On the events trail, you hear proclamations galore about wondertech. You need a sharp BS filter, which means you need "go-to-smarties."
I'm not the only one. If you're sick of "data-driven" platitudes and want to get results out of data, Girardi is a guy you want to talk to. If you want your executives to see the impact of the data that your data analysts love to geek out, Girardi's your guy.
I first met Girardi at an Acumatica cloud ERP show (DataSelf is an Acumatica partner with a self-described mission of giving Acumatica users a "360 degree view of their business"). Girardi speaks to the data plumbing you need to make analytics integrations work. Given that BI vendors will never tell you their integration toolkits aren't seamless, that's a crucial conversation.
Overcoming the barriers to business user adoption of BI
So at this year's Tableau Conference, I finally put Girardi in the interview hot seat. My burning question, at a show filled with data geeks:
How do we get business users fully invested in the data conversation?
My past chats with Girardi were about technical nuts and bolts, but he surprised me this time. He pointed directly to the human disconnect:
Some of our latest insights into the BI world are actually about human behavior. When we go to events like this, it's all about us the data geeks. But we are a very small portion of the population.
We love data, we love charts, we love Tableau, we love all these things, but when you give this to decision makers who are usually not data analysts - they might be data-driven, but they're not data-analysts - there's a huge gap between what we do and how many these people get it.
Which brings us to the problem that has plagued so many BI projects: business user adoption.
This is, to me, one of the reasons why so many data projects have user traction issues... I was talking to one of my new people. He was going to train one of our clients, and they told him, "Well, guess what? We're going to be starting a new group of users."
You have some major human hurdles to get over. These users, they're not BI people. They might be salespeople who need to sell demand planning. They are there, not because they want it, but because their boss told them, "Hey, great news. You've got to be trained. You should be selling more planning," or whatever it is.
The first problem data geeks run into during BI training? Grasping a different agenda: "Business users go there with their own priorities, and we need to try to reel them in, right?" Example: don't overlook the log-in. Girardi has seen business users quickly drift to their phones and inboxes if they run into friction with system access. So his BI trainer made sure every single person was logged in properly. But that's just the beginning. If they don't quickly see data they care about, you've lost them:
Once you log them in, your next big challenge is you: you've got to show them something they care about. Not what we care about - but what they care about. So you have to ask the BI team at the client what to record, just to make the training relevant for everybody.
But training days fade quickly into the rearview. So there's one more crucial action you have to achieve: data subscription.
The very important thing is teaching them how to subscribe. Subscribe means that they'll get the email every day, or every week, with that report. Now even if these people never log back in, they will still remember.
Of course, opening a useful email is hardly the pinnacle of engagement. But for the BI team, it's a chance to prove their value:
They got an email in their inbox and one day, hopefully, that email will help them make better and they'll say, "Wow, you guys helped me. What else do you have for me?"
Girardi doesn't claim to have all the answers for setting business users on fire. But he sees progress:
We're starting to explore this human behavior problem that many of us techies don't pay much attention to. I think we are onto something to increase user interaction.
I brought up the power of alerts. Girardi isn't sure if helping business users configure alerts is the best approach for an early training, but he agreed that alerts are a piece of puzzle. Getting the right data notifications at the right time to spur actions is, in my view, just as powerful as a dashboard - and for some users, better. You don't have to remember to check your visuals.
Can single page analytics help?
But a vendor like DataSelf has to go beyond advisory: you have to build a solution that addresses the problem head-on. So what did Girardi's team do about it? For years, Girardi has obsessed on the problem of financial reporting - a problem that persists on many application platforms.
He's found that finance users and CFOs tend to get lost in data, losing the thread in drill downs. Or, they invest in a tool like Tableau, but the BI team struggles to deliver the right view. Girardi says the typical BI person that comes in and installs Tableau doesn't know how to set it up in a way that hooks finance users.
Typically, tools like Tableau, Power BI and Qlik, for P&L and balance sheets, are not designed for that purpose out of the box. They can do it, but it's not done very well.
DataSelf provides what Girardi calls "Single page analytics - Tableau with data warehousing." It's not a Tableau extension, it's what he calls - "Tableau functionality in a smarter way."
I told my team, "Let's come up with a way to do this whole analysis on a single page. They can see where it started, where you can drill down in all the filters on a single page. You can start with the P&L, you click you and go down - without leaving the main page. You can see all the clicks highlighted.
But there's a twist. When it comes to delivering value, DataSelf provides customers with views out of the box. That made me wonder: after all Girardi's talk about the need to deliver relevant/personal information, why would he offer out of the box visuals? The answer? Over the years, he's built literally thousands, to give his customers a head start:
If you have a Microsoft system, Sage, Acumatica, Salesforce, or whatnot, in hours or days, on a fixed budget, we can give you a tangible number of reports and dashboards, usually thousands of them, to cover sales, inventory, financials, payables, receivables, a bunch of things. So instead of taking two, three years to get there, we give you a fixed budget, fixed timeframe.
How customer feedback changed DataSelf's project approach
Customers can use Tableau or Power BI to access the visualizations. But how does this work in practice? Girardi says half of DataSelf's customers find that out of the box is all they need. Girardi has learned the hard way: don't go far down the road of custom BI projects until you give your clients a chance to see their data in a new way. He told me about a customer who gave him a big lesson on this. The CFO told Girardi:
In my first project, I asked for a very complex BI project that took you and IT five freaking months to complete. Anyhow, once you guys did your thing, it was awesome. You fixed my problem, my and that's how I got over the hump of my pain. Let me see what else DataSelf has, just out of the box.
After browsing out of the box report, the CFO made a discovery. Girardi picks up the story:
The CFO found a report showing some major inventory issues. If he had opened that report five months before, he would have said, "Forget my original priority, we have bigger fish to fry." Now he says, "I don't know what I don't know." He's totally out of the box today.
Now the CFO tells Girardi:
Let me sit on the driver's seat. Look at my own data. I understand data. This is bad data; this is good data. And let me, the decision maker, I'm paying you - we need you guys to do what I really think is going to be the priority.
Girardi learned his lesson. Let the data breathe - and spark discoveries. That may shift priorities.
Since this client, I always tell my new clients, "Do phase one out of the box. Let's put the few decision makers that are willing to look at their data out of the box together." Usually there's a bunch of things, so they can see their data, and tell us that's okay. Then I ask them, "Do you want to continue with phase two being whatever you told us before, or maybe there was something else that you found out?"
Girardi warns that this approach isn't for the faint of heart. You have to nail data security, integration, and other enterprise-grade concerns. Girardi's been cracking away at this for years. Now he believes he's turned a corner:
You know when you go to sell something but you're like, "It doesn't blow my socks off"? I think now I actually have one of the best financial frameworks out there, so I'm going to be kicking ass, pardon my French. I think it's going to make a big change in my market.
End note: I consider myself at least 50 percent data geek, so I feel comfortable using that term, and do not mean it in anything but an affectionate manner.