Sometimes, it’s better to let the experts do what they do best.
A couple years ago, I made the decision I was going to tile the countertop in my kitchen. Prior to that, I created a tile backsplash in both of my bathrooms. I figured if I could do that, then the kitchen counter would be a natural progression. And, because I would do the work myself, I could spend a little more since I wasn’t going to pay for installation.
I figured the entire job, from start to finish, would take a couple of nights after work. It ended up taking a week. While I had no problem laying tiles in open spaces, I ran into difficulty cutting and placing the specialty tiles along the edges and corners. The end product wasn’t a disaster, but it’s not something I’d put in a DIY publication either.
Oh, and did I mention that the morning after I completed the tiling, I woke up with searing back pain that kept me in bed and away from work for three days? The bottom line is while I got the job done, it came with several costs. I spent more money and time on the project than I expected. I spent more on take-out as I kept the kitchen out of commission for longer than I expected. I ended up taking paid time off to recover. And, was the result comparable to the professionals? No. If I had a chance to do it over, I’d have left the job up to the experts.
Self-serve Business Intelligence (BI) may not come with the physical hazards of other DIY projects, but that doesn’t mean it can’t be as harrowing. While there are several self-service solutions out there, independent industry experts are concerned about the impacts and implications.
A recent Gartner report indicated that within the next 18 months, a majority of businesses would have access to self-service tools. However, the same report also indicated that 90 percent of those operations would lack the structure to foster successful BI initiatives.
There are many reasons why self-service BI fails. Here are the top seven:
1. The data is wrong (my numbers do not match yours)
Who owns the data, and where is it located? Who can vouch for its accuracy and make changes or updates when necessary? If multiple unskilled users have access to disparate data sets, or views of company data, then it’s impossible to generate and maintain a single source of truth for your data. And, with no clear workflows or processes to approve content built either, it’s impossible to trust the quality of the reports and dashboards produced.
2. Can’t use the tool (built for data experts, not for me)
What kind of interface does the tool offer? Is it intuitive to the user? Also, what kind of training is offered for the users, and where can you go if you have questions or encounter problems? Without sufficient support and guidance from experts, attempting to build reports and perform data analysis unassisted is a daunting prospect for non-technical business users. Without assistance, the more difficult it is for business people to use BI software – even if it is relatively ‘intuitive’. And, the more arduous the user experience is, the less likely they are to keep using it.
3. No context to the numbers (what are they telling me)
Many self-service BI tools lack the collaboration features needed to effectively connect business users with analysts to collectively interpret and understand what the numbers are telling them. Then, once you have that insight, does your BI tool empower you to share your insights with others? Context is crucial. Gaining context by being able to collaborate on and share your data insights allows you to ask ‘why’, not just ‘what’. ‘Why was there a spike in revenue?’ vs. ‘what was the growth in revenue?’ Without context, it's increasingly harder to effectively act on the data.
4. Who is responsible (someone needs to fix this)
Who do you go to if you find an error in the data set with which you’re working, or need additional material that you can’t find? As is often the case with the mainstream approach to self-service BI, business users have to act as their own data analyst and data quality steward. Not only is such data wrangling a complete distraction for business users, it demands skills most of them don’t have, leading to errors being made and increasing the risk of poor decisions being made from dodgy data. Data that’s not trustworthy will eventually stop being used.
5. It takes too long (I need to do my real job)
Dashboards look easy, but behind the façade are layers and layers of data and multiple reports. If you’re a business user trying to create your own dashboard, you might be able to create your own report, but pulling together data from multiple sources, creating multiple reports, and combining those in a coherent manner to form a dashboard? That’s a tough ask, and significant time spent not actually doing your day job. Without the proper support, many business users will make significant errors, revert to spreadsheets, or simply give up. Besides, when was the last time you met a sales manager who begged for the chance to compile their own reports from scratch?
6. Low adoption (no one is using this thing)
As Gartner has noted, the adoption rate for self-service BI tools has wallowed below 20 percent. If only one in five available users are taking advantage of the solution, the BI tool is as good as useless to most the business. Why is there low adoption in self-service BI? Because the DIY approach creates erroneous data, takes too long and eventually becomes too difficult. Self-service is instant gratification at the expense of long-term gain.
7. Difficult to access (why can’t I access via the web)
Part of the reason why adoption rates are low could be due to the difficulty business users experience when attempting to access self-service BI. While self-service BI tools appear to empower the business user, they often become decentralized desktop applications. If users cannot access their reports and analytics online, then they can’t utilize and act on their data when they’re away from the office - whenever and wherever they need those insights to make decisions. If a sales person can’t access their prospects data using a mobile device when travelling, they’re less likely to use it even when they’re back at their desk.
In the BI world, the expert is the data analyst. They work with data and the necessary tools every day. They save companies time and money by providing a service that’s necessary for organizational success. You can do BI yourself, but it comes at a cost. Are you willing to pay that cost down the road in an attempt to get a fast, but most likely false, start?