A Business Stakeholder’s Quick Start Guide to Useful Analytics
3 Common Myths Exposed that You’ll Want to Avoid
by Mike Sturm
“We want to do analytics and machine learning on our Big Data.”
At Hashmap, our primary mission is to help our customers generate measurable business outcomes leveraging Big Data, IoT/IIoT and AI/ML applications. Many of our customers are experienced players using Big Data as part of comprehensive business focused strategies, but we talk to a lot of customers who are just starting out in the early stages of that journey. The statement above represents a loose paraphrase of how these customer engagements start.
Avoid the Technology “Rabbit Hole”
It’s easy to go down the technology rabbit hole right away when you hear something like this. We’re technology people and as we are often talking to IT professionals, everybody there is quite happy to immediately talk about the relative merits of different technologies. But it’s critical early in the engagement to get to the basic question of why?
No business simply wants to create a big data architecture, classify data, train machine learning models and build analytics dashboards for the fun of it. Rather they want to use those technical methods and techniques to create a business impact.
Getting to “Why”
Now at this point, I’m sure most of you are probably thinking that this is pretty obvious. And the reality is that there will be a stakeholder in the organization who is feeding the business requirements to the IT team. When we ask the why question, the IT team points us to that business domain expert.
As the guy who is responsible for understanding the customer’s business objectives and insuring that the outcome fits the objective, interacting with the domain expert is where I earn my keep.
As I work with the business stakeholders, there are three common myths that appear consistently. I refer to them as myths, not so much because they aren’t true, but more because they represent a lack of experience in how analytics and machine learning can be applied to solve business problems. Exposing these myths and clarifying the business challenge is the first step towards helping my customer move to a meaningful outcome and avoiding wasted time and effort as they move into the analytics and ML arena.
So if you’re a business stakeholder and you see others in your industry exploiting analytics successfully, here are the three common themes to be on the lookout for.
Myth #1: If you build it, they will come
This is the fallacy of more data must mean better analytics. It sounds good to be able to say that you are storing everything and that all the data is accessible somewhere. This is one of the most prevalent concepts that I run into and is the result of someone from the business side telling the IT group they want to do ‘analytics’ without a clear idea of the question they want to solve.
I like to capture this strategy as follows:
Step 1: Acquire the data
Step 2: Connect that data to a BI tool
Step 4: Profit!!
Now clearly that’s a bit tongue in cheek, but it usually gets the point across. Data by itself is not a cure-all. A domain expert with enough data can probably get to some useful insight given some time. But in most cases, you’re much better off defining a few specific problems and then identifying what data is needed to answer those questions.
For an organization just getting started with analytics, keeping the initial problems well scoped with clear business objectives will help keep everyone focused on the target while building the experience and competencies to move into more advanced techniques.
Myth #2: Analytics produces business value
Analytics is a powerful business tool and good analytics data can lead to very valuable insights about your business. However, those insights are not an end point in themselves, they need to be translated into actions that the organization can implement.
Let’s assume you have a defined problem to tackle, have well classified data, create a simple user-friendly report, and have now generated several powerful new insights. Now the question to ask is so what? Business owners often start from a presumption that the analytics will lead to measurable value. This can be true, provided that the organization is capable of following the direction that the analytics points to.
It’s a useful exercise to game out some possible bookend answers for the problem you’re trying to answer. Think about some double this, reduce half of that kind of scenarios. Can you overcome your organizational inertia and hurdles to achieve these? This helps to understand the potential value of the answers you may find versus the effort required to act on them. In the worst case where the organization is unwilling or unable to act based on the answers, then focusing on another problem that is more tractable is the right move.
Myth #3: Your IT department is a vendor
We see this repeated often, the business defines the requirements, provides the funding, and then hands the project off to IT to deliver. The IT department implements then shows the solution to the business owners only to hear what we really wanted was…
This traditional view of the IT department as the vendor and the business as the customer is valid for enterprise level software solutions (think email and networking). But for building successful analytics platforms and workflows it’s important to view your IT department as a partner and not a vendor.
This requires an investment in collaboration from both sides. For the business stakeholders, they have to acknowledge the time commitment needed to work directly with the IT team during the implementation. For the IT team, they have to become comfortable with requirements that evolve in real time. On both sides, the people will be moving out of their comfort zones.
To be clear, I’m not advocating a full agile approach. This guidance is simply a recognition that good analytics is based on clearly defined and classified data. Having the business and IT working closely together during the process will result in an accelerated path to useful insights.
Getting your first analytics project off the ground can be challenging. From the business side, being outcome focused with a clear statement of the question to answer is the place to start. If you can avoid the three myths you’ll be well on your way to capturing the value locked in your company’s data.
Feel free to share on other channels and be sure and keep up with all new content from Hashmap on our Engineering and Technology Blog.
Mike Sturm is Director of Industrial Practice at Hashmap providing business guidance and operational expertise across industries with a group of innovative technologists and domain experts accelerating high value business outcomes for our customers.