I have been visiting some of my clients for the last four weeks. One of them was a fishing company that transformed its business using data.
They’ve come a long way to do this, but the results are stunning! Today I want to share with you a couple of things we can learn from them.
Let’s start with a quick introduction. Our main character is one of the largest fishing companies. They are operating in four different countries on three continents. A couple of years ago, they were not an IT company and didn’t collect any data. Today they are a data-driven organization, which gives them a massive advantage over their competition! Based on data, they can optimize their revenue, directly update their fish harvesting processes, decrease their environmental footprint, and even cure fish diseases by changing some environmental variables — and this is only the beginning of their data-based abilities 😉
So what can we learn from them?
Lesson 1: Try to understand why data is essential.
First, let’s start with why we need data in our organization. There are many reasons for that. We need them to:
Make informed decisions — not based on your gut feeling, but on the actual data
To track and improve the performance of specific processes — to optimize something, we need to measure it
To ensure compliance
To make data-driven decisions
To automate and create predictive systems (like legendary predictive maintenance applications 😉)
In the case of our client, many decisions people made based on their gut feelings and previous experience, but no one could tell the exact reason for specific feeding patterns or harvesting time. Moreover, these decisions were specific to different locations. Because of this, it took a lot of work to improve something or to standardize it. These challenges led to one conclusion — we needed to collect data in our organization and become a data-driven company!
Lesson 2: Start collecting your data as soon as possible!
Working with data is tricky — at the beginning, you need to know which of them you will need in the future and store them. I always recommend collecting all possible data at the start — sure, some of them will be junk — but thanks to that, you are building your organization’s data awareness culture. At some point, you will need to curate those data and create data governance standards, but at the beginning, start collecting them.
To do this, you need to figure out to which data you already have access and what other metrics, logs, or data you need to start collecting. Sometimes it will require new IoT systems that can measure new parameters. In other cases, you could stream data from existing applications or build simple ETL processes to collect them. At this stage, concentrate on one thing — how to collect data.
Lesson 3: Create first insights based on stored data.
At this point, you should have some data stored. Try to prepare simple reports showing their utility to others in the organization. It could be elementary reports with some KPIs or metrics that will offer some basic but essential information for them. Start with small steps and ask your colleagues what information will be most beneficial for them, and then make them more visible using BI tools.
In this specific case, together with the client, we created simple PowerBI reports for business owners, c-level, and people that work directly in their facilities.
In this step, our primary goal is to deliver first insights that can help others and build fundaments for a data-driven decision process.
Lesson 4: Add more advanced analytic mechanisms and create your data architecture.
At this point, your organization already collects data and is aware of the benefits of using them. It is the time to use their full potential. To do this, you need to do two things. The first one is to create your data architecture and data governance standards. If you are using the public cloud, you should develop your data landing zone and data management zone (I will explain what they are in another article).
The second one is the process of adopting advanced analytic solutions. To do this, start asking questions and defining fundamental problems that can be solved using data. After that, go to the technology part 😉 Try to use some data science tools and methodology to create models that will be able to answer those questions. Create more advanced reports with dynamic customization based on specific parameters. Make simple predictive systems that can tell users which actions they should take to have specific results. Explore and try to adopt the advantage of data fully.
Lesson 5: Create insights personalization and embrace data-driven decision process.
You now have many reports, systems, and insights for your co-workers. They can see every single thing using data that your organization collected. But wait, for some reason, they are not using them! The problem lies in complexity and data chaos. Too much data have the same value as no data. There is a way to overcome this. Here you will need to use some AI mechanisms and create personalized data feeds for our end users. Imagine that every day a user will receive an e-mail that they can read in five minutes with their morning coffee. In this mail, they will have every vital insight, data summary, and recommended actions for today. It will prime their brain to make data-driven decisions and increase user experience. If they are interested in specific information, they will find it themselves using one of the reports. If the system detects something important throughout the day, it will notify the user.
To do this, you need to create automation based on AI mechanisms that will react to particular conditions and inform the user about them. That will be able to create a simple data summary every morning with selected data for the user.
Only then — when the user receives filtered data with suggestions and essential insights the data-driven decision process will work on daily basics in your organization.
Those are five lessons learned in this specific case. I think they will help you to adopt data fully in your organization. Please share any questions, suggestions, or other ideas in the comments!
Top comments (0)