Share this post:

”In God we trust. All others must bring data.” This quote, made by W. Edwards Deming, refers mainly to the importance of data measurement and analysis when doing business. In IT, like in business, data analysis is equally important. Thus, for the last decade, we have been talking about Data Warehouses, Big Data, Data Lakes and lately Data Science. Data Science has become an IT discipline by itself and one of the hottest things you can become these days is a Data Scientist.

In most companies, there seems to be this perception, that we have all this data and we just need to handle and analyze it right, to gain insights and benefit from it. Conclusion: Let’s use Data Science!

That’s all good, but how do you do it?

Data Science is a key element when working with AI, Monitoring and Automation. You use Data Science to analyze and group data, create correlations, look for clusters and essentially gain insights into data, that you cannot get from standard reporting of the tools and systems creating and storing the data. Most Data Scientists use either Python or R to perform this work.

My personal experiences with Data Science are mainly related to Service Management, where we use Data Science to analyze ticket data, Incidents, Changes, Problems, Request for Service, CSAT and more to identify specific patterns and automation potential, which in turn leads to a number of recommendations for how to make the Service Desk and Service Management processes more effective and less resource consuming.

However, it’s not just a matter of getting a ticket dump and applying Data Science. You also need to understand the context of the data you’re analyzing, how it’s being used and what processes, it is supporting. Data Science, by itself, does not add value, unless you understand the context of the data you’re analyzing. It might seem obvious, but there sometimes seems to be a misperception, that:

“”We have a lot of unstructured data, we know nothing about. Let’s add some Data Science to make some sense out of it””

It doesn’t work like that.

5 recommendations for using Data Science:

Use Data mining and perform an initial Quality assessment

1. Start analyzing using a Top-Down approach and build/analyze clusters as you come along
2. Use benchmarking against industry standards, where possible
3. Explain the findings using Story-telling
4. Deliver the end-results to the customer focusing on the value, including:
   • Quick-wins
   • Main recommendations
   • Input for the roadmap and associated business case(s)
   • Next steps

I could, of course, elaborate on all of the above, but that would make this blog post too long. Instead, please read more about what IBM can offer here:

Or … reach out to me, if you’re interested in Data Science and how YOU can use it to bring value out of YOUR company’s data. There’s a lot of opinions out there, but do you have the evidence to prove your points? If not, bring data!

If you have any further questions, please do not hesitate to contact me at HHansen@dk.ibm.com.