How to own Data Preparation to ensure a successful Dashboard Project


Table of Contents

computer-1245714_1920-1500x996 (1)

I work as a Delivery Manager at Toucan Toco, and managing dashboard projects is my everyday job. I wrote this article so you could spare thousands of hours on managing endless dashboard projects. Then you could dedicate your precious time too long term benefits: adoption, ROI and data ownership.

#1 When it comes to dashboards, the world is flat…

Working on data projects made me realize an unexpected thing: the world is flat. I am not trying to fake-news you here, the world is flat for real. At least when it comes to reporting data and build dashboards. Let me explain.


When I joined the industry years ago, I was expecting to deal with super-connected companies, gathering trillion lines of data on gigantic databases. Cubes. Multi-dimensional data system, series of tables based on its self-structure. Well, if most of them have actually built their cubes today, we won’t directly use these cubes to create our client’s new dashboards. Why?

Because when we manage dashboard projects, we always do it with 2D data. This is 2 dimensions data (columns and lines) that most people use for their graphics or Powerpoint presentation. They are simply used to this way of handling data, as there is all they need in this 2D format: crossed results of lines and columns they want to focus on.

Bear with me: our mission at Toucan Toco is to help you building clear and simple data stories for those who use your reporting.

Who? Board members, C-levels, operational business users, our applications should be understandable by everyone.

How? By representing the data simply: in two dimensions.

Trust me, nobody wants to lose their time trying to decipher this kind of chart every time they need key take-away info.

At Toucan Toco, we took the simplicity stand, and always query the client’s data in two dimensions (column by column, line by line). Because that’s what serves best the end-users need, and help them to take actions based on the data they see.

Super easily understandable data. Immediate food for thoughts and actions. Like a pizza slice.

This is why, most of the time, when I run a dashboard project and connect data to our Dataviz tool, I end up using excel files/csv files. Flat data extracts. No multidimensional databases.

Just. Flat. Files.

#2 … and quite often, it’s a bit blurry for those who are using this data to track their activity

More than flat, the data world is also a bit blurry for most of our clients. Why?

Well, first of all, because that’s not their job. To get data from cubes/databases, you need to know its structure, dimensions, hierarchy. Accessing complex data structure, explore trillions of lines requires a specific set of technical skills and a strong data culture. Inside a company, only a few people can access, understand, and dig into databases. These people are data scientists, system administrators, data analysts, IT experts, BI managers, etc.

They work in the Business Intelligence department (BI) of the companies. And usually, they stand a bit far from my business owners (finance manager, marketing director, etc). Appears that these are the ones who would like to use a dataviz tool to track their activity.

That’s when you ask me “And so what? What’s the consequence of all that?”

In most cases, that means one thing: the content of this data (how it’s “made”, collected and calculated) is often unknown by our business interlocutors. Although they use this data daily. Usually, when people start a dataviz/dashboard project, they know which KPIs they want to see, but do not know how they are calculated.

the figures our business owners are used to seeing on your monthly Power-Point presentation has made a super long journey before getting there.

#3 Finally, what happens when you don’t know your own data?

Your project fails. “How do you know the project is a failure?” you’d ask. Well, if your dashboard is not used, as simple as that. Usage is key. And the milestone of this usage is the quality of data. If you don’t ensure good data quality, by running carefully your data preparation, you can be pretty certain that everything built on top will crumble.

And this will cost you energy, time, money, even credibility.

What are the benefits of a good data prep?

So first, let’s think usage, therefore data! Internal resources can handle data preparation (cleaning, treatment, formatting, and extraction of your data). Or you can externalize it to an integrator partner. In both cases, you better manage this data prep phase carefully, as it can lead you to several unwelcome situations.



  • Accelerate and secure data check phases

How to ensure a super clean, fast and efficient data checking? Anticipate! To check that all calculations have been reviewed, approved and that the figures are 100% correct, it is necessary to :

– Write the exact calculation rules: “My benefits = my expenses — my charges”

– Identify several test scenarios: “When I cross this KPI with this date, I expect this result”

– Identify which data tables are tested, in which environment: “We will request the SQL base XX-1, in a live data context, with a low internet connection”

– Dedicate time and valuable human resources to carry this data checking phase, identify mistakes and bring solutions

– Create data checking booklets, use them and version them to measure progress in one glance

– Once your business rules are locked down, time to think of the future.

  • Prioritize maintenance and knowledge

Be a responsible owner of business figures and calculation rules. To be so, you need to set up strategic reviews, which require to keep control of the data journey.

Documentation is key to ensure knowledge transmission and control of data. Because what’s done by the data owners (partner/internal services) is not always documented and understandable by the people who will take over, you absolutely need to write and store this knowledge somewhere. It will help you transferring information, implement evolutions to your dashboard. And at the time you’ll need to apply evolutions on your dataviz tool, by recalculating/adding new KPIs, you’ll be ready to face this serenely!

  • Control the scope of evolutions and the budget allocation

Let’s make an assumption: you don’t know how your data is calculated within your company data system, but you managed to build a dashboard tool, and you want now to implement evolution to your dashboard. When you try to add a new KPI, or to modify the calculation of one of your metrics, you’ll have to dig back into your data. And if nothing is documented, you’ll face a black box. You’ll lose your energy trying to gather all the info you need, and lose time preparing the implementation of the upcoming evolutions. And time is money. Don’t waste your budget in lightening the black box. Dedicate your time and energy to implement changes to your dashboard.

  • Basically, it forces you to go back to a “project mode”, lose more time and to spend more money.

Buy new data sprints. Mobilize new resources to get the treatment done. Other resources to proceed the full data integration and checking. And with all that, you might even miss your launching deadline…




  • Data as a usage booster

Making sure that the data restituted in your dataviz tool is correct and fits the Business need is a very essential step toward a successful data project. You need to show the right KPIs to those who take actions based on it, to ensure usage and actionability of your tool. Also, you need to make sure that each time you will update the data, you will not baised its interpretation. Data has to be readable durably by your dashboard’s users, KPIs have to be calculated the same way this quarter than 2 years before, to allow comparison. As your end users will base their decisions/actions on what’s displayed on your dashboard, what you show better be accurate!

  • With clean data, you’ll get your users’ trust

If you don’t show the right figures to your users, they will notice it immediately. They won’t take your dashboarding tool for granted, and just stop using it. That means months of work building a tool are going in the garbage bin. And they won’t adopt/use the app anymore. Because no usage = no ROI = project fail. At the opposite, by showing quality data, you’re bringing trust in the game, and get more usage at the end! Because your user has faith in the figures they see every day in your dashboard, they can serenely take actions — all these thanks to a careful data prep!


Enjoy digging into the benefits of good data prep? Feel free to like this post and share your best practices with us in the comments!


And always bear in mind that even if data work can be challenging, it is worth it! It brings usage to a dashboard, and this is priceless 👑

Get a demo

This is a heading 2

This is a paragraph

This is a heading 2

This is a paragraph

Table of Contents