Visual representation of location data is truly a modern miracle everyone should have access to.
You’ve probably heard the cliché “a picture is worth a thousand pictures”, right? Well, it’s a cliché because it’s true.
If you think about, what’s the first evidence of human beings communicating with each other?
Paintings, not writings.
As a matter of fact, the language in general is evolutionarily speaking a brand new invention.
The point is loud and clear: a visual representation communicates the important points SO MUCH faster than any spreadsheet ever will.
In this blog post we’ll have a look at a compilation of some recent mapping cases we’ve done for our clients. Client confidentiality is very important to us, so we won’t be going into the nitty gritty with these cases. However, you should be able to get some pretty dang awesome insights if you read through the whole thing with an open mind.
Discovering hidden correlations open up the stairway to heaven.
Imagine if you could have these magical 3D glasses that would instantly show you what’s really going on underneath the surface in your business…. You would have an advantage so big your competitors would be thinking 10 years from now what was it that hit them.
So, this is a BIG thing to tackle.
With that, let’s first have a look at some of the problems with the traditional way of doing things.
You can count on the fact that we’re not going to sell you on dropping 1-dimensional data intrepretation altogether from your toolbox. There are certainly good and valid reasons for only playing with numbers. The trick lies in this question:
“Should the need arise, do you have what it takes to project the data into a locational map?”
If not, you have a look at some of these problems we’ve identified, and then move on to read about the benefits and how to get this thing handled efficiently.
1. Blind spots
Let’s say you have a logistical process of some sort. As an example, you could have 100 vehicles touring around a certain area, with the main purpose of distributing the goods you bring to your customers.
Naturally, over time you would be interested to maximize the efficiency of you process, right? In practice this could mean things like optimizing the fuel consumption, ROI of routes, density of customers and profit, time efficiency and so on.
While putting all of this data into a spreadsheet would be an excellent first step, where do you go from there? How do you catch the seemingly irrelevant and unrelated correlations in your investigations?
2. Inability to do effective multi-variant mulling
To continue on the previous one, let us consider this one as a problem of its own.
You see, the truth of the matter is this: life is complex.
RARELY do we have 100% accurate, black and white causal connections between individual variables. And yet, as a leader you need to have a clear grasp on the big picture, and some sense of how it’s working as a whole.
Business is a SYSTEM, where one piece tends to heavily influence the other pieces.
So, let’s go back to our example about the 100 vehicles. Let’s say the management says we need to optimize fuel consumption because the prices are going through the roof (which they are.)
While the intention is certainly good and worthy of implementation, what if we cancel out the most profitable areas simply because they are a little bit farther ahead from out terminals? Sound like an extreme example? Unfortunately many companies succumb to mistakes like these without taking into account ALL of the data they have.
3. Being forced into siloes
Before we go into this, let us say that siloes definitely have their place. As a matter of fact, simplicity can be one of the greatest gifts you could possibly receive when trying to make sense of complicated business systems.
However, thinking in this way comes with a costly paradigm that goes like this:
“Our data is like a cellar: messy place, but with strong enough intention you could actually find something worthwhile.
Right? If something, we’ve learned to get answers to specific queries. So good have we become at this in fact, that because of technological limitations and paradigms we’re no longer able to zoom out from the siloes, and we’re forced to work inside of them.
And even if we WERE able to do it, we wouldn’t have a good method for mapping out all of that data into one, coherent and easy-to-grasp visual.
If you can get the aforementioned problems solved, we can assure you you’re making a tremendous progress. But there’s so much more to it.
It’s not just about Big Data and it’s not just about visualization, but BOTH (more about this in our case study portion.)
1. Access to high-value process optimization
While money is not the only purpose of any business, we are yet to see a manager that wouldn’t want to have create hefty savings under his or her name. And to be honest, this is where the rubber can meet the road the fastest.
Going back to our example of the 100 vehicles, let’s say the management tells us to optimize the fuel consumption.
It’s quickly obvious that it’s probably not the fuel they want to optimize per se, but COSTS in general. So, what if we could find a totally novel way to deliver our goods that not only saves us costs, but is way more profitable at the same time?
It might sound like some kind of faraway dream, but we’re seeing this quite a bit.
2. Instant comprehension of complicated data
The cave painting metaphor is so good because it brings home a very important point about how we human beings comprehend things. It truly is a miracle to be able to take a set of complicated data and draw a picture from it. The reason this is so cool is that everyone can instantly “get” the message, even if they have not done the legwork you have.
3. Fast visual iteration cycles
This is where it gets interesting. Imagine you put all your data on an actual MAP, and then you start refreshing it based on queries you have in your mind. You can be scrolling and role-playing different scenarios as much as you please, and any insights you draw will be based on data, not just gut feeling.
So, how do you tie all of this back into the real life? That’s where our case compilation comes to play.
This case study is a compilation of some of the most recent locational data cases we’ve made. We want you to be able to play along with, so we chose to put the important stuff into the context of an app. You can substitute this with just about anything that uses locational data.
So, without further ado, let’s get going.
Let’s say you have a gaming app. Simply put, you have one of those juicy-looking apps you can download from iTunes or other similar marketplace.
The app is a game of some sort that relies on locational data (kind of like the dating app Tinder, if you’ve heard of it.)
Let’s also say you’ve done a good job at setting up the data tracking, and you have set just about everything in your app to trigger a data tracking response of some sort. You know, whenever someone upgrades their character, purchases something, wins a battle or whatever. Every single one of these actions are collected into a vast pool of gaming data.
On top of all of the things that happen inside the game, you are also dutifully tracking the locational data (kind of like Foursquare, if you’re familiar with the famous Facebook app).
All right. Your app has really made it big in the world. Congratulations! Millions upon of millions of downloads, and an endless stream of positive feedback from the users!
With all of these downloads happening, the amount of data is also accumulating at a rapid rate. So fast indeed is this expansion, that collection of all the data is starting to be in jeopardy as you run into storage problems.
But the biggest problem of all has to do with optimizing the user experience. As we know, the app world is a do or die kind of an environment, where SHARP decisions need to be done on a lighting fast speed and precision.
Thus, the problem you have is that you have to first of all work in a very limited set of data (let’s say 4 weeks worth of time), and then you need to dive deep into these siloes within that limited set of data. As an example, you could figure out how many downloads there were in any given day, but you could totally miss what time of day the biggest spikes are. Further, you have no idea what locations SPECIFICALLY are the most stressful ones for your servers.
Let’s say that you show incredible foresight and business intelligence by letting us help you out with the data mess your precious users are creating. Where do go from here, and what can you expect?
Well, the very first thing we do is this: we will take ALL of the data you have, and put it in one, secure place. Yes, ALL of it. Right from the beginning of publishing the app. Everything.
Then we will start iterating, in other words doing the data magic.
Now we can start asking a lot more intelligent questions.
First of all, let us notice the areas where the biggest growth centers are right now. And not just at a country level, but all the way down to the addresses of users (if we so choose). Then we can start mixing in all kinds of other data. Are the people in these growth centers purchasing the X type or Y type accessories? How long does it on average take for the people here to grab the offers? Are they clicking ads more or less (thus directly affecting our revenue)?
Further, how about identifying some new growth clusters? Maybe some specific language is trending, but you have not localized your app. Who knows, maybe these people are spending 78% more on the game, but with their limited English skills, most of the people can’t comprehend the game (and as a result, they play something else.)
Oh yeah, and as you start making better choices based on the data we have right now, you can re-evaluate these decisions one week, one month or one year from now. That’s the beauty of this! You can project incredible amount of data into one picture, and then see how it moves over time as you make changes to your app. And everything you do will be accompanied with that knowing feeling that you have this under control. Say goodbye to guessing.
So, how could this relate to you?
It’s pretty simple to be honest. Let’s say you have a set of locational data lurking around somewhere. This could be an IoT setting, logistical environment, e-commerce or marketing setting etc. The source of data doesn’t matter much.
We have a very simple 2-step process for putting the data into visual format.
1. Import ALL the data into one place
In order to prevent siloing ourselves into a corner, we first import all of the data into one safe place. You can leave the worries about data size at the door.
2. Refine the data at analytics platform
This is where we’re going to grab the data and project it into visual format. Understandably you have tools you’re already familiar with, so we’ve built a wide support for all the most used tools out there, like ESRI.
Long time ago we arrived at the understanding that data domains of most companies are in no need of even more complicated solutions.
While some bigger analytics companies might have an army of young guys ready to jump on their clients challenges, sparing no expense in finding a solution, we like to simply offer a plug ‘n play solution that’s producing results as quickly as possible.
“Cooperation was great. Our Architect sent afterwards special thanks on how well the day went. The same architect had previously suspected that how on earth it could be possible to take over the entire database in just one day.”
If you’ve got big enough set of data (locational or otherwise) that you would like to put into visual format, there might be a fit between us. We do require, however, that the amount of data is significant enough.
The reason for this qualification is simply to make sure that we are both committed, and that there’s a potential for extraordinary ROI for you.
So, if you’d like to skip the traditionally long planning phase and simply jump into it right away, let us know through our contact form.
COO, Service Delivery
+358 40 550 2524
Fredrikinkatu 61 a 6. krs