Making a Data-Driven Culture.
Introduce a Data Driven Culture and other Healthy Habits.
Thinking about making your business a little more “Data Driven” and less “Intuition Driven”? These two are not necessarily in opposition of each other, however one should consider which role each factor plays in their decision making process. Introducing a “Data Driven” process into a company’s culture is a popular refrain nowadays. So why is it so difficult to implement such a popular concept? What follows here are a few of the reasons why making these changes pose such a challenge, along with a little advice.
First of all the psychology of change is something to be understood, and the lucky thing is there are many parallels in human behaviors and psychology we can draw from. I think we can all agree at the outset; everyone resists change! So, get ready for an uphill battle.
The first stage is usually to get an idea of where everyone is, and identify any pain points on which to gain a little leverage. To do this we recommend a mix of interviews and observation. …Interviews? Nobody likes Interviews, especially at work where it feels a little disruptive and maybe even intrusive. So the point is that they take time and can be perceived unfavorably among staff, especially in a “protective” environment.
This is where many introverted and/or tech-centric folks are stopped dead in their tracks. Soft-Skills are a must in order to gain trust, identify skepticism, and respond appropriately. Active listening and engaging at the right level will go a long way toward gaining a partner-like relationship. Using information gained here is what you’ll be using to develop your plan, so be as accurate and thorough as you can.
So you’ve had your interviews and learned a ton! You have developed your plan and can’t wait to get started showing how your solution can be the best thing to happen for the company. Not so fast; Try not to bite too much off at one time OR force sweeping change, especially in an unreceptive environment. The journey of a thousand miles, begins with a single step. Change may need to come in a small step-change and highly iterative way. Just keep the big picture in mind and don’t be afraid of change yourself, making appropriate corrections and small adjustments as needed.
Make sure whatever you propose is measurable and likewise is actually measured. If the measure is too small or atomic the mere task of measuring and presenting the results can be disheartening. Take for example, weight loss. If you weigh yourself every couple of hours, you’ll soon give up. However, if you weigh yourself once or twice a week, you will start to see what is working and what is not. Weighing yourself once a year will not work either. So the trick is when developing measures, be sure to try and identify the sample rate that makes the most sense.
Using an array of measurements also helps develop a sense of correlations and drives insightful measurements. For example, selecting both caloric intake, mood, and activity levels will give a much better picture than if you were to just measure weight and calories, or weight and activity.
Business Intelligence Servers versus Cloud
On Premises versus Cloud
We have been seeing lots of blog posts, and articles on the question of "BI on Premises or in the Cloud"? I personally remember the days when it was a simple choice, with the exception of one solution everyone using DSS / BI were using the systems in a Client - Server architecture. Almost every product was limited to installing servers and installing a BI Client on their users desktops. Then along came the push to make it all available through any browser.
This meant the server now needed to have a webserver implemented and didn't rely on proprietary application servers, rather a Tomcat or any other web application server would do. This of course meant that you had to deploy your BI solution as a full fledged web application, with security and all the rules imposed by good web deployment practices to make sure the appropriate user had access to the appropriate level of information no matter which browser they happen to be using.
Eventually this evolved into the discussion here and now; "do we host or do we use cloud?" is the question of the day in 2017.
as I said, lots of resources are out there, but our question is "Why not a blended solution"? We are able to deliver an application which uses the latest in API utilization to minimize the need to make the decision at all.
even so, here is a nice write-up for those who might want another perspective:
Data Pipeline Blog