<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Plot + Scatter</title>
    <description>Data for human consumption.
</description>
    <link>http://plotandscatter.com/</link>
    <atom:link href="http://plotandscatter.com/feed.xml" rel="self" type="application/rss+xml"/>
    <pubDate>Tue, 18 Jun 2024 05:28:41 +0000</pubDate>
    <lastBuildDate>Tue, 18 Jun 2024 05:28:41 +0000</lastBuildDate>
    <generator>Jekyll v3.9.5</generator>
    
      <item>
        <title>A Voronoi-based exploration of transit times</title>
        <description>&lt;p&gt;&lt;em&gt;Heather Armstrong did an internship at Plot + Scatter in fall 2016. Here she describes the capstone project she completed as part of a course she did at the Vancouver Institute of Visual Analytics. Heather offers valuable insight into some of the design decisions and challenges she faced when building her project.&lt;/em&gt;&lt;/p&gt;

&lt;!--more--&gt;

&lt;h1 id=&quot;background&quot;&gt;Background&lt;/h1&gt;

&lt;p&gt;I completed the &lt;a href=&quot;http://plotandscatter.com/transit-times&quot; target=&quot;_blank&quot;&gt;Transit Voronoi Explorer&lt;/a&gt; in 2016 for a course I was taking at the Vancouver Institute of Visual Analytics. The aim was to build an interactive map with an overlay of transit times to downtown Vancouver across the Metro Vancouver area. I hoped this could give insight into each neighbourhood’s accessibility to public transport.&lt;/p&gt;

&lt;p&gt;I envisioned the tool helping transit planners with the optimization of transit routes. Or assist new home buyers to make informed decisions when deciding where to live. It would specifically help target and minimize &lt;a href=&quot;https://transportist.org/2016/10/06/what-do-we-know-about-the-first-milelast-mile-problem-for-transit/&quot; target=&quot;_blank&quot;&gt;last-mile travel&lt;/a&gt; times, encouraging increased connectivity to transit across the Lower Mainland.&lt;/p&gt;

&lt;h2 id=&quot;initial-research&quot;&gt;Initial research&lt;/h2&gt;

&lt;p&gt;I found the &lt;a href=&quot;http://tenyearvision.translink.ca/downloads/10%20Year%20Vision%20for%20Metro%20Vancouver%20Transit%20and%20Transportation.pdf&quot; target=&quot;_blank&quot;&gt;Translink 10-Year-Plan&lt;/a&gt;, while doing some initial research.  The plan consisted of creating new B-Lines and rapid transit lines, which I later added as an overlay to the map.&lt;/p&gt;

&lt;p&gt;I contacted TransLink to inquire about other useful resources they might have.
They responded with several links and said a data-driven project could be valuable to help validate the plan. The 10-year plan includes some of the biggest changes ever to be made to Vancouver’s transit routes. And at the time was the only plan based on prior passenger flow data.&lt;/p&gt;

&lt;h2 id=&quot;1-data-acquisition&quot;&gt;1. Data acquisition&lt;/h2&gt;

&lt;p&gt;After identifying the visualization’s purpose and design, I gathered coordinates for the travel time data. The City of Vancouver had a shapefile which included coordinates of all city street intersection. But I had to contact other cities in the Metro Vancouver area and inquire for similar datasets. Sadly, most didn’t have any available. This highlighted how difficult sourcing data can be.&lt;/p&gt;

&lt;p&gt;Being a UBC student was a great help in this case though. I used the university’s GIS database to locate a CanMap Content Suite, and found one that contained a shapefile for every intersection in Canada. I then cropped specific areas and reduced the point density with QGIS. Finally, I had a set of coordinates that I could use to get travel time data. But I now needed to find public transportation data.&lt;/p&gt;

&lt;p&gt;Selecting an API to collect data was more complicated than I thought. Many APIs either lacked public transportation data, or weren’t free. I settled on the &lt;a href=&quot;https://developers.google.com/maps/documentation/distance-matrix/&quot; target=&quot;blank&quot;&gt;Google Maps Distance Matrix API&lt;/a&gt; for its ease of use.&lt;/p&gt;

&lt;p&gt;Google Maps charges  a fee though based on the number of data points you collect. The travel time for a coordinate at a given time would be considered one data point. The problem was I could only access 2,500 free data points per day. Since I needed data for about 25,000 coordinates, cost became a limiting factor. I decided to only gather data for a snapshot in time (9AM on a Wednesday morning during rush hour).&lt;/p&gt;

&lt;h2 id=&quot;2-visualization-tool-planning&quot;&gt;2. Visualization tool planning&lt;/h2&gt;

&lt;p&gt;The next step was to plan the visualization. I had several use cases, some rough UI sketches, and an idea of how I would build my project. Yet, I still needed to select a tool or framework to use.  The &lt;a href=&quot;https://developers.google.com/maps/&quot; target=&quot;blank&quot;&gt;Google Maps API&lt;/a&gt; requires that all data collected be displayed on a Google Map. So I needed a framework with a Google Maps base-map. It had to be flexible enough to create customized SVG overlays. I found that, &lt;a href=&quot;http://leafletjs.com/&quot; target=&quot;blank&quot;&gt;Leaflet&lt;/a&gt; did the job.&lt;/p&gt;

&lt;h2 id=&quot;3-building-the-map-visualization&quot;&gt;3. Building the map visualization&lt;/h2&gt;

&lt;p&gt;The process of building the visualization was challenging. I had little experience with JavaScript. Having API documentation, tutorials, and visualization examples were key.&lt;/p&gt;

&lt;h3 id=&quot;challenge-1-using-google-maps-base-map&quot;&gt;Challenge 1: Using Google Maps base map&lt;/h3&gt;

&lt;p&gt;Leaflet has a plugin section with many different base-map providers. This allows you to use almost any map-tiles with little to no configuration. One plugin allowed you to use Google Maps tiles as a base-map,  &lt;a href=&quot;https://gitlab.com/IvanSanchez/Leaflet.GridLayer.GoogleMutant&quot; target=&quot;_blank&quot;&gt;GoogleMutant&lt;/a&gt;. But it had been created only two months earlier. So documentation was limited and the plugin generated many bugs in the project.&lt;/p&gt;

&lt;p&gt;I did consider several other options. At the time, &lt;a href=&quot;https://www.mapbox.com/&quot; target=&quot;blank&quot;&gt;Mapbox&lt;/a&gt; didn’t provide transit data through their API though. &lt;a href=&quot;https://mapzen.com/&quot; target=&quot;blank&quot;&gt;MapZen&lt;/a&gt; did, but it required that you to know each location’s stop ID. I finally managed to use GoogleMutant &lt;a href=&quot;https://gitlab.com/IvanSanchez/Leaflet.GridLayer.GoogleMutant&quot; target=&quot;_blank&quot;&gt;GoogleMutant&lt;/a&gt;  after some debugging,&lt;/p&gt;

&lt;h3 id=&quot;challenge-2-json-api-data-post-processing&quot;&gt;Challenge 2: JSON API data post-processing&lt;/h3&gt;

&lt;p&gt;Once the base-map was running, I need an appropriate way to format the JSON data so that I could create a customized SVG overlay. To do so, I extracted the latitude and longitude coordinates and the travel times from the JSON. I changed all time data to total number of minutes by writing a post-processing script in Python.&lt;/p&gt;

&lt;h3 id=&quot;challenge-3-visualizing-point-based-time-data&quot;&gt;Challenge 3: Visualizing point-based time data&lt;/h3&gt;

&lt;p&gt;I collected point-based time data for the selected intersections within each city. But I needed to visualize the area around each specific coordinate.&lt;/p&gt;

&lt;h4 id=&quot;the-basic-svg-circle&quot;&gt;The basic SVG-circle&lt;/h4&gt;

&lt;p&gt;Initially, I used equally-sized SVG circles to test my colour scale before shading the polygons. Each circle represented one coordinate. I shaded them based on their corresponding time interval.&lt;/p&gt;

&lt;p&gt;&lt;img src=&quot;/assets/images/projects/point-based-map.png&quot; alt=&quot;point-based map of Vancouver&quot; /&gt;
This helped confirm that my method fulfilled the purpose of my project. It also ensured that I was able to capture areas where it took longer to get downtown via transit.&lt;/p&gt;

&lt;h4 id=&quot;moving-to-a-voronoi-diagram&quot;&gt;Moving to a Voronoi diagram&lt;/h4&gt;

&lt;p&gt;Once I validated the idea, I then had to find a way to create an layer composed of polygons for each point. But I could not use a regular polygon. This was because the coordinates were not evenly distributed. So there would still be spaces between each polygon.&lt;/p&gt;

&lt;p&gt;I decided to use a &lt;a href=&quot;https://en.wikipedia.org/wiki/Voronoi_diagram&quot; target=&quot;_blank&quot;&gt;Voronoi diagram&lt;/a&gt; instead as an overlay. A Voronoi diagram creates polygons across a plane so that each polygon contains only one point. It maximizes its area without encroaching into another point’s space.&lt;/p&gt;

&lt;p&gt;&lt;img src=&quot;/assets/images/projects/full-voronoi-map.png&quot; alt=&quot;Voronoi-based map of Vancouver&quot; /&gt;
Leaflet happens to have its own pre-defined function for creating a Voronoi diagram. It’s based on a set of points, but I had having trouble creating one using Leaflet’s &lt;code class=&quot;language-plaintext highlighter-rouge&quot;&gt;d3.geom.voronoi()&lt;/code&gt; function. So I followed &lt;a href=&quot;https://chriszetter.com/blog/2014/06/15/building-a-voronoi-map-with-d3-and-leaflet/&quot; target=&quot;_blank&quot;&gt;Chris Zetter’s tutorial&lt;/a&gt;, which was extremely helpful.&lt;/p&gt;

&lt;h3 id=&quot;challenge-4-speeding-up-the-load-time&quot;&gt;Challenge 4: Speeding up the load time&lt;/h3&gt;

&lt;p&gt;The original set of coordinates contained a very large number of data points to process every time the map reloaded or moved around. This meant it took  a while to redraw the Voronoi diagram each time. To reduce the load time and have a better user experience, I took a random sample of the data using QGIS instead.&lt;/p&gt;

&lt;p&gt;This hardly made a difference to the accuracy of the visualization at a high zoom level. But when a user would zoom in too much on the map, the data points no longer gave an accurate representation of travel times.&lt;/p&gt;

&lt;p&gt;To to solve this issue, I thought of switching datasets at certain zoom levels. This meant that when a user zoomed in at city block level for example, the Voronoi function would draw the diagram with finer data.&lt;/p&gt;

&lt;p&gt;But I was unable to get this aspect of the map working in time for my project, so the map only currently uses the random sample dataset. I am still hoping to implement this functionality at a later date.&lt;/p&gt;

&lt;h3 id=&quot;challenge-5-containing-the-voronoi-diagram&quot;&gt;Challenge 5: Containing the Voronoi diagram&lt;/h3&gt;

&lt;p&gt;The Voronoi diagram presented some other challenges. A Voronoi diagram doesn’t have any constrained boundaries. So it will spread outwards infinitely. As a result, the areas of the outer most polygons are infinite. This makes for a messy visualization when a user tries to navigate outside of the main map area.&lt;/p&gt;

&lt;p&gt;To ensure the Voronoi diagram was contained, I had to limit the minimum zoom level. I also made the map bounce back to a set of defined bounds when a user tries to drag the map out too far.&lt;/p&gt;

&lt;h3 id=&quot;challenge-6-creating-a-colour-scale-that-shows-the-differences-in-travel-time&quot;&gt;Challenge 6: Creating a colour scale that shows the differences in travel time&lt;/h3&gt;

&lt;p&gt;Creating a colour scale that allowed users to see both small and big differences in travel time was difficult. I couldn’t just use a linear scale.&lt;/p&gt;

&lt;p&gt;&lt;img src=&quot;/assets/images/projects/colorbrewer.png&quot; alt=&quot;Colorbrewer diverging, sequential and&quot; /&gt; Image: &lt;a href=&quot;http://www.gnuplotting.org/tag/palette/&quot; target=&quot;blank&quot;&gt;Gnuplotting&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;First, I tried using a single-hue sequential scale. But it was hard to distinguish accessible areas from remote ones. The diverging scale made it somewhat easier to identify areas that lacked transit access.&lt;/p&gt;

&lt;p&gt;The travel times started at 0 minutes and went all the way to over 200 minutes. To capture this range of time, I varied the intervals. The time scale started with at 5 minutes, and went up to 40 minutes. Times increased in a logarithmic-like manner.&lt;/p&gt;

&lt;h3 id=&quot;challenge-7-integrating-translinks-10-year-plan&quot;&gt;Challenge 7: Integrating TransLink’s 10-Year-Plan&lt;/h3&gt;
&lt;p&gt;The last challenge was to see how well the new planned routes targeted underserved areas. I thought to overlay the routes on top of the Voronoi diagram. This proved to be more tedious than difficult.&lt;/p&gt;

&lt;p&gt;I ended up creating a new overlay on the map, and plotted the planned B-Lines and rapid transit routes by hand. I then used SVG polylines to create paths out of them. This layer allowed users to see how the new routes would improve travel commute to downtown Vancouver.&lt;/p&gt;

&lt;h1 id=&quot;results&quot;&gt;Results&lt;/h1&gt;
&lt;p&gt;Overall, the &lt;a href=&quot;http://plotandscatter.com/transit-times&quot; target=&quot;_blank&quot;&gt;completed visualization&lt;/a&gt; met my goals.I noticed that most skytrain stations were surrounded by a blue or green polygon. This indicates faster transit times to get downtown. Stations farther out will have a different colour polygon as travel time increases. This make intuitive sense.&lt;/p&gt;

&lt;p&gt;&lt;a href=&quot;http://plotandscatter.com/transit-times&quot; target=&quot;_blank&quot;&gt;&lt;img src=&quot;/assets/images/projects/b-lines.png&quot; alt=&quot;voronoi map with Translink B-lines&quot; /&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The focus on increasing flow in more highly traveled corridors&lt;/strong&gt; becomes obvious when the new B-lines and new train routes are layered on the map. Translink seems to be addressing the high flow of traffic to universities, by implementing new B-Lines to both SFU and UBC.&lt;/p&gt;

&lt;p&gt;The 10 year plan also includes a new rapid transit route which connects VCC-Clark to the Arbutus and Broadway areas. Surrey, Langley, and Coquitlam all gain rapid transit routes as well. This increases the ease of commuting to downtown Vancouver.&lt;/p&gt;

&lt;p&gt;Parts of New Westminster, Richmond, Port Coquitlam, and White Rock do not seem to benefit from new transit routes though. The center of each municipality appears to be well connected, but &lt;strong&gt;most people do not live there&lt;/strong&gt;.&lt;/p&gt;

&lt;p&gt;The lack of transit connections to outskirt neighbourhoods increases last-mile travel times, or people’s commute times to downtown Vancouver. Unfortunately, the data and map don’t capture population distribution in these areas. This may be a factor for not developing transit routes in these areas.&lt;/p&gt;

&lt;h1 id=&quot;limitations&quot;&gt;Limitations&lt;/h1&gt;

&lt;p&gt;This visualization has several limitations. I have outlined them throughout this text, but to summarize:&lt;/p&gt;

&lt;ul&gt;
  &lt;li&gt;The data visualized is only a snap shot in time at 9AM rush hour on a Wednesday morning. It is likely that the Voronoi overlay would change depending on the day of week and time of day. (I realized afterwards that 9am might even be too late to capture the morning rush.)&lt;/li&gt;
  &lt;li&gt;The data used is currently a random sample, hence accuracy decreases as a user zooms in on the map.&lt;/li&gt;
  &lt;li&gt;There are still several areas to add in Metro Vancouver (e.g. Maple Ridge and Langley).&lt;/li&gt;
  &lt;li&gt;The map doesn’t capture the number of people traveling to and from places. Nor does it show which are the most popular routes and stops.&lt;/li&gt;
&lt;/ul&gt;

&lt;h1 id=&quot;future-improvements&quot;&gt;Future improvements&lt;/h1&gt;

&lt;p&gt;A few improvements could be made for a more in-depth and accurate analysis of the TransLink plan. Some are more feasible than others:&lt;/p&gt;

&lt;ul&gt;
  &lt;li&gt;Create another layer to highlight inaccessible areas that need to be addressed.&lt;/li&gt;
  &lt;li&gt;Increase point density and accuracy as the user zooms in.&lt;/li&gt;
  &lt;li&gt;Visualize data from different times and days of the week.&lt;/li&gt;
  &lt;li&gt;Add all municipalities in the Metro Vancouver area.&lt;/li&gt;
  &lt;li&gt;Show the movement of people to and from downtown Vancouver, as well as most popular routes.&lt;/li&gt;
  &lt;li&gt;Be able to change destination points.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Still, this visualization is a good initial step. A bigger project in the future could include other features and aspects of the transit plan.&lt;/p&gt;
</description>
        <pubDate>Mon, 02 Oct 2017 00:00:00 +0000</pubDate>
        <link>http://plotandscatter.com/ideas/2017/10/02/a-voronoi-based-exploration-of-transit-times.html</link>
        <guid isPermaLink="true">http://plotandscatter.com/ideas/2017/10/02/a-voronoi-based-exploration-of-transit-times.html</guid>
        
        
        <category>data</category>
        
        <category>project</category>
        
        <category>transit</category>
        
        <category>map</category>
        
      </item>
    
      <item>
        <title>Sprinting through data projects</title>
        <description>&lt;p&gt;&lt;em&gt;In our last post, we talked about &lt;a href=&quot;/ideas/2017/09/27/tearing-down-the-data-silos.html&quot;&gt;tearing down data silos through collaboration&lt;/a&gt;. In this post, we look at what Lean and Agile UX methodologies can teach us about sprinting through our data projects.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;As we mentioned in the &lt;a href=&quot;/ideas/2017/09/26/start-goobing-with-data.html&quot;&gt;first article of this series&lt;/a&gt;, the three most common mistakes we see people making with their data projects are:&lt;/p&gt;

&lt;!--more--&gt;

&lt;ol&gt;
  &lt;li&gt;Setting unrealistic goals because they’ve misunderstood the drivers of their target user / audience / market.&lt;/li&gt;
  &lt;li&gt;Underestimating the extent to which local conditions and environments determine how, where, and why their target audiences behave in certain ways.&lt;/li&gt;
  &lt;li&gt;Ignoring how these can change across regions / organizations / departments / industries / socio-demographics.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;We’ve found that these can often be avoiding by applying &lt;a href=&quot;https://www.nngroup.com/articles/doing-ux-agile-world/&quot; target=&quot;_blank&quot;&gt;Agile&lt;/a&gt; and &lt;a href=&quot;https://www.interaction-design.org/literature/article/a-simple-introduction-to-lean-ux&quot; target=&quot;_blank&quot;&gt;Lean UX&lt;/a&gt; methodologies. One of the most useful methods is to integrate sprints into our data projects.&lt;/p&gt;

&lt;h1 id=&quot;avoid-big-and-costly-mistakes&quot;&gt;Avoid big and costly mistakes&lt;/h1&gt;

&lt;p&gt;While there are &lt;a href=&quot;https://www.justinmind.com/blog/choosing-a-ux-strategy-that-suits-you-and-your-users-agile-vs-lean-ux/&quot; target=&quot;_blank&quot;&gt;some differences between an Agile sprint and Lean iterations&lt;/a&gt;, the general idea remains the same: &lt;strong&gt;don’t wait till the end to find out if your project will fail or succeed&lt;/strong&gt;.&lt;/p&gt;

&lt;p&gt;Instead work in short iterative cycles or sprints. Fail early and catch mistakes before it’s too late.&lt;/p&gt;

&lt;p&gt;Agile and Lean UX methodologies encourage you to validate or teste your ideas often and incrementally. You then iterate upon them with the feedback you’ve received. Sprints or work cycles can be as short as one week or as long as four. It all depends on the nature of the project, team, and the organization.&lt;/p&gt;

&lt;p&gt;As Hilary Mason, founder of machine intelligence research firm Fast Forward Labs, &lt;a href=&quot;https://hbr.org/2017/07/how-ai-fits-into-your-data-science-team&quot; target=&quot;_blank&quot;&gt;points out&lt;/a&gt;, iteration is powerful because at the beginning of a data project it’s difficult to know if it will work or not. By nature, gleaning insights from data requires an exploratory approach.  &lt;strong&gt;“Data science requires having that cultural space to experiment and work on things that might fail,”&lt;/strong&gt; says Mason.&lt;/p&gt;

&lt;p&gt;Approaching data in an iterative way makes sense. By being  intentional, and always looking at where and why your data is failing you, you’ll learn to refine your data. You will can then make better decisions and take adequate action.&lt;/p&gt;

&lt;p&gt;This approach also works when you’re developing a data-driven product. For example, while working with &lt;a href=&quot;http://knomos.ca&quot; target=&quot;_blank&quot;&gt;Knomos&lt;/a&gt;, a legal knowledge management start-up, we integrated Lean UX sprints into the company’s workflow.&lt;/p&gt;

&lt;p&gt;Within two sprint cycles, we were able to help the company pivot their data-driven product. The user research and testing data we collected showed that their legal visual navigator didn’t meet their target audience’s needs. In less than four weeks, Knomos was able to refocus its product on a better audience, saving months of costly development work.&lt;/p&gt;

&lt;h1 id=&quot;fail-fast-and-rebound&quot;&gt;Fail fast and rebound&lt;/h1&gt;

&lt;p&gt;Sprinting or iterating on your data projects in short time-boxed increments will give your team the necessary time to explore your data while remaining focused. The goal should be to fail early rather than later. Identifying mistakes and deficiencies quickly will allow you to find solutions sooner.&lt;/p&gt;

&lt;p&gt;So avoid the temptation to create a single, intimidating deadline. For better results and less waste along the way, envision a series of energetic 5k runs, not a marathon.&lt;/p&gt;

&lt;!-- Read other articles from our data and UX blog post series:

1. [GOOBing with data](/ideas/2017/09/26/start-goobing-with-data.html) 
2. [Tearing down the data silos](link) --&gt;
</description>
        <pubDate>Thu, 28 Sep 2017 00:00:00 +0000</pubDate>
        <link>http://plotandscatter.com/ideas/2017/09/28/sprinting-through-data-projects.html</link>
        <guid isPermaLink="true">http://plotandscatter.com/ideas/2017/09/28/sprinting-through-data-projects.html</guid>
        
        
        <category>data</category>
        
        <category>agile</category>
        
        <category>ux</category>
        
      </item>
    
      <item>
        <title>Tearing down the data silos</title>
        <description>&lt;p&gt;&lt;em&gt;In our &lt;a href=&quot;/ideas/2017/09/26/start-goobing-with-data.html&quot;&gt;last post&lt;/a&gt; we talked about GOOBing (Getting out of the building) with your data. In this post, we explore what Lean and Agile UX methodologies can teach us about working together.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;One of the biggest challenges people face when trying to execute data projects is navigating departmental, organizational and team silos. This problem is more common than we think, and spans organizations and industries.&lt;/p&gt;

&lt;!--more--&gt;

&lt;p&gt;For a long time in the tech sector, project owners, designers and developers would work with a &lt;a href=&quot;https://en.wikipedia.org/wiki/Waterfall_model&quot;&gt;waterfall model&lt;/a&gt;. This involved a very linear and hand-off process. Each individual was responsible for one step without really understanding the impact of their work on the overall project. (This model was adopted from construction and manufacturing, and is still used in many industries today.)&lt;/p&gt;

&lt;p&gt;The waterfall model often resulted in very costly projects. Project owners and designers couldn’t see flaws until the project was finished. This meant more design and development would then be needed. The process also required stakeholders to write extensive instructions. These would be passed on to the stakeholder in the next phase – a very time consuming task, as you can imagine.&lt;/p&gt;

&lt;p&gt;Then came Agile and Lean User Experience methodologies. While one stems from the development side of tech and the other from design, both embrace similar values and principles that focus on:&lt;/p&gt;

&lt;ul&gt;
  &lt;li&gt;People (instead of documentation and processes)&lt;/li&gt;
  &lt;li&gt;Collaboration throughout teams and different roles (breaking down silos)&lt;/li&gt;
&lt;/ul&gt;

&lt;h1 id=&quot;working-collaboratively-or-tearing-down-the-data-silos&quot;&gt;Working collaboratively (or tearing down the data silos)&lt;/h1&gt;

&lt;p&gt;Lean and Agile UX methodologies have encouraged close collaboration between designers, developers, business execs and product owners. This ensures that a project evolves in alignment with everyone’s goals. It also reduces time and  wasted resources, while increasing the project’s likelihood of success.&lt;/p&gt;

&lt;p&gt;Remember, people should be at the centre of your process. You shouldn’t be spending more time documenting your efforts than actually working with your team. Otherwise you’re doing something wrong.&lt;/p&gt;

&lt;p&gt;It all starts with building the right team. Select people with different backgrounds and various perspectives from within the organization – otherwise you may overlook important aspects. Someone who works on the floor will bring a very different perspective than someone at the exec level, and vice versa.  Your team benefits from having both because you need buy-in from all levels of your organization.&lt;/p&gt;

&lt;p&gt;Also consider bringing someone from outside of your organization to be part of your team (either a third party consultant or the identified stakeholder/audience of your data project). If this is not possible, designate someone on your team as your audience champion. They will be responsible for reporting on your &lt;a href=&quot;/ideas/2017/09/26/start-goobing-with-data.html&quot;&gt;GOOB activities&lt;/a&gt; and should challenge any assumptions you are making.&lt;/p&gt;

&lt;h1 id=&quot;less-friction-and-better-project-outcomes&quot;&gt;Less friction and better project outcomes&lt;/h1&gt;

&lt;p&gt;Our clients have told us that navigating the silo culture within their organization is one of the greatest problems they face in working with data. Teams and departments have difficulty accessing and sharing data with each other. Corporate policies cause friction and delays in making the data available. They also create overly protective data gatekeepers. Inexperience and unknowns nourish fears that lead to ineffective data projects or inaction.&lt;/p&gt;

&lt;p&gt;Ultimately, data project goals should align with the organization’s business and strategic goals. Resources and knowledge should be shared across teams. The outcomes of data projects should also be shared in order to benefit everyone.&lt;/p&gt;

&lt;p&gt;This doesn’t mean everyone should be involved at every step of the project. But it does mean that everyone should be considered at some point (ideally at the beginning). Remember, people benefit from knowing the project’s progress.&lt;/p&gt;

&lt;p&gt;So next time your organization is planning a data project, take the time to set up a team with a range of invested stakeholders. It could mean the difference between a successful project and one that gets tangled up in organizational politics or red tape.&lt;/p&gt;

&lt;!-- _Up next: Sprinting through your data project. (link)_ --&gt;
</description>
        <pubDate>Wed, 27 Sep 2017 00:00:00 +0000</pubDate>
        <link>http://plotandscatter.com/ideas/2017/09/27/tearing-down-the-data-silos.html</link>
        <guid isPermaLink="true">http://plotandscatter.com/ideas/2017/09/27/tearing-down-the-data-silos.html</guid>
        
        
        <category>data</category>
        
        <category>agile</category>
        
        <category>ux</category>
        
      </item>
    
      <item>
        <title>Start GOOBing with data</title>
        <description>&lt;p&gt;At Plot + Scatter, we’ve seen our clients make the same three mistakes across all industries — big and small, corporate and non-profit. These errors cost time and money. In some cases, they lead to the realization that the entire project needs to be re-imagined. Not ideal.&lt;/p&gt;

&lt;p&gt;These common mistakes are:
&lt;!--more--&gt;&lt;/p&gt;

&lt;ol&gt;
  &lt;li&gt;Setting unrealistic goals because the project team has misunderstood their target user / audience / market.&lt;/li&gt;
  &lt;li&gt;Underestimating the extent to which local conditions and environments determine how, where, and why their target audiences behave in certain ways.&lt;/li&gt;
  &lt;li&gt;Ignoring how these can change across regions / organizations / departments / industries / socio-demographics.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;People tend to only look at basic or traditional indicators in their data projects. As a result, they miss great opportunities that would allow them to dive deeper into their data. Sometimes they totally miss the boat.&lt;/p&gt;

&lt;p&gt;To help our clients avoid making these mistakes, we’ve adopted some useful &lt;a href=&quot;https://www.nngroup.com/articles/doing-ux-agile-world/&quot; target=&quot;_blank&quot;&gt;Agile&lt;/a&gt; and &lt;a href=&quot;https://www.interaction-design.org/literature/article/a-simple-introduction-to-lean-ux&quot; target=&quot;_blank&quot;&gt;Lean User Experience (UX) Design&lt;/a&gt; methodologies. In this series, we’ll show you how they can apply to any data project.&lt;/p&gt;

&lt;h1 id=&quot;smart-methodologies-for-working-with-data&quot;&gt;Smart methodologies for working with data&lt;/h1&gt;

&lt;p&gt;The terms “Lean UX” and “Agile” are often heard in the tech world these days. More than buzzwords, they are processes and methodologies that have allowed the tech industry to evolve and innovate at a fast pace. Luckily, they can  be adapted and applied to industries and domains that go beyond the tech sector.&lt;/p&gt;

&lt;p&gt;While Agile stems from the development side of tech, Lean UX stems from design. Both embrace similar values and principles that focus on:&lt;/p&gt;

&lt;ul&gt;
  &lt;li&gt;people (instead of tools and processes)&lt;/li&gt;
  &lt;li&gt;iterative work (to minimize overall risk and adapt quickly to necessary changes)&lt;/li&gt;
  &lt;li&gt;collaboration throughout teams and different roles (breaking down silos)&lt;/li&gt;
  &lt;li&gt;shorter feedback loops (test or get feedback often, sooner rather than later).&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;(To fully understand the difference between Agile UX and Lean UX, check out this &lt;a href=&quot;https://www.justinmind.com/blog/choosing-a-ux-strategy-that-suits-you-and-your-users-agile-vs-lean-ux/&quot; target=&quot;_blank&quot;&gt;link.&lt;/a&gt;)&lt;/p&gt;

&lt;h2 id=&quot;so-how-does-this-apply-to-understanding-and-communicating-data&quot;&gt;So how does this apply to understanding and communicating data?&lt;/h2&gt;

&lt;p&gt;According to Don Norman, “guru of workable technology” and co-founder of the Nielsen Norman Group, user experience in general is defined as “…everything. The way you experience the world, the way you experience your life.”&lt;/p&gt;

&lt;p&gt;Data is exactly this: tidbits of numbers and information that show us on how people experience their lives and the world.&lt;/p&gt;

&lt;p&gt;Lean UX and Agile techniques are all about setting your assumptions aside and listening for multi-dimension answers instead of jumping to conclusions. It all starts with peering beyond the computer screen and into the world.&lt;/p&gt;

&lt;h1 id=&quot;get-outside-of-the-building-goob&quot;&gt;Get outside of the building (GOOB)&lt;/h1&gt;

&lt;p&gt;GOOB is a UX and product design exercise that should confirm assumptions that product teams in tech make about their projects . It works exactly as it sounds: you leave your desk to go collect relevant information. Then you confirm (or refute) your assumptions.&lt;/p&gt;

&lt;p&gt;This can be done in different ways with:&lt;/p&gt;

&lt;ul&gt;
  &lt;li&gt;Surveys&lt;/li&gt;
  &lt;li&gt;Stakeholder interviews&lt;/li&gt;
  &lt;li&gt;Ethnographic research&lt;/li&gt;
&lt;/ul&gt;

&lt;h2 id=&quot;the-perils-of-tunnel-vision-when-working-with-data&quot;&gt;The perils of tunnel vision when working with data&lt;/h2&gt;

&lt;p&gt;One of the worst mistakes people who work with data make is focusing only on the data immediately in front of them. Step away from your spreadsheets and computers. You will discover information you hadn’t considered before.&lt;/p&gt;

&lt;p&gt;A recent &lt;a href=&quot;https://hbr.org/2017/08/3-things-multinationals-dont-understand-about-africas-middle-class&quot; target=&quot;_blank&quot;&gt;article&lt;/a&gt; in the Harvard Business Review explains how multinational corporations (MNCs) failed to understand their targeted markets in Africa. In their example, MNCs only relied on GDP and demographic growth data to build their business strategies.&lt;/p&gt;

&lt;p&gt;Their approach to data didn’t capture how wealth trickles down through society. This led MNCs to overlook the informal economy. More than 50% of working adults in Africa earn income from activities not reflected in official statistics.&lt;/p&gt;

&lt;p&gt;Organizations tend to jump-start their data projects without building a foundation. They’re eager to gain insights without thinking about what they might do with them. They don’t consider how they will take action afterwards. And most importantly, why they want to do the project in the first place.&lt;/p&gt;

&lt;p&gt;Sometimes project leaders already have an idea of a desired outcome. But if they don’t question their biases and remain curious about what the data is not saying, they will miss big opportunities.&lt;/p&gt;

&lt;p&gt;In the &lt;a href=&quot;https://hbr.org/2017/08/3-things-multinationals-dont-understand-about-africas-middle-class&quot; target=&quot;_blank&quot;&gt;Harvard Business Review study&lt;/a&gt;,  the author and her team talked to people ranging from hairdressers to public relation executives on the ground. This allowed them to rebuke several myths and assumptions about middle class African consumers. They discovered that many Africans valued durability over flashiness when buying higher-cost items.&lt;/p&gt;

&lt;p&gt;The team found that MNCs also overlooked an important fact. In Africa, traditional ties between individuals and large family networks remain important. As a result, they may have ignored opportunities in regions that would appear as poor in the data.&lt;/p&gt;

&lt;h1 id=&quot;in-short-start-those-data-projects-right&quot;&gt;In short, start those data projects right&lt;/h1&gt;

&lt;p&gt;These examples may appear anecdotal at first. But collecting these stories on the ground will help identify gaps in your data and your approach to it. It will also save you time and resources in the long run.&lt;/p&gt;

&lt;p&gt;Adopting Agile and Lean UX strategies to the way you approach data will lead to better results for your projects. So start GOOBing with data today.&lt;/p&gt;

&lt;!-- At Plot + Scatter, we bridge the analytical and creative realms of data to create impact in everyday life. Our work is guided by a simple question: “How does this serve the end-user or audience and make data meaningful and truthful?”  If your organization has been working on a data project, we’d be happy to chat about the possibility of working together. Send us an email and say hello. --&gt;
</description>
        <pubDate>Tue, 26 Sep 2017 00:00:00 +0000</pubDate>
        <link>http://plotandscatter.com/ideas/2017/09/26/start-goobing-with-data.html</link>
        <guid isPermaLink="true">http://plotandscatter.com/ideas/2017/09/26/start-goobing-with-data.html</guid>
        
        
        <category>data</category>
        
        <category>agile</category>
        
        <category>ux</category>
        
      </item>
    
  </channel>
</rss>
