Data best practices vlog with BI specialist, Joseph Hobbs. Season 3 Episode 4 Hobbs talks about how to free up time for complex business decisions by choosing automation over reporting.

Video Transcript

Hello everyone, welcome back to TNT. My name is Hobbs, this is Valorem Reply, and we're going to talk about business intelligence best practices. Specifically, the question I want you to ponder today is whether or not a report is needed at all or if all that's needed is some automation.

Welcome back. My name is Hobbs, and I would like to share an experience that I had recently. I was working with one of my clients and we were talking about reporting needs. Reporting is a really common request. You've got a person who needs to make a business decision and they understand that they need to see these particular data points, these KPIs, these metrics in order to make the best possible decision. And as I was interacting with this client I started thinking, they have a really clear grasp, I mean a really remarkably clear grasp of what they want to happen and what the standards are to cause that action to occur.

So, I proposed to them, just like I'm going to propose to you now, that maybe what they needed was not reporting. Maybe what they needed was automation. We've got the ability pretty straightforwardly now. And again, I spent a lot of my time in the [Microsoft] Power Platform, so that's where my brain goes. You’ll have your own tool sets so you can design this in whatever tools you're comfortable with. But you can take a data-driven question that has, based on the results, a particular set of actions you want to take. And you can really automate that process beginning to end such that a person never needs to look at a report to decide what needs to happen. Now that's sort of, it's a long ways down [in data and business process maturity], and that's a lot of trust in your data pipeline, right? You're saying, ‘I trust the quality of this data well enough that a person doesn't necessarily need to look at it. I know that if A is here, then I need to do actions one, two, or three. And if B is here, then we're going to take this other set of actions.’

There is sort of a middle ground as well and that is reporting with alerts. Here's what I mean by that. Let's say that you’re a manager. You've got a certain number of employees and you've built a report that monitors their performance, and their employee satisfaction, and all of these things. And instead of going to that report every day, looking at it and saying, ‘OK this person is happy, this person is happy, this person is happy.’ [Then] close it. Come back tomorrow. Open it up again. Recheck those few people. Close it. Instead of repeating this process over and over and over and over again, what you really want to know is did something go wrong or possibly very right? This then becomes a matter of automation.

You could say, I want to receive an email or an alert on my phone if any employees’ happiness falls below a given point. And otherwise, I never want to look at this report again, right? My employees are doing OK, I don't necessarily need to dig in any further. And it's not specific to employees, you could have it be stores, you could have it be any number of pieces of any number of different enterprises and businesses. But that core concept to say a person's time is pretty limited. The time you've got to dedicate to reading through reports, really can be quite small. And there's much too much data out there. So, what if instead, every place you knew what you wanted to happen and when it would happen, you just automated that. And let the computer do its thing, do what it does well. Anything that was in the middle, where you kind of knew that you wanted eyes on this, you weren't ready to trust full automation yet. In that case, you build out a series of alerts to let you know if and only if you need to look at something. That way the report is there, but you don't care about it until something either goes right or wrong, depending on your business scenario.

And then finally on the other side, you would have the reports themselves. And this is a scenario where no amount of automation is going to help you out. The decision is complex enough, or there's enough variables involved that you don't really feel you could successfully offload this to a truly automatic process. A person is needed to deeply look into an issue. And that's where you got into reporting. To sort of summarize and encapsulate that, whenever possible, automate stuff. Let computers do what they do well, which is repetitive, clearly defined processes, right? And then that lets you have some time back to focus on areas where perhaps a broader perspective is needed than what a computer could give it. And then in cases where truly there's not even a middle ground, where you can only look if there's an alert and something has gone right or wrong, but you just need to have eyes on something every day, then and only then start focusing on reporting.

Thank you for joining me today, I hope you enjoyed this video. If you've got any comments for me, if there's a topic you'd like me to talk about, by all means let me know in the chat below. If you want to reach out and find me on LinkedIn, you certainly can. Our company website has lots of information about Valorem Reply and we would really love to be able to do business with you if there's a way that we could partner and help you be successful in a business intelligence project. See you all next time.