Expanding the Field: The Data Solutions Space for Self-Service Business Users

Don’t forget the business users in the data debate! The most used data analytics tool is still Excel and there is a huge learning gap before mastering top-notch tech frameworks in the data space. I have always been an advocate of self-service analytics and the open source KNIME, based on many years of experience and massive adoption in the companies I work for. With this piece, I want to start a series of blog entries about this data movement. Today’s starting point is about the business user’s options and their learning curves vs. addressable business complexity.

It is always the same starting point for the business user. And with business user, I mean the huge number of employees in controlling, logistics, sales, HR, and other departments who don’t have IT or software programming backgrounds. These users all have, and all know one tool very well: Microsoft Excel. It’s a great tool and will always play its role in the data space, I believe. But my observation is that

➕ the biggest advantage of Excel is the near 0 entry barrier to learn and accomplish first number crunching tasks,

however this is offset by two disadvantages that

➖ business users need to put a lot of logic into combining the limited functions to succeed just little more complex tasks, and

➖ the limits are hit quite fast: number of rows, processing time, complexity.

One day, a user told me her anecdote that a finance manager summoned his staff to a meeting in 15 minutes. The ‘reporting guy’ answered:

“Boss, that’s impossible.”
“Why?”
“I usually open my spreadsheets 25 minutes before meetings for everything to load. I won’t be able to present any figures in 15 minutes.”

A definite example of too much formula complexity and data volume in the wrong tool for these purposes.

But fortunately, we have found KNIME. It’s open source and we could hence vastly adopt it (4 digit numbers of users) at zero license costs and very reasonable internal maintenance effort. With either KNIME’s great generic training content freely available online or our target-audience adjusted training courses with real-life examples from the respective industry, learners make fast progress. It will require a couple more hours to feel comfortable than the ‘instant start’-feeling of learning Excel. But once users master the most important set of commands, they can already achieve things of the complexity that spreadsheets see their limits in.

Users then quickly progress into the ‘expanding the field’ area: With learning more functionality of KNIME, they create fully automatic, sophisticated data workflows that solve complex business problems. This article isn’t about use cases, but to give you an impression: We are talking about e.g. automating the budgeting process with 10 different ERP data sources and dozens of special rules that need to be considered.

At some point, even the great Big Data integrations of KNIME won’t solve the dilemma that spreadsheets were in at a much earlier point of time: It’s not a tool that will replace an entire IT landscape.

Of course, at some point in time there is only the good old programming language and the typical tech stacks that will scale even to the highest complexities of business requirements. But before you will reach this point, your business users will already have closed a huge gap between Excel as the previously solely available tool and finding a programming guru who can code stuff for you.

I forgot to mention: Business users hate programming. They often don’t want to become the IT nerd and most will never dare to learn a programming language. With great visual self-service analytics tool like KNIME, I can understand this. The skill level and need for training to solve the first programming projects is just too high. Even with easy to learn languages like Python and the great documentation out there. KNIME is much closer to the admirably intuitive Excel in this regard.

Stay tuned for more to follow: typical use cases, how to nurture a data movement, why artificial intelligence (AI) is not needed to solve most users’ data challenges, how self-service analytics compares to Business Intelligence (BI), …