Go to navigation menu

GIGO from ITSM tools

As an ITSM consultant I have had many opportunities to review and assess ITSM tools. Many of these tools including fancy “cloud positive” had a very poor setup. Imagine that you cannot even distinguish between incidents, service requests, problems and changes. At the end all were handled as a ticket, very randomly categorized. Based on such data, many decision-makers believe reports produced by ITSM tools are correct and then often make very questionable decisions based on that.


Like Wikipedia says: "Garbage in, gospel out" is a sardonic comment on the tendency to put excessive trust in "computerised" data, and on the propensity for individuals to blindly accept what the computer says. Since the data entered into the computer is then processed by the computer, people who do not understand the processes in question tend to believe the data they see and take it just as seriously as if the data were Gospels.

In 2018 we heard about many AI failures. Quite well known are AI simulations about the winners of the FIFA World Cup. I like this article by Nick Burns about Understanding why AI has failed to predict the FIFA World Cup 2018 and his lesson learned from IA failure. He stated “no matter how good your models are, they are only as good as your data.

In the near future, many organizations within the ITSM domain will start with AI. But if their current data is garbage, then what could be a possible result of using AI? Obviously even the best AI will simply do garbage in, garbage out (GIGO). The real danger will be if such AI output will be used as a gospel.

So, before you start with AI, or for example automation within ITSM, assess how prepared you are for that. I know that “old” fashioned topics like well-established ITSM processes which can produce reliable data are not so fancy, but it seems that without it any advanced level of ITSM is not possible. But some organizations will try to find a shortcut.

Finally, one good recommendation. In any area where you process data, you should report not just the results but also the uncertainty of it. Because without the stated uncertainty using the results for day-to-day decisions are not possible. There is a big difference between “the tram is leaving around 3pm” and “the tram is leaving at 3pm”. It is a question of trust and transparency to show the uncertainty.

Miroslav Hlohovský, Head of Consultancy Services, OMNICOM

Back to News