I just this week for a 1-1 with an attendee at our Latin American Gartner CIO Summit here in Cancun, Mexico. We were exploring how to gain improvements in the use of information to drive better business outcomes. We focused on the use of business intelligence and analytics solutions. The attendee said something that caught my attention. He said, “We are using self-service now to allow our users to create their own reports.” This sentence might sound innocent and it might even apply to your business. But I think the point is very important because I think it describes a problem we see a lot of. What exactly is self-service capability and when should we use it, offer it or acquire it?
Where I work we have implemented self-service travel booking. In year’s past we used to email or speak to an assistant, or call the booking office directly, and talk through what was needed. All the while the other person would do the actual key-punching into the relevant systems. Some years later this process was “improved” in that new systems were introduced to permit, allow or require me to do the booking myself. As such the work didn’t really go away – the same work was executed. What changed was that I now spent my time doing the work. The overall net in terms of time spent on the work dropped somewhat since only one person was now involved. But there was no real increase in productivity.
In effect this self-service effort just moved work around. It is this principle that the attendee who sat in front of me just repeated back to me. He noted that in the past his IT department used to create and service reports for his company’s business users. After implementing “self service”, those same business users could now set up their own reports when they wanted them. What was explained to me did not quite fit with my understanding of what self service ought to be about. So the question becomes: what is self-service and where does it actually add value?
I think it might help if we first looked at the work being undertaken. Setting up reports is, in general, pretty easy if you think about. I recognize that it can easily become complex and ugly but reports tend to be focused on historical data and simply report on what has happened. This is the most basic of “business intelligence and analytics”, if you will. It’s what we have all been doing for 30 years. We used to put such work into groups called, “data processing” or “management information systems” and now, “IT”. I think applying self service here is not overly smart. What if each business user set’s up the exact same report? Such users might feel they were empowered but the firm’s costs would rise without any improved value creation.
A better way to think of self service might be this: When the business users spots a problem, or a challenge, or a risk, or some other condition that does not fit normal business performance or behavior, they need to solve a problem or improve something as a result. It is kind of work where self service here makes perfect sense. This is not about “setting up reports” but actually doing some root cause analysis, some ad-hoc exploration, some exceptional activity. Empowering those users with self service is really smart. But note this is not repetitive, standardized work. It is complex, good thinking work. So the attendee and I concluded together:
- Don’t seek to apply expensive self-service capabilities to standardized routines – its just a waste of money
- Do seek to apply self-service capabilities to non-standard people-driven problem solving or opportunity exploration work.
This seems to fit with where others have noted the use of self-service capability in the data and analytics market as a whole. We have seen self-service data preparation and data discovery tools – which provide a set of flexible capabilities that non technical people can use to find, discovery, search and collect data. We have not seen self-service MDM or information governance solutions, but I think we are seeing information stewardship solutions with self-service capabilities. This is because information stewardship is all about policy enforcement, so periodic root cause analysis is needed.
Now, before you shoot me down here, I have to accept one other obvious point. What starts out as a one-off, periodic problem solving exercise might end up becoming a repeated solution. As such, that work starts to look more repetitive and even standardized, even if there are the odd exceptions. So clearly there is a transition between non-routine and routine. But at least we might agree on the principle here.
By the way – there is an update to my experience with self service and travel booking. Over the last few years the effort to keep booking all these trips became too great. The task is, at its root, a standardized task with just a few unique variables each time. So after the loss in output that came as a result of end users booking all their travel, we have moved on and now have assistants to do this for us. So in reality we have gone full circle. The work moved around the place, from person to person, and the balance of administrative and operational work changed too. In the mean time the vendors got richer and richer selling ever more solutions to wider and wider set of folks. Interesting, don’t you think?