Has your organization reached optimized business outcomes?
The rapid growth of analytic solutions has not transformed every organization into a data-driven powerhouse.
Though it seems there is a new, modern technology introduced every month with the power to change your organization in unimaginable ways, you may not have been able to implement appropriate technologies and connect the dots — yet.
As with any endeavor, your company will face obstacles as you strive to transform into a data-driven organization.
In order to succeed, it’s important to evaluate where you are today, the challenges that lie ahead, and what exactly “ahead” means for you and your team.
- How to determine your company’s current analytic maturity level
- How to handle indecisive or non-supportive leadership team
- The best ways to work through data quality issues
- Tactics for getting people on board with the digital transformation
Here are the highlights from the webinar. We’ve listed the approximate location where you can listen or you can just read these highlights.
You first need to talk about what the analytics maturity model is and it’s very simply just a way to measure the stages that organizations go through to use data and analytics effectively. So basically it goes from the very technical term of Zilcho to being completely data-driven. Now, as you can imagine, the better business outcomes live in the level four and level five areas. But how do we know that? So there’s been various studies done and one study that looked at different companies said that when they were at the fourth level of maturity, they were often on the top company lists. So when they were ranking higher among their peers as being better. Also, when you look at analytics, maturity and financial performance, there is a positive correlation and us as datas professionals love to see positive correlations. But another part of this is how many people are in the lower levels?
How many companies would you estimate in the lower levels? Well, I’m going to tell you that it is in the disappointing 87% from a recent survey [Dec 2018] that Gardner did where they asked the companies to rate themselves. So I’m asking you that if the better results are in the higher levels, then why are so many companies dragging behind and we actually have some thoughts about what’s causing it and some ideas on how you can move forward. And that’s what we’re going to talk about today.
First of all, we’ll review the analytics maturity model and we’ll talk about the common blockers, what happens, and then we’ll talk about some success strategies. So the analytics maturity model, if you’re not familiar, there’s a a sci organization and they have the capability maturity model. And what happened was in the 1970s the US federal government was having issues with software.
The projects were not being delivered on time and they didn’t have high quality. And that’s a very unique thing where I know that doesn’t happen to any other companies or anything.
So let’s talk about what the [Analytics Maturity Model] looks like [when applied to real organizations]. The first stage is Chaos.
[12:54] [First blocker is] leadership, you got to have a strategy. And I am often amazed at how many companies actually purchase. A very fancy software like SAS via or other software and they have no idea what they want to do with it. We’ve had customers call us and say, all right, we got the software, come tell us what to do. And it’s kind of strange if you think about it because data can answer so many questions and it can help in so many ways, but you’ve got to know what question you’re trying to ask.
How do you overcome it? Well, successful companies have a strategy and here’s what’s important about a strategy. It forces decision making and priority setting. Otherwise you’re in that situation where [managers are] trying to make decisions when they don’t understand what’s going on [with upper management].
Blocker number two, there’s no buy in and it literally looks like a tug of war management is going one way and the staff is pulling them the other way. And maybe as part of the data staff, you recognize yourself in this picture. There’s different ways this goes on. So a lot of times leaders want to trust their gut over the data and they just had this big issue with trying to move further with the data. They just [want to do] what we’ve always done in the past has worked.
[Part of blocker number 2] happens if the leader just doesn’t understand statistics. They haven’t had it since college or maybe they didn’t even have it in college. And when you start talking about machine learning it just sounds like something so scary!
One thing I like to recommend is an analytics coach. Analytics coaches are useful for several reasons. They’re particularly good in the level two to level three transition. If you think about a coach, a coach is someone who’s going to come in and work with an entire team to win games and to win a season.
Someone who’s got the big picture in mind, someone who is trying to make everything better. The first thing a coach would do, is level set the [leadership team]. They would want everyone to understand: here’s what statistics mean, here’s how you use machine learning, and here’s what it does for the company.
So another strategy is if you just do a good ole a proof of concept. So look for small scale projects with achievable results. And when I say small scale, I mean something where you can get a result within a month. You know that the data exists and you can see something where you’re going to get a high return on investment. I would look for areas where the company spends too much money or something where it’s close to the customer.
[Blocker number 3] Data is a huge area where we see a lot of blockers put in place. Now, data presents a lot of opportunities and challenges all along the analytics spectrum curve, right?
So as you’re trying to move up the analytics maturity model, data can be one of those big, big roadblocks that you run into. And one thing that a lot of our customers don’t realize is that the farther you get along the analytics maturity model with a doubt without addressing these data issues, the bigger those problems become that because they simply compound over time.
So what, what is the recommendation? How, how can we fix this problem? Well, um, you need to get control of your data and steer it back into the right direction. Forgive the picture Pun, but oh, we have three steps that you need to take in order to get control of your data and steer it back into the right direction.
Now the [fourth] blocker that we see in the data world is siloed data. I’m sure everyone here has experienced with siloed data. I’m sure that it even infiltrates your dreams at night and giving you nightmares. But for those of you that don’t have experience with siloed data, it is exactly what it sounds like.
You get these walls in between different data sets that basically provided or prevent eager, easy aggregation or joining of the datasets. And so, uh, you need to focus on breaking down these silos. Now, um, we see data silos and a number of different ways. Um, silos aren’t always a bad thing. Oftentimes they’re legitimate reasons to have silos put in place such as protecting customer private data, right? Other times we see that they exist simply because there hasn’t been a need up until now to break down those silos.
So what’s the, what’s the best type of success strategy that we can put into place? Well, when we go come in to a customer site, we often hear this idea of let’s create a data lake, tear down the silos, get rid of them, and throw all the data into this massive data lake so that we can get this beautiful pristine location where all the data works harmoniously together. Well, the problem is that when you break down those silos without any sort of strategy or ETL processes in place instead of a data lake, you often get a data swamp and nobody wants to live in a data swamp!
So what’s the actual, um, best solution for this? Well, again, find a proof of concept, right? Specifically, you want to focus on high value projects, especially, especially ones that are going to have a fairly quick turnaround. Okay. You don’t want to take on a project that’s going to take you four years to complete because there’s going to be burnt out and management isn’t going to want to continue to be that sponsor for you. And then finally, um, you also need to focus on utilizing data across data silos as critical when trying to, um, show to management the value in and breaking down the silos because you need to show them that there’s a lot of great cross pollination that can happen within the data.
The [fifth] blocker that we run into when you’re talking about people in skillsets is the culture issue, right? Culture problems present themselves in a number of ways. Oh, you can get stubbornness to adopt new ideas. When you’re bringing on a new project or product, this fear that my skill sets are becoming obsolete. And so there’s this kind of resistance to change or this idea that the current projects just the flavor of the month.
So this passive resistance [occurs] until the fad fades away. As a statistician, I get this a lot, “My Gut knows better than your math, nerd.”
So, you see these cultural issues pop up quite a few ways, but you have to address it. And one of my favorite quotes on, on a culture is from Peter Drucker. No, I know I’m a millennial, but even I know who Peter Drucker is and he says that culture eats strategy for breakfast. And we all know that breakfast is the most important meal of the day.
That’s why you need a good analytics coach! You don’t just need somebody to create buy in from the management level. You need somebody that can break down the vision into those bite size morsels and explain it to your people, to your talent to help get them on board and buying into your vision. And so that you can take progressive steps up the analytics maturity model.
Now the [sixth] blocker. So [fifth] blocker is culture, [sixth] blocker are those sweet, sweet skills. And I’m not talking about nunchuck skills or bow hunting skills. I am talking about technical skills.
A lot of companies, we go in and talk to them and they were like, if only we could find one or two unicorns to solve, solve all of our problems and we’d be golden. Well, well that’s true for a minute. There’s kind of this unicorn paradox because while they do fit that perfect blend all in one person, you’re not going to keep a Unicorn tied down. They are free spirits. They know that they are valuable and the grass is always greener on the other side. And so after a year or two of them just churning out these amazing results for you isn’t going to inevitably be this day where they’re going to come to and say, I’m sorry, but our blessed for marriages over, I’m leaving you. And that’s devastating because what happens is after you spent,
So much time and resources in this person to be that, um, Jack of all trades, the moment they leave, they creating massive brain drain for your company. And so you need to avoid that. So instead of looking for those magical beasts out there, focus on creating a team, right? You, so build out that end to end solution within your data science team.
Look for the people who have the data engineering skills, the database skills. Look for the people who can do the data science of programming the statistics, and then find those people who can generate those insights and disseminate the actionable insights down to, um, the management level so that they understand what, what to do next.