I wouldn't be surprised if your business processes are wasting more time than creating actual value.
That's typically what I see when I do Value Stream Mapping (VSM) for a client. VSM is a lean manufacturing technique that helps pull the most value out of your processes. It's a visual technique that brings your experts together to solve important problems about the performance of your business processes.
The ready, fire, aim nature of the way our business processes are typically born provide little room for critical thinking and design; instead, we use common sense and rapid problem-solving to quickly assemble a process that meets the customers' demands. Over time, the process grows organically—a tweak here and there in reaction to ever-shifting customer demands. At some point (usually when a competitor figures out how to deliver a similar product faster and cheaper) you wonder if you're doing things smart, and that's when you might want to turn to the smartest group in your company: the data scientists.
Traditionally in a VSM effort, the team is comprised of product experts: operators, designers, production managers, etc.; however, in this new age of big data analytics, data scientists can help tackle this challenge.
SEE: Hiring Kit: Data Architect (Tech Pro Research)
What's the current process?
The first step in VSM is understanding the current process. But before we get started, it's important to understand that data scientists use critical thinking to arrive at tenable conclusions. Good data scientists will ask a lot of questions, which may seem time-consuming, challenging, and even combative for outsiders to question the production experts. Critical questions are a vital part of the process, and they must be addressed for the sake of the goal and for the team's well-being.
So, as your experts assemble the current-state process, expect your data scientists to ask questions that challenge assumptions. The goal is to create a representation of the way the process really is, not what the experts think it is.
What data should we collect?
Once the current process is mapped out, it's time to determine the metrics that contribute to good performance; your production experts will have good ideas about what's important: inventory, cycle time, quality, and so on. Those familiar with lean manufacturing will be looking for areas of waste such as overproduction and transportation, so you might hear some of these metrics discussed.
Your data scientists might be thinking in another direction, though. If the goal is to find the biggest contributors to performance in each process step, then how do you know if you're right? There may be a metric that has a huge impact on performance that nobody knows about. It's important to encourage this critical line of questioning.
SEE: Free ebook—Executive's guide to IoT and big data (TechRepublic)
What does it all mean?
Once we know what data to collect, it's time to start collecting and analyzing it. You expect data scientists to have copious ideas about collecting and analyzing operational data, but keep in mind that production experts also know a lot about collecting and analyzing the data that matters to them, and these experts know a great deal about the equipment that's generating all this data—something the data scientists know nothing about. So, this is not the time for data scientists to be didactic about data collection or even analysis. What's important to discover are the essential questions the production experts can't answer or explain and that will eventually lead to a better tomorrow.
What process should be used in the future?
The final step in VSM is to design what the future process should look like. Up until this point, the data science team has been involved, but it doesn't seem like they have contributed much to the solution other than asking a lot of questions. Now it's time to see how data science brings VSM into the 21st century.
To understand how the process should perform, you must determine its theoretical best performance targets. Data scientists will use operational intelligence to see what's truly possible instead of production experts making educated guesses. Then, data scientists can build sophisticated simulation models to see how the process will perform under different configurations. This would be impossible for production experts to do, even with a mildly complex process. So, it seems we can teach a few old dogs some new tricks—with a little help from big data analytics.
SEE: 2017: The year data science will live up to its potential (TechRepublic)
Good data science makes good sense
We've just walked through an old lean manufacturing process (i.e., VSM) with a big data analytics twist. The basic framework remained the same:
- understand the current process
- determine the data to collect based on the largest contributors to process performance
- collect and analyze the data
- develop the future process based on findings
What's different is how data science thinking, practices, and techniques add so much more value to the outcome. It's more than just throwing in fancy algorithms and tools—it's a data science philosophy.
At the heart of all good data science is curiosity. I find that most experts have more answers than questions, but data scientists are different—they have a lot of questions that generate more questions. It's all part of critical thinking and the scientific process. This is a culture shift in most cases that requires proper leadership if you're going to include data scientists in your efforts with experts from other disciplines. But it's well worth it.
In the case of VSM, data science adds a lot of value in forcing the team to think critically about the current process, the metrics that contribute to performance, and the data collection and analysis process. But of course, it doesn't stop there—when it comes to future-state process mapping, data scientists pull out tools and techniques that are not accessible to others.
If you're currently wondering how to get the most value from a critical business process, take some time today to organize a group of experts, including data scientists, to run a VSM effort. With the critical thinking that data science brings, you'll have valuable insights flowing in no time.
- 6 myths about big data (TechRepublic)
- How critical thinking, or lack of it, makes or breaks projects (TechRepublic)
- Staying agile: data-driven IT operations (ZDNet)
- Iconoclasm of the day: Is lean manufacturing long in the tooth? (ZDNet)
John Weathington is President and CEO of Excellent Management Systems, Inc., a management consultancy that helps executives turn chaotic information into profitable wisdom.