Designing Data Driven Applications in Tableau: How We Scope

What I love most about Tableau is that it lets me do some pretty amazing things without writing a line of code.  Dashboards, yes, but also phone apps, mockups, and even interfaces for products.

Which brings me to my latest hobby/project: a smart home.  Now, I understand loads of people are plugging their Nests into Alexa and having all kinds of fun.  Call me what you will, but this will not do at the Cogley house.  So, the engineer and I have decided to embark on a little home hobby that is equal parts terrifying and delightful.  So, what does this have to do with Tableau?

The equipment may be industrial, but the data lives in MySQL.  And data means I can pop it in Tableau to make the interface I want.  I’ve practiced making mini-applications for IronViz feeders, in part out of curiosity and it in part because we’re moving towards ‘dashboards that do’- that link to existing systems, that make exploring existing software packages easier, and that support both information finding and acting.

Any successful application requires planning.  My process comes from design thinking:

For now, we’re going to start on the ‘what is’ phase.  That’s our understanding and problem scoping phase.  The people that like paper might call this a project plan.

This is the point where we take stock of our world.  We understand what data we have, we find what data we don’t have, we make a list of what we still don’t know, and map our way through what is today.  Notice what’s missing – a plan.  Right now, we’re at the understanding point.

Your checklist, if you will, looks a bit like this:

• Document the world:
  • Who is involved?
  • What happens today?
  • What does this look like?
• What data is available?
  • Where is it?
  • How accessible?
  • How complete?
• What external data is needed?
  • Public data, such as weather
  • Shape files
  • Visual assets
• What data needs created?
  • Short term (sample)
  • Long term (& how?)
• What does the user experience today?
  • Job/task
  • Where will this piece fit in?  What’s there today?
  • What else happens around this job/task?
  • Who else is involved?

For my smart home, the as-is list looks like this:

• We have 1 data analyst and 1 engineer
• We have a server, which includes a MySQL database
• The house is a particular size and we’re mapping out the floor plan. We also have access to the government site data, but it only provides so much granularity.
• The electrical outlets exist in X places which tie back to Y fuse in the fuse box
• We already have a water filtration system, which documents our water usage and pressure readings
• We’ve also added various sensors to monitor the water heater and furnace
• We have other sensors to log the temperature in various places
• We have a rough idea what we want to use to control lighting.

Today, we already have a basic programmable temperature control (the kind with 2 buttons that comes from the 1990’s).  For the most part, I don’t have to touch it.  My phone gives me an idea of outdoor temperatures and, yes, I can even open the door if I want to go retro.  Alexa doesn’t control my lights or audio – I do.  So, for now, we’re fairly low tech, considering others (AKA, there is no Alexa).  And, no, she’s not welcome here.

Me and voice-activated tech

As someone classically misunderstood by voice-activated products, I have no interest in yelling at HAL to turn on my lights.  I like the Star Trek ideal of touching things or, dare it say, having things work seemingly by magic.  So, part of that is the vision and my inability to be understood by anything ‘techy’ goes on the journey map.

The journey map logs my life today.  It’s where I highlight my pain points in contrast to what the Jetson’s have.  You know, turning on the lights…

So, what does this look like?

• I rely on my bill providers to inform me of my energy usage, so I can’t accurately correlate it back to behaviors that I could reduce or outside influences that affect my data (cold days, roasting marshmallows on the fireplace, leaving windows open).
• I turn lights on wherever I go.  Sometimes, they stay on…you know, because the cupboard monster might eat me.
• My music lives on my computer or my phone or YouTube.  Not all devices have the same things and YouTube commercials are getting unreal.  Seriously, 2 minutes for a 3-minute song?!  Or, my favorite, 5 minutes for the 30-second clip.
• I like a long hot shower and others do too.  Sometimes, this means I get a short cold shower.  No one is happy with that.
• I have data where I can monitor on temperature readings and water usage.  I need to go to my computer, which means going into the (cold) basement.  Unless there’s a fire and hot chocolate, you may as well write this one off, kids.

(It really is that cold in my office.)

This, kids, is step one.  You’re probably as miffed as the engineer who got excited when we went to the whiteboard, only to discover we weren’t brainstorming…yet.  Come back for part 2 and I promise there will be stickies!

Designing Data Driven Applications in Tableau

Part 1: How We Scope or What Is

Part 2: Brainstorming or What If?

Sneak peek

Part 3: Prototyping or What Wows?

Part 4: Deploying or What Works

Happy hexagons!