just really quick? What's your thumbs down on the beach?

**Audience**

[3:49](https://youtu.be/_D2qL0P7rEg?t=229) I am just not a beach person.

**Richard Dunks**

[3:51](https://youtu.be/_D2qL0P7rEg?t=231) All right. That's good to know about yourself. That's cool. All right. So we got some reactions here. This is helpful. If anyone had any problems with the reactions, please just quickly let us know. Like I said, the more, we'll get your reactions. We have some people doing it in the chat. That's cool, too. Multiple ways of doing it. Awesome. And I'm doing this because I have an intention. We'll be using reactions for feedback in our talk today. So there will be times where I'm going to ask you to react, give you thumbs up, thumbs down on it. And that will be really helpful. So thank you for going through the exercise. I really do appreciate it. Time wise, we had to go through the whole, like, how to unmute yourself, all that kind of stuff. We moved beyond it. So just to give some more context about me, you know, Hannah's wonderful introduction. So I have almost 20 years experience working as a public sector analyst in different ways. Working for the federal government, working local governments, and nonprofits, social service nonprofits, all kinds of kind of areas of this.

[4:57](https://youtu.be/_D2qL0P7rEg?t=297) I am a graduate of NYU Center for Urban Science and Progress, if you're familiar with that program. I like to call it big data for big cities, where we learned a lot about how governments, city governments in particular, work departments, how services are delivered and how to analyze that data. And providing support and training in a number of areas that touch on this, including even design thinking, user research, and all of those kinds of aspects to this work. I'm a former adjunct at Columbia Pratt, and I'm currently an adjunct at Tulane University. If you were here earlier, you heard me talking about that. So I've had an opportunity to kind of teach this material, as well as work kind of in the trenches with lots of people actually doing the work. So enough about me. I'm going to get into this now. I want to set a purpose for this, for this talk. And I'm going to postulate, so I'm going to make a provocation here. There's a lot of cheerleading for data. And I want to be very clear, there's nothing wrong with that. I think it's very important that we advance the cause of data. And I'm going to assume then that you've gotten a lot of these talks already in various different capacities.

[6:01](https://youtu.be/_D2qL0P7rEg?t=361) Maybe you give these kind of talks often in your work. And again, so nothing wrong with that. It's very important to go out there, evangelize, give the gospel of data for those, for the uninitiated and for those that are out there. But I'm also going to suggest that we don't have to talk about why we aren't more data driven. And we talk about the benefits, we talk about why we should be doing it. But it's challenging for it to materialize, right? It's this kind of moral imperative. And I want to talk about that today. Kind of break this down a little bit, talk through it. Sometimes this kind of topic can become a therapy session. That's fine. That may happen. That's cool. We'll roll with it. But I really want to break down this idea of why aren't we more data driven? Why aren't we thinking with data? Which I argue is where we really want to be getting to, where data is really much in the midst of everything we do. So that being said, I kind of want to get some context here with a poll. So I'm going to ask Hannah or Kate who is running the poll here to put this up for us. And the question here is the biggest barrier to being more data driven on my team is our leadership, our team, our tools, our data, organizational culture.

[7:09](https://youtu.be/_D2qL0P7rEg?t=429) I'm going to give folks a few minutes to do that. And for the person just joining us, so we are taking a poll right now on the biggest barrier to being more data driven on your team. Yeah, we've got almost 80% of people have responded. So that's a great number. And it's climbing. Yeah. Go, go, go. This is awesome.

**Unknown speaker**

[7:55](https://youtu.be/_D2qL0P7rEg?t=475) Okay, I'll end the poll now to show the results.

**Richard Dunks**

[7:58](https://youtu.be/_D2qL0P7rEg?t=478) Great. Can you all see the results? Is that visible to you all?

**Audience**

[8:11](https://youtu.be/_D2qL0P7rEg?t=491) Yeah.

**Unknown speaker**

[8:12](https://youtu.be/_D2qL0P7rEg?t=492) Yes, it is.

**Richard Dunks**

[8:15](https://youtu.be/_D2qL0P7rEg?t=495) Awesome. Yeah, thanks for that. These are hands up, thumbs up reactions. So interesting. So our top one is tools. Our middle or tied for second place is our data and organizational culture. Awesome. 12% is our leadership and 6% are our team. Awesome. Cool. So we'll keep that in mind here. So we have 41% are saying our tools. I'm going to try and keep this window open so I can, I can refer to it. All right, so I have for me personally, my experience. I really think it's culture. And at the root of everything is the culture and there's this wonderful quote from Peter Drucker, culture eats strategy for breakfast. And this is also speaking back to my background as a sociologist, my undergrad was in sociology. The culture is extremely important. And I want to talk through some of the other aspects aspects to it. But first I want to get your reaction to this so agree or disagree.

[9:15](https://youtu.be/_D2qL0P7rEg?t=555) This is culture. Thumbs up, thumbs down. I got some thumbs up, looking for some thumbs down. No one's coming down on this. All right, someone want to be brave a thumbs down tell me why they don't think it's culture. I'm sorry Debra I guess because you weren't able to see the poll results. So don't be brave and say why it's not culture. I see a hand up.

**Audience**

[9:56](https://youtu.be/_D2qL0P7rEg?t=596) I, I. Hi, this is Jennifer, I would say, the reason why I don't necessarily think it's culture is because leadership creates the culture. And so the culture they create is what's going to eat the strategy for breakfast in my opinion, and they're the one that's setting the tone for it so by extension, they're the ones that eat strategy for breakfast not necessarily the culture itself.

**Richard Dunks**

[10:19](https://youtu.be/_D2qL0P7rEg?t=619) Okay, Jennifer, thank you so much for that. I'm going to talk through this and Jennifer I want to come back to you because I want to ask you what you think after I've made my case, if you will, around this. I think we have a few lawyers or law department folks in the room so hopefully this will they'll they'll be able to judge on this. So, first of all, I want to talk about why I don't think it's the tools. Spreadsheets are amazing and spreadsheets are awesome, and they are something that we almost everyone in government has on their machine whether you use it or not, there is a spreadsheet on there. And if you don't, there are options like open office Libra office, not to mention Google Sheets, to be able to use this and spreadsheets are powerful tools. By way of example some of you may remember the London whale scandal, this is a scandal where a JP Morgan Chase and or trader lost almost $7 billion in trades, and there was another billion of fines on top of it. But what people miss in this conversation I think it was oh it's just this rogue trader doing all this kind of crazy stuff.

[11:23](https://youtu.be/_D2qL0P7rEg?t=683) He was doing almost all of it in Excel. One of the problems with this was the Excel wasn't configured right, but it's an example of how billions of dollars in our economy upwards of trillions of dollars are traded using spreadsheets, and when we every time you hear about the Federal Reserve Bank or any kind of policymaking kind of work or, you know, financial markets. It's being done in Excel. So much of what we do is being powered by Excel because Excel can do it. Now of course there's lots of problems there's lots of challenges as a London whale exposed, but spreadsheets can do a lot of work when you get into it. I know it can be challenging to unlock all that potential, but it is there. And that's why it's really important to remember we have a powerful tool in our pockets in our on our desks, and it's just about using it. There are more tools you can use more specialized tools but the spreadsheet will do a lot of the work that we can and need to do. I'm also going to argue, it's not the data, I'm going to use myself by way of example.

[12:25](https://youtu.be/_D2qL0P7rEg?t=745) This guy here. Back in the early 2000s was deployed to Iraq, and we were being supported by a platform, and we knew the platform was not doing what we needed to do for our mission success, but our superiors like no no it's all you guys you're not doing it right. What we did I actually went through and I was pulling down from their chat logs if you're more like IRC chat logs like basically guys instant messenger kind of stuff, or teams now I guess I was going through that because that was where all this data was being reported so there's a lot of text mining but I was able to pull it out and be able to demonstrate just how bad the platform was was performing. I think was when it was supporting us it was like, you know, we had a 25% chance of actually succeeding what we needed it to do so. It was not fun. It was not enjoyable, but it just revealed to me it's like the data is there. Yes, it can be very hard to work with with some of the silos that we have and we definitely have silos and restrictions and things like that. But often where there's a will, there's a way to get the data that we need. So the key lesson here the data is there, you may just have to work hard at getting it, but often we have what we need when it comes to the data.

[13:35](https://youtu.be/_D2qL0P7rEg?t=815) Also argue it's not the people, usually. And again by way of example and talk about the New York City Department of Transportation stood up an analytics unit for their meter collections, right so this is the early to 1020 10 this is a 2013 or so. And, you know, and some meters still require coins and if you've never kind of worked with this kind of operation or thought about it. There's a challenge because too many coins in there will lock up the machine so you won't be collecting anything more, but too little and you kind of are wasting your time doing the collection. Manhattan, right, you have these guys and women dodging traffic taking their lives in their own hands in order to go to one of these machines and then it's like, oh two quarters drop out or something like that. So the Department of Transportation stood up this unit. And in order to kind of do some work analytics around, when is the optimal time to get access this meter and pull it down and get the cash. They staffed it with people who were the ones collecting this collecting these quarters or coins or whatever else from it. So they were highly motivated to find answers to this challenge, but they had very little training or experience to do it and they did some amazing things, you know, kind of like acted together did it on their own. It was really amazing to see. And I've seen that time and time again, where these kinds of units are stood up or these tasks come down, and people take it upon themselves to teach them what they know.

[15:01](https://youtu.be/_D2qL0P7rEg?t=901) Because people will find a way to accomplish the task with the data that they're working with. And it's an amazing thing to see, and I'm sure it's happening in all of your organizations. It may not be happening a lot, but it's there, and people really can impress and give them the situation. And I want to talk about this, too. I don't think it's not just the leadership. And Jessica, I hear your point about this, but... I'm going to mute you, Lori. Sorry, we're... Sorry, I'm going to cut your conversation. Not just leadership, but here's an object lesson for this. I think this is true for those of you who work in departments where... Typically, I find this where there's a lot of, like, labor, service delivery kinds of things. So, like, parks, public works, any of those kinds of things. Your executive directors, your commissioners, often, like, 30 years ago, were the ones that were, like, laying gravel, or helping pave roads, or set up playgrounds, or whatever else.

[16:02](https://youtu.be/_D2qL0P7rEg?t=962) Leadership comes up through the organization, and they are often brined by it, right? They're conditioned by that culture. And this came to a head in 2018 in New York City. In New York City parks, there was a massive storm that came through New York City. They had 6,000 reports of downed trees. It was really bad. I mean, a lot of things were problematic about it. The leadership at the parks department, of course, is getting requests from the mayor's office and other electives. When is this going to be clear? When are you going to be done? And they gave an estimate that literally, like, back of a napkin, envelope kind of thing. Of course, that was not accurate. There were all kinds of problems with duplicate requests. So, even that 6,000 requests, it wasn't clear. Was that 6,000 downed trees? Or was that, like, 1,000 downed trees were reported six times each on average? And all these kinds of things. The workflow system, the work order system was very manual. So, it was really hard to know what's your crew doing? How long does it take to do these things? What are they finding?

[17:02](https://youtu.be/_D2qL0P7rEg?t=1022) All that kind of stuff. So, they would say, oh, this is going to be a few days. And then a few days, it still wasn't done. It'll be a few days more. It'll be a few days more. And that's not a way to win friends and influence people in government. You're not knowing what your organization is doing. Now, admittedly, this was an extreme event. But it was symptomatic of what they were dealing with day to day. Where they didn't have the data. They weren't really tracking what the crews were doing. They didn't know how long it took to perform their operations. And when I talked to the leadership about this, they were like, oh, yeah, this is just a one-off. This is just really bad. But, no, we're usually pretty good. I'm like, no, you're kind of not. I mean, you need to recognize that you need to better track and better understand this to be able to better manage your people and your resources. But because they had come up through this culture of this is just how we do it, it was really hard for them to see the alternatives and really understand what was going wrong. So, I would say it's not usually primarily leadership because they're just as affected by the lack of a data-driven culture as everyone on the front lines and everyone else in the organization.

[18:06](https://youtu.be/_D2qL0P7rEg?t=1086) So, that being said, dig it back to it's the culture. Culture eats strategy for breakfast. So, Jennifer, did I convince you or are you still not convinced?

**Audience**

[18:18](https://youtu.be/_D2qL0P7rEg?t=1098) Nope. No, I definitely think it's a very compelling argument. And I do think, yes, from a holistic landscape culture, we'll eat strategy for breakfast. But I think that at times, and I could just be speaking obviously from the lens where I work and what I see, I think at times leadership might minimize the level of influence and impact that they have. That even when you have a toxic culture, not necessarily that one person in leadership is going to help turn it all around. But when you can have a complete strategy that takes into account all levels of leadership and how they reinforce the work that's being done, I think that, yes, a lack of it can eat strategy for breakfast more than culture can.

[19:18](https://youtu.be/_D2qL0P7rEg?t=1158) And I think also if the culture is that everyone in the organization has a responsibility of leadership in some way, shape, or form, then in that way, it could. But I do see your point completely as well. So, I think it's a combination of things. But I think sometimes we don't focus enough on the level of impact that the individuals who are leading the organization have in this work. And to your point, to the data, how we're utilizing the data, what story we're using the data to tell, and the vision that we're casting forth based upon that data. Right.

**Richard Dunks**

[19:58](https://youtu.be/_D2qL0P7rEg?t=1198) I'm going to offer some thoughts on that really quick. And thank you, Jennifer, for standing up in front of this crowd of 102 people and saying all this. I really do appreciate it. It's not easy. Leaders are stewards of culture. And when you talk about the mechanisms by which leadership affects culture, it is, by and large, I would say the biggest is by example, by what people see. It's not what you hear, it's not what's being said, it's by what they do. And I think too few leaders, and you find this, I think, in different fields, a different level of awareness. When you get into the social services end of things, people tend to be a little more sensitized to their actions and what that does, how people respond to it. In the more kind of structured engineering, kind of those kinds of things, it's a little less obvious to people about how their actions impact how people perceive them. It's just kind of been my experience of this. But all leaders need to be sensitized to the impact of their actions because it gets bigger and bigger as you go out.

[20:58](https://youtu.be/_D2qL0P7rEg?t=1258) And I think this is one of the things that I've noticed as well as the military of, you know, you have a senior officer, a colonel or general or someone who will just literally like just run their mouth and say something kind of random about like, oh, I wonder about this or whatever. And then you have all this staff work is sped up and people do all this time and energy and there's a whole briefing and everything. And this officer walks in and like, what are we doing guys. You know, they forgot about as soon as words came out of their mouth, but they just don't realize, you know, the impact that they have and what they're doing. So it's very important with that for them to be sensitive to the impact that they have for it. And I lost my other thought but I know it'll come back to me so thank you very much that Jennifer I really appreciate it and I want to go through this I want to talk about then, in terms of the culture and what we're trying to do. What is it we should be looking for and what are the actions that we should really be promoting in this work that we do. So, I'm going to start off with and against going back to my you know just what is culture.

[22:06](https://youtu.be/_D2qL0P7rEg?t=1326) A great definition I think from the Cambridge Dictionary, the way of life, especially the customs and beliefs of a particular group of people at a particular time. And we talked about this in sociology right it's the prevalent beliefs values and norms of that group. In this case, our organization our team, our agency, you know our state how the whole government is run things like that. When we talk about creating a data driven culture, I think, for me the real North Star the real, you know, proof in the pudding if you will Burmese Road, whatever it is, is this quote from Judy Arsenal. Do you have data to back that up should be a question no one is afraid to ask, and everyone is prepared to answer. Everyone is entitled to their opinions, you can you can have your opinion, whatever you know you think, but having data and evidence to back it up is really fundamental to this idea of a data driven culture and being able to not just throw out statistics, because we know 80% of statistics are just made up right, but really being able to say, how did you arrive that where did that come from, how is that generated.

[23:13](https://youtu.be/_D2qL0P7rEg?t=1393) That's really the essence of what it means to think with data and being open to being challenged, you know, obviously it should be in an appropriate way it should not be, you know, attacking all that's not true. You know, help me understand where that's coming from, help me understand where you're getting that information so I can understand the context for it. Because facts without context or just mere curiosities. It's having the context for it. So I'm going to kind of talk about this culture piece and encourage you to share your thoughts in the chat if you'd like. This is inspired coming mostly from a book by Carl Anderson creating a data driven organization, highly recommend if you're interested in this kind of topic. He was a. He's a number of things but he wrote this while he was working with Warby Parker, who many of you, some of you may be actually wearing Warby Parker glasses right now. He took them from a very startup kind of kind of playing with data lots of spreadsheets and things like that, into something was much more data focused and he talks about, you know, the key things putting in place in his experience.

[24:15](https://youtu.be/_D2qL0P7rEg?t=1455) It's a great resource for anyone really trying to create this culture, and he's the one really that you know made that that impact that observation right yeah we can give them tools we can give them the techniques we can hire smart people, but it's really about the culture we're creating around this, and this is what he came up with as part of that. This is idea of understanding what the culture is. Well, sorry. One second. So, he offered this as a list I do it as a compass and I'll, you'll see why I do it as a compass it to kind of highlight what I think are the main points of this. What are the goals of the organization, what are you trying to achieve. What is the point of what you're doing. This is essential, it is that's why it's at the north of this compass it's got to be the North Star, what are you trying to achieve. I'm curious, you think about this they'll share it to chat or just kind of take it away. You know what does it mean is what's a blockers to us is in setting goals.

[25:22](https://youtu.be/_D2qL0P7rEg?t=1522) Think about that. Because we can have a lot of activity but very little movement. Sometimes there's that activity itself as a barrier right where's Oh no I'm too busy to set goals. I'm so day I'm too busy to kind of set this up and say what I'm actually doing. The other element here is data leadership. And this is really important because it goes something Jennifer said about kind of everyone taking ownership of this, the hiring a chief data officer is not going to make you data driven, it'll help you know help kind of someone wrangle it, but everyone should be taking responsibility for leading the work with data to creating a data driven organization. It's more incumbent on the senior leadership to model that behavior, and how they use it and how they encourage their staff and mentor their staff to use it. It's not having a person come in all the other data personnel, unless we create space for them to do their work. It's the same for a chief innovation officer for even back in the day when they started having chief technology officers like it was important to create space for them to do their job and insert have input into the different aspects of the organization, really help remake the culture around the work that they were doing.

[26:36](https://youtu.be/_D2qL0P7rEg?t=1596) So you think about this, what's keeping these key leaders from stepping into their role. What are some of the challenges maybe in that. Then another kind of piece of this is an open and sharing culture. And this is usually when I'm talking with government folks where they start to laugh and snicker and everything is open and sharing is not usually what government is known for. And there's some interesting reasons we can talk about why that may be and why that is. But when we talk about building a data driven culture is extremely important. You know, we understand the context we have access to the information the things that's going to make our data work, meaningful. And then also with this. It's very important to have broad data literacy. It's really hard to have a data driven organization, when you have people who are uncomfortable, if not outright scared of numbers charts and graphs. Because what happens, that would be ignored or let's accept it uncritically, and both are dangerous, right people like, I don't trust those numbers, where those like oh okay yeah you have a chart so it must be it means it's true.

[27:42](https://youtu.be/_D2qL0P7rEg?t=1662) No, there's got to be a middle ground where people feel comfortable taking the information asking questions, how did you arrive at this what does this actually mean what's the context I need to know for this in order for this culture really to happen, really to catalyze. So there's barriers to this. Obviously, you know in terms of training resources all those kinds of things but find that fundamental level of comfort level. I think is essential for this type of work. Yeah, Jennifer thanks for that. Yeah, exactly. There's so much work. I have a great story about this by the way, there is an agency in New York City. The deputy commissioner required a report to be done every two weeks on all the operations everything that was there and it's this huge report and it took. At least two or three days of an analyst time to produce. And, you know, went up the chain got reviewed by the chief of staff, and the different boroughs and all this stuff went up the chain.

[28:42](https://youtu.be/_D2qL0P7rEg?t=1722) An analyst once forgot to do this he's got distracted and been able to do it from the boroughs, and it went like three or four days and no one said anything and he's like, oh, I gotta go do this and he did it and you know spent the time on it and then no one said anything. So, what does that mean when you're saying like, Oh, I did all this work and give you this report. No one even knows it wasn't there, is that really driving decision making, how is that helping your organization be more efficient. So, and again the question is was it something where they thought they needed it but they really weren't using it they know how to use it they weren't comfortable with it, it was just being filed away, and what does that mean for the organization and in terms of how data is perceived. So, also talk about this once we have goals and we have an open sharing culture, then we have the opportunity to be inquisitive and questioning, you know, is this really how's this advancing our goals, how's this getting where we want to be going. What else can we be doing in this in our work. And then being able to be iterative, you may have heard this term I kind of like test getting things out there and seeing okay, we try this going to happen and and do this kind of learn see if this is going to happen.

[29:49](https://youtu.be/_D2qL0P7rEg?t=1789) And this is a story I like telling with this, the Department of Environmental Protection New York City is responsible for the water supply, and they tested this idea of how to get more people. It's a perennial problem in New York City. You have basically the champagne of water, right, they like to say, but no one wants to pay for it. So they tested this idea. It's actually grounded in science about how to nudge people to pay their bills. And what they did, when they were going to shut off someone's water for nonpayment, there would be someone who would go by a few days before and outline where the shutoff was, a kind of paint on the sidewalk, wherever it was, and then a few days later, a week or so, someone would come by and actually do the shutoff. And they tested this idea of when they marked out the cutoff switch to be turned off, they wrote the word off on the sidewalk. Now that's kind of appalling, you might think, but there's actually interesting science around. And again, it comes from a book called Nudge by Pat Sandstein and the Behavioral Insights team, if you're familiar with them, has done work on this.

[30:55](https://youtu.be/_D2qL0P7rEg?t=1855) There was some work done where they put the utility usage for buildings in the UK out in front, the people agreed to it and they did it. And what they saw people would conserve resources in because now the public available, see what's there. So DEP tried this, you know, it's a big off kind of thing, see if people would actually pay their bills and actually, you know, so they wouldn't cut off service. And to their credit, they tested it and it didn't work out. So, but this is again idea of like, well, let's try this and let's see and let's have a test case against our usual, do we see a change and everything and then learning from that. Instead of just, you know, doing this whole effort and not really testing the assumptions underlying it. You know, and I guess the other thing too, and that part of that data, that leadership there that was there and understanding the goals, we want people to pay for the water they're using. And then be able to test, right. And again, this is the leadership supporting it and the broad data literacy. So people can understand how to do these tests, do the statistics, run the testing on it and understand the results.

[31:55](https://youtu.be/_D2qL0P7rEg?t=1915) But all of this, all this is for not if we don't have self-service. You know, it's in this again, this comes out of the open and sharing culture, right, we want to share information with other people and people can use the information that we share, then the self-service becomes possible. So if you've tried to do a project like, oh, this is a great idea, I know we can do this, but the day is going to take you six weeks or more to get. And there's probably not going to be the data that you want, you have to go through all these hurdles to get it and all these kinds of things. That is not, that does not encourage innovation. Like you lose the momentum of the moment, you lose the timing of it. So that's to say, that's to say that, you know, the janitor should have access to your restricted private, you know, student data or something like that. But it is to say if someone should have access to it, why don't they, right? And if it's going to help their job, like give them the access that they need in order to really advance this. So these are what he said, again, the key elements of this data-driven culture, bring it all together for that. And like I said, really exemplified in my mind about this idea, do you have data to back that up?

[32:58](https://youtu.be/_D2qL0P7rEg?t=1978) The question anyone would be asking, commissioner on down, you know, and where is it and how do we know it? So, and yeah, great to see the behavioral insights team is going to be talking with you all. This is really great, a really great group of people. And again, doing this kind of work and how can we kind of help encourage and advocate for kind of the best, better things in government service and with our residents or people we serve. So I want to come back to this idea of what does this culture look like? How are we doing cultivating this thinking with data approach to what we're doing? So some key beliefs, again, beliefs, values, norms. What are some of these beliefs? That data is a useful guide and should be used whenever possible. You know, reaching for the data whenever possible. Everyone can use data, should have access to tools and information they need to be successful. Not just assuming that someone will, they don't know what they're doing, so I'm not going to give them the data. Okay, well, let's, how can we support them? How can we mentor them? How can we show them best practices? How can we make sure those best practices are followed in order for us to achieve what we're trying to achieve our goal?

[34:03](https://youtu.be/_D2qL0P7rEg?t=2043) And realize no one knows everything. Even data doesn't know anything, right? Data never changed anything or did anything, or it doesn't answer our questions just, you know, by itself. How can we come together and do this work better together? We, in our classes, we talk about don't data alone. Data analytics is a team sport, right? We come at this together and we use data together, much like innovation. You can't, there's no innovator who works on their own. It's always with other people together, kind of collaborating, whatever arrangement, whatever works best for them. Some of the key values when we talk about this, a single source of reliable truth. So that it's never the question about, oh, well, you know, my version of this data says this. Or, you know, oh, did you know that James over here, he has the data and it's only on his machine and we can't get access to it. So we can't answer that question. It's like having a solid, a single source of truth is essential to this. And then managing that like anything else you would manage.

[35:05](https://youtu.be/_D2qL0P7rEg?t=2105) It's an asset for the organization. If data is driving our strategy, if it's driving what we're doing, then it's just as important as, you know, the machines we're using or the buildings that we own or whatever it is that we're doing with this work. Valuing tested hypotheses over untested assumptions. It's not to say that assumptions aren't helpful. They definitely are. When I was talking about this when I was deployed, our assumptions were really helpful, but then being able to take it the next step and let's test this. And it's really important to discern the difference. Is this an assumption or a tested hypothesis? Do we actually know this is true? Carl talks about the hippo, the highest paid person's opinion. And it's one of the things that can kind of rule when we don't have facts, we don't have evidence. It's just, well, whatever the highest paid person's opinion is. Well, you know, commissioner thinks this, this must be the truth. This is the mayor thinks that or governor or whoever's in charge of this thinks this, this must be the truth. And a value is a constant improvement in how things are measured, evaluated, communicated. Almost always you're going to need to iterate.

[36:05](https://youtu.be/_D2qL0P7rEg?t=2165) You know, a metric is going to be maybe initially helpful, but you've got to change. You've got to improve on it as you go forward. It's very important. But knowing it is that organic process as you move forward. Some of the norms, I think, are really important. What's the data tell us on this? This gets to something I think is very important, which is holding our assumptions a little bit more loosely. Being able to know when, and again, it goes back to what we're talking about. Some of these leaders in our organizations that have been doing the work for so long. You know, they just kind of instantly go for, well, this is the answer. This is the answer when I did this work. You know, let's say that's not true, but having a little bit more space to say like, well, this is what I knew to be true back in the day. But what's the data show me about now? What are some of the things we can explore and kind of look at this about and understand? Offering up data, even when it's not explicitly asked for. So this is, again, a way to kind of like manage up or manage across your organization. Where the question may be, just tell me X.

[37:09](https://youtu.be/_D2qL0P7rEg?t=2229) And being able to add in some more context. Well, in order to answer X, here are some things to know. Here's Y and Z to also know. And I'm going to give you some context and add this extra data to it. So that the context is there. And I think that is a good norm to get into. Getting the data, pulling it down, working with it. Again, it's problematic when you work in silos and the data is not easily available. And you've got to call through, you know, three months worth of chat logs in order to do a performance analysis. But, you know, it's offering that in there and kind of encouraging this extra information in there. One of the very interesting things, talking about this, that comes to my mind. Those of you who are familiar with New York City and everything, you may not be appreciating all my New York City stories. But it is some interesting kind of test cases for a lot of these things. After 12 years of Mayor Bloomberg, who was very data driven and always when trying to advocate for a policy in front of city council. People would come with all the data about it and why they made decisions and all these kinds of things. I was there for the transition to the de Blasio administration.

[38:11](https://youtu.be/_D2qL0P7rEg?t=2291) And what was interesting is with the change in personnel, they didn't quite have that same approach to it. And some of the first meetings with council where they were trying to advocate for these proposals. It was interesting because council was just like, okay, where's the data? Where's all these kinds of things? You've done this analysis, right? You've done this and that. And it was like, oh, oops, we haven't. We didn't do that. It was a really interesting moment where you saw how the institution of council had been so habituated to seeing the data. When it was absent, it was noticeable. And I think that's also a good way to look at this is like when you keep offering this up and keep offering it. When it suddenly goes away, it's likely it'll be noticed. And that's a good thing. Oh, yeah, I really liked having all that extra data. Let's keep that in there. So that's a kind of a norm I think is really good. And we talk about a data driven culture. Deferring judgment until the data is ready again, if possible. There's always that, it's never a hundred percent certain. You don't want paralysis by analysis. But really encouraging.

[39:12](https://youtu.be/_D2qL0P7rEg?t=2352) It's like, I'm not sure we really quite understand this fully. And I'm sure every one of us here could offer an example of when something was done without the right data or enough data. Or something that could have waited just a little bit longer, would have known a little bit better what the right or a better course of action would have been. And again, getting to this idea of the discomfort when we're making decisions without data. Sometimes we have to. We just, we have to. But having that be an uncomfortable situation. And then in getting to this idea, right, of getting leadership on board, everything else is then helping encourage everyone in the organization to understand what it means to then devote resources to this. What does it mean to actually take the time to get our data systems in place and collect the data and all these kinds of things. So, that's Richard's vision of, you know, a data driven culture, a data driven world. What it means to think with data and creating some of these things. I really want to take this in the moments that we have left and have you take this into our Q&A as well about this kind of work.

[40:13](https://youtu.be/_D2qL0P7rEg?t=2413) About why isn't this easier to do? Data has a better idea, right? It's one of those little taglines. It seems so easy, I think, for a lot of us who understand the benefits, who have experienced the benefits very viscerally, to see it as like a moral imperative. This is just what you should be doing. But not everyone in our organizations typically have that perspective. And it's helpful to have some thoughts about that. So, I want to share with you here. I'm happy, again, to take some more comments or thoughts on this in our Q&A or in the chat now. I'll start with the line staff. In my experience, when you talk about the line analysts, people that are out there doing it, much of their perceived value is in their skill and knowledge. They really feel that this work that they're doing defines them, defines their value to the organization. And their judgment is relied on for sound decision making. They're the people that they go to for answers. They often work in silos and are respected for the work they do, no matter how inefficient.

[41:15](https://youtu.be/_D2qL0P7rEg?t=2475) I was working with New York State Department of Education on the Uniflow Pre-K, which became Pre-K for All. The Early Childhood Education Department had one person doing all their data analysis. I mean, this is a school district of 1.1 million students, right? Their Early Childhood Education alone is larger than many school districts, if you think about that. And there is one person, it was a dude, a guy, I forget his name, but he was doing all their data work. Very much in a silo, very much like this was his bread and butter. He got a lot of respect for it. I'm sure he's a wonderful person, but that created some dynamics that were challenging. Because, well, what if he gets hit by a bus or goes in sick or wants to go and do other things, right? There's so much knowledge that's locked in his head. And then there's the question about, is this the best way maybe of doing this work? And it's really hard for him to let that go. It's a loss of prestige. It's a loss of his value to the organization. You're trying to push me out. What are you doing here? It becomes really problematic.

[42:16](https://youtu.be/_D2qL0P7rEg?t=2536) And also, it's important to understand people drawn to government. This is really important innovation work. People drawn to government work often do so because they value continuity and consistency. Now, we can do a psychological profile. We can break that down. Why? And their history, biography, whatever. But I think it's important to understand that change, or at least a perception of change, becomes very disruptive. And really hard for some people to really embrace. As much as people maybe still say, this is what you're going to do. It's going to be so much better. It's really hard for them to get on board with that. When they feel more of the loss than they see the gain. So that's really important to acknowledge. They can feel really pushed aside in these kinds of things. So that's one thing blockers that I see to this kind of work. And then we talk about leadership. Again, leaders retain their position. They were highly skilled and knowledgeable. Their judgment could be relied on for sound decision making. They were rewarded for knowing all the answers. They were the smartest person in the room. And they like being in control.

[43:16](https://youtu.be/_D2qL0P7rEg?t=2596) There's a lot of responsibility to being in charge, as some of you know. Because you're in control of things right now that are maybe going off the rails. You've got to get them back on, whatever. There is that aspect to it. Data comes in here. And they can feel like they're losing control. You're questioning me. You're not taking my opinion. You don't trust me. You want to use this machine learning algorithm to tell me something that I already know. It becomes really problematic. Some new way of doing things. I think that's changing. And that's great. It's more where people really start to embrace what innovation, what data and technology can provide. But there's still that aspect of negotiating the loss of control. And how do we say, this is a tool. This is help. It's going to give us better views into things than you may be getting right now. So those are kind of two aspects to it I kind of wanted to address. I want to talk too about some thoughts on how to manage this change. What gets people thinking? How gets an organization on board with a more data-driven culture?

[44:18](https://youtu.be/_D2qL0P7rEg?t=2658) I'm going to start off with some things not to do. I don't see probably Wins Friends and Influences people. Start the conversation telling colleagues what they should or have to do. Especially to leadership. It's like, hey, you should be doing this. They probably should. But that may not be the best way to frame the conversation. I think it's better to tell them what data will enable. And definitely show, don't tell. Having some examples to really kind of make it clear. Hey, you know this horrible experience we had with the downed trees? Here's a better way of doing this. And I noticed someone actually posting there about some of the post that storm. Some of the work's been done by Parks. And yes, it's their IPM team, Innovation and Performance Management team. Dashboarding things. So I think that's a good way to start. I've seen this kind of work, you know, in so many examples of this, you know, you talk about teachers, you talk about law enforcement, firefighters, paramedics, you know, even, you know, construction workers, people out in the trades and things like that. You know, we want their data, we want them to enter all this data in for us. No one in these fields got into it because they wanted to do data entry. That was not what they wanted to do. You know, teachers want to teach, law enforcement wants to get out there and patrol, do their job. Firefighters want to put out fires.

[45:28](https://youtu.be/_D2qL0P7rEg?t=2728) There's a great book on this called The Fires. It talks about some of the work in the 60s with New York City Fire Department. You know, and they made the mistake, let's measure everything. And people were resistant to it because it gets in the way of what they really want to be doing in their job. So I think it's very important to kind of, you know, this will just lead to exhaustion if you aren't measuring what matters. So I think it's important to kind of measure everything. And people were resistant to it because it gets in the way of what they really want to be doing in their job. So I think it's very important to kind of, you know, this will just lead to exhaustion if you aren't measuring what matters. So identify what matters and really help them understand the value of it. And if possible, automate it, have proxies, do other kinds of things so that you're getting the data that you need without necessarily overburdening the people that are collecting it. With the work I've done around this kind of thing, I'm always trying to like, how does this, this data that we need, how can we make it useful for you so that you can make best use of this? That's really important. Another thing not to do, don't promise them everything will be solved with data. It's your silver bullet. It'll do everything. It won't. And then when it doesn't, they'll come blaming you or whoever told them this would be the case, right?

[46:44](https://youtu.be/_D2qL0P7rEg?t=2804) It's just the nature, especially if you've got vendors out there in the space, not to mention any names, IBM, who are saying that like their data system solutions will solve every problem the city could ever possibly have. It won't. And then when it doesn't, they'll have even less support for being data driven, right? So it's being, I would call it humility, having humility about what data can do and can't do, and really being clear, like, this is not going to answer everything. When I was a government analyst, I was working with the deployed soldiers, I briefed a general who had command over a regional command in Afghanistan, and was briefing on some techniques and things we were doing and some opportunities in this battle space. And he looks at me like, wait, what should I do? Like, you're the one in charge, or you're the general, not me. Like, I'm not the person to tell you what to do, sir. I can help you make that decision. I can give you the information you need to make better decisions, but it's your responsibility to do this. It's not just going to be like, go do this and make that simple for you. So it's really important to message that, manage the expectations, but really, you know, under promise, over deliver kind of thing, I think is really a successful strategy. Many of you probably have experienced this in your work.

[48:00](https://youtu.be/_D2qL0P7rEg?t=2880) Yeah, Christine, thank you for that. Yeah, you collect data, what do you need it for? If you don't need it, don't collect it, right? That's also the privacy piece, that's everything else in this as well to be very mindful of. And it hurts me as an analyst, I want all the data, I want to swim in data and just have everything possible, because I don't know, it's like, I've got to be more thoughtful in government, especially about, okay, making the case for it and being iterative and testing, this is why I need it and going through it. So my thoughts about getting buy-in, I've already mentioned some of these, start small, set a good example, have tangible results, like show, don't tell, make sure things are really kind of clear. And the value of being more data-driven should be self-evident and speak for itself. If you're not familiar, careful of my story, so we'll go over, I have too many stories. If you're familiar with the Peak Academy in Denver, really great example of this kind of piece. The Peak Academy was set up to bring Lean Six Sigma into Denver government. It's a unified city-county government, and they're going through budget cuts, and they were trying to do this, you know, do more with less.

[49:10](https://youtu.be/_D2qL0P7rEg?t=2950) And Brian Elms talks about this book, Peak Performance. The DMV was having this horrible problem, as most DMVs do, right, service delivery and four-hour waits and all this kind of stuff. So the answer, right, was like, oh, we'll put an ice cream machine in the lobby, and that'll make people happier, because we're getting deemed with these, like, satisfaction surveys and all this kind of stuff. Yeah, okay, it gets, you could say that was data-informed, maybe, you know, this might help, we'll test it and see. The better option was, you know, the students that they were trained through the Academy who worked in the DMV did their analysis and looked at, okay, where are the bottlenecks? What's happening here? What can we do to improve this? And they identified, you know, why people were waiting so long, they were able to streamline the process, they were able to measure everything about it, and, you know, actually create an experience where people were in and out, and I think it was like 20 minutes or something, like, how do you get people to actually do better, actually do the better service and improve the metric that you really want?

[50:10](https://youtu.be/_D2qL0P7rEg?t=3010) That was a clear-cut case of everybody, you know, of using the data to be, you know, in a small way, like, just kind of fix the line at this DMV, but being able to have results then that they could show that advocate for why this process was of benefit to the city, more so than having an ice cream machine out in the lobby. So, and we should be able to show that a data-driven approach is more efficient, more reliable, and provides repeatable results towards outcomes. Another story about this that I really love is there was an experience with a foreman who was heading a parks firm, not in New York City, another city, and he, in this meeting where they're trying to help him be more data-driven and think about this, he just kind of gets really frustrated and says, I can do things, or I can measure things, but I can't do both. You didn't use the word things, but the word isn't to be used in polite company. So, and to the point, right? It's like when you're delivering things, it's really hard to measure it. Fast forward a few months, and there was an incident of a rise in crime in the parks in the city.

[51:18](https://youtu.be/_D2qL0P7rEg?t=3078) And so, he and his team said, okay, we need six more rangers to deal with this crime. Okay, what are six rangers going to get us? Was the question that came back to him. What are six rangers going to get? Why six? Why not four? Why not eight? Are we going to, with 12, would we eliminate crime entirely in the parks? Like, what is this? And that really, you know, talks about your people that were in the meeting talk about like, you know, that was kind of his light bulb moment where it's like, oh, you know, measuring things is an important aspect of doing things. These have to be kind of in tandem. Because, you know, someone famous once said, you can't manage what you can't measure. Right? If you can't, you don't know what you're doing and why you're doing it and what the outcomes are, it's really hard to make these decisions. And that's where things become just arbitrary. Oh, well, six just sounded good. Well, in our research and training environment, it's really hard to justify that anymore. Like, oh, yeah, just 10. 10 will solve our problem. It's like, why 10? And what if I gave you 12? You know, it's in, we have this problem here in Dallas, all these cities, all these municipalities are with this problem, right? What are the outcomes we're actually getting for what we're doing?

[52:23](https://youtu.be/_D2qL0P7rEg?t=3143) So this is where a data driven approach, I can really be able to demonstrate that, you know, it becomes really important. And if we can't show results, then we may not be doing it right ourselves. And this is really important. You know, trying to hand wave this or anything else, because again, that undermines the case that we can and should be making. So just some some thoughts here, you know, have patience, have empathy. Know this work isn't easy, but it is important. And I want to give you a thought about why this is important. For me, I believe centering data more in our conversations is what can help us towards creating a more fair, equitable and just society in which everyone benefits. So this is really, I think, why we do this. It's to take that, not to say we shouldn't be subjective or, you know, objectivity is a thing, but really putting more evidence, more kind of reliable, less kind of human bias into this. Not that, you know, data has its biases, but when we rely just on opinions, I think this quote from Jim Barksdale really resonates with me.

[53:23](https://youtu.be/_D2qL0P7rEg?t=3203) If we have data, let's look at data. If we have opinions, if all we have are opinions, let's go with mine. And that's what becomes like it's just the opinion of the highest paid person, the loudest voice, the squeakiest wheel, whatever it is. And that's where things can become become really problematic when we talk about fairness, equity, justice in our societies. With that, thank you very much. I really do appreciate it. Please feel free to contact me. I have my email information. Happy to connect on LinkedIn.

* * *

Original page: https://innovate-us.org/thinking-with-data-leveraging-analytics-to-better-define-and-solve-problems · Licence: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/), as stated on innovate-us.org
