The most common technical errors in Business dashboards to watch out for

The most common technical errors in Business dashboards to watch out for

Since the pandemic, there has been a meteoric rise in the adoption of business dashboards.  It makes sense.  Businesses want to understand more about their businesses and be able to ask a whole new range of questions about their corporate data. When building business intelligence solutions, these are the most common technical errors in Business dashboards to watch out for.

Many organisations have created their own dashboards using one of the off the shelf dashboarding tools.  It’s a great path to take because it’s cost effective, it’s relatively easy to get started and an outcome is usually front of mind.

The most common technical errors in Business dashboards

As businesses become more proficient with their chosen dashboarding tool, they start to hit some common hurdles:

  • Is your data model incorrect?
  • Are your data feeds automated?
  • Is your data out of sync?
  • Have you defined its purpose?
  • Are you using the right graphs?

Is your data model incorrect?

Anyone can build a dashboard with a single data source.  This could be an Excel spreadsheet or even a connection to a single table from a database.

But dashboards really come to life and really add value when data is blended together or merging with other data sources.

For example, you might bring your sales data into the dashboard and then connect it to your staff data to look at the performance of staff against revenue.

Defining the data model is a really crucial step when creating a dashboard like this.  It is commonly thought of as a simple step.  It’s not always.  This is called Data Engineering and it’s emerging as its own profession.  There are some very clever professionals that deal with this challenge explicitly.

Justifiably so.  It’s the foundation of a successful dashboard.  Making sure that all of the data is brought into the data model correctly is critical.

Typically, most of the data sources that are blended together are not connected in their incumbent systems. There’s an inherent knowledge in the business about how they’re connected but the systems may not have that process implemented.

Making sure that logic is persisted into the data correctly and accurately is very important.

Are your data feeds automated?

Probably one of the most disappointing phases for anyone that goes through after they’ve completed a dashboard is to realise that they have to maintain the data feeds for their dashboard.

Creating a dashboard from scratch usually involves working with static data extracts and bringing in a whole range of information from different systems through something like Excel, or flat files.  This helps keep things fluid as the data model is defined and established.

When you publish the dashboard, the expectation is it’s going to be up to date, at least on a monthly basis. This expectation becomes a whole maintenance process around making sure that data is available.

If you haven’t automated your data feeds, then the burden that this dashboard is going to give you is quite excessive. And one that might be a showstopper as far as the dashboard being relevant.

Is your data out of sync?

You have your data model sorted, and you have automated your data feeds.

Have you considered the timing of these data feeds? They can be a real problem if they’re not thought through cleverly.

We hear a lot of talk about real time data analytics.  But there isn’t an agreement on what ‘real time’ means.  It means different things to different people.

Real Time financial data, for example, could be your live accounts presenting on a dashboard. But to be frank, because of the importance of financial data and the audit process that goes around it, your finance numbers probably aren’t going to be signed off until the end of each month. So while she can pull in live data from the finance system, you probably shouldn’t.

So the speed of finance data is determined by its governance, its quality.

Similarly, HR data might revolve around timesheets.  These could be updated on a weekly basis. But again, there might also be an approval process to make sure that they’re accurate.  Surely you shouldn’t present data that hasn’t been audited.

And then consider sales data. Sales data could be available as frequently as daily. It could even be real time and as it happens.

Again, there’s probably a QA process around this that needs to be put in place.

Either way, your dashboard is drawing in data from different data sources that have a different QA timing process, then a great deal of care has to be put together about what you present when the rule of thumb is the lowest common denominator for this the slowest data feed.

Have you defined its purpose?

Every dashboard needs a purpose.

Do you know the reason your dashboard exists?  Most dashboard projects start out with a laser focussed objective in mind.  During the development phase, excitement takes over and every nut and bolt is added to the dashboard to ‘provide additional context’.  What is really happening is the focus of the dashboard is being diluted.  The ‘reason’ becomes unclear.

If there is no specific reason why this dashboard exists, then revisit the point of coding in the first place.

For example, does your dashboard present corporate KPIs? If so, who is it for?

If it’s a sales dashboard, what key information has been delivered from this dashboard? Is it what your team really want to know?

Make the purpose a core part of the development process of your dashboard.

Are you using the right graphs?

If you have a pie chart on your dashboard then throw it out and start again!  Pie charts are garbage!

Data can be presented in many different ways (even more reason not to use pie charts). There are so many different graphs types to choose from and each has its own purpose.

However, most people gravitate to the more commonly known charts such as bar, line and pie.

These are all valid, but are they appropriate?  Download your guide on How to choose the right graph

You must choose the right graph in order to get your point across effectively and efficiently.  Your aim is to press home the importance of the data in your dashboard. It’s crucial to use the correct graph.

In summary

Many organisations are building dashboards and are producing great work.  I hope you are one of these.  But at the same time, many are coming unstuck at the mercy of the most common technical errors in Business dashboards.  I hope you don’t fall prey to these common technical errors.