No-code business intelligence means that you - the end user - do not need to code or write Excel macros in order to perform a given analysis.

It also means that Tom, the Power BI guy in the IT department, can now develop the CEO's new sales dashboard better, in less time, and with less need to be a DAX magician.

However, while this is great for Tom, it is not the core of the no code mania. Tom already knows how to code in DAX (it is his job after all), so making "BI no code for the coders" is nice but not all that exciting.

What is exciting is if no code BI finally empowers self service business intelligence. That is you.

Hey, but I already have self service BI!

Tableau and Power BI dashboards let me filter and visualize data as I please, unconstrained by the limits of older generation reports!

So why do you continue to use Excel?

Because dashboards are unable to answer all questions, especially the new questions that nobody asked before. A dashboard is a collection of a given set of charts that Tom chose at one point in time for a given analysis. Problem is that you, the user, might be interested in seeing different things in different circumstances, and today might need a view - a chart - that is different from the one that, last quarter, served you perfectly well for the "same" analysis.

One example, three similar chart types.

Stacked bar charts are ideal if the receiver is interested in a ranking of the biggest absolute numbers in total and/or a substructure. Normalized 100% stacked bars are best to compare the parts of a whole between different elements. Marimekko charts are best for comparing the parts of a whole between different elements and still seeing the absolute numbers of the elements as a kind of weighting.

What if Tom can't put all three in your dashboard? Tom can choose the best given today's data, but cannot guess which will work best with tomorrow's data.

Another limitation of dashboards is that they are visualization engines at their core. Not ideal - due to issues of coding and of testing/maintenance - to model complex calculations. Python and Excel are much better at that.

With no code BI you - the end user - can answer to that new business question that popped up this morning. You do not need to be an Excel master. You do not have to worry about bugs in your formulas. You do not have to worry that the chart you need is missing from the dashboard.

You also do not need to think about how to format your charts to make them pretty, clear and insightful. There are "good" and "bad" ways to set up a visualization. But they have already been discovered and documented. Trying to reinvent this particular wheel is not a good use of your time.

IBCS is an initiative that puts the best practices in structured data reporting together into a formal standard. You should absolutely visit their website. It is full of great suggestions.

One example: it makes no sense to label a column total as 1.234.567$. 1.2M will do fine.

Another: turning long column labels diagonal is legibility hell. Turn your column chart into a bar chart and leave your labels horizontal as they should be.


'Mparanza recognizes the limits of dashboards and the power of Excel/Python analysis. 'Mparanza automates what is generally done in Excel or Python and returns the result in seconds.

Spend your time on data interpretation, not on number crunching.


'Mparanza works for sales and costs analysis.

It requires a dataset in a flat 'tidy' format, probably the format of the dataset you already use for your sales analysis.