Dashboards

Dashboards are great tools to assist in your day to day operations. While spreadsheets can be good starting points, a well built dashboard can bring you more benefits, including freeing you from the task altogether!

Dashboards can bring you one of the fastest wins you can get. Technically, they are not hard to set up. You are probably already familiar with the concept if you spreadsheets or other analytical tools (they can be your first version as well).

What is a dashboard ?

The easiest way to understand what a dashboard is is to look at your car's dashboard. Those various indicators located behind the steering wheel.

To make it simple, a dashboard is a set of data presented to you to help make decisions.

If the speed goes to high, you need to slow down. The engine sign shows in red, you're in trouble, ...

In the case of your business, a dashboard can help keep track of where you are, where you want, what's the status of various activities, ...

Why is it so useful ?

A well built dashboard is very useful. It gives you a in single view all the information you need to act and make decisions.

A great dashboard is something that gives you the answer you need as quickly as possible so that you can move on to something else.

It can range from a single metric report following a specific situation to a full blown decision making tree driving your business.

A well design and autonomous dashboard can also give you a great starting point to automation. It can help freeing you from performing tasks while still allowing a level of control if required.

What makes a good dashboard

There is no single way to build a dashboard, and a lot of factors can enter into account. However, to enhance your chances of building a great dashboard, you can follow some simple principles:

The goal of the dashboard

In the best world, each dashboard would have one goal only. It would assist you in making decisions related to one topic, one question only.

It is not your car dashboard where the goal is to have everything at once because changing view would distract you and introduce some safety risks. One your computer, phone, ... you can have multiple dashboard accessible within one click. and that's very likely fine.

Keep your dashboards simple, one dashboard one goal

Use appropriate renderers

Each piece of information can be displayed in many different ways depending on the context. If you stick to one goal per dashboard, it will be easy to determine how to best show what you need to know.

Let's take a simple example: today's temperature.

If your dashboard is built to help you decide which clothes to wear, you may be interested in knowing what is the range to expect and when. You want to know whether you'll wear trousers or shorts. Based on when you go out and come back, the temperature will vary, its value is important. In that case, you may want something like :

But, if what you need to know is if there will be ice on the road in the morning. Then, what matters to you is whether or not the temperature will be below 0°C then. Then, a Yes/No boolean display is more appropriate. It would give you the asnswer you need in that context without you having to infer it from a chart like above.

Ice / No Ice

Trust issues ?

I understand it, sometimes, you may find the data hard to believe for some reason.

You want to get deeper into the details of how a value gets calculated.

In that case, I recommend two options. Both aim at keeping the original dashboard as simple as you can. First one is to have an advanced mode that would display more, second is to link to the source data. In both cases, try to not pollute you dashboard with more than what's necessary.

The components of a dashboard

Presenting data

That's the visible part of the dashboard. The first thing you need to focus on as that will be what will save you time in the long term. Making sure everything is there in the most effective way.

Presenting your data has two important parts:

  • The layout: That's how you organize the visual rendering, how big is each part, which data points make sense to be clustered together, ...
  • The data point renderers: (See above), how each value is represented. Make sure that it gives you the answer you need at one glance when possible.

Retrieving data

Another essential part of the dashboard is the data itself. How do you get the value to display.

The first step may be manual, going to each source and copying the values from there into your system, but, you will want to move away from that as quickly as possible, if not from the start. Automating your data retrieval will be your greater ROI when done properly. Even more than how the data gets displayed.

You can benefit from even a simple data dump  (showing everything on the same page).

Let's imagine your goal is to have a dashboard to help you ordering some goods online. You may have multiple browser windows open, each one on a different product at your supplier(s), multiple windows open on your inventory levels and be switching from one to the other to identify which product, how many of each and from where you should order.

How much time can you save by having all one a single page. Scroll up and down the page between sections is already an improvement over looping through tabs to find the right one and then looking through pages for the right data point etc...

We may become more technical on other articles, but for this kind of task, APIs, scrapers, ... are useful tools.

Tell you the decision

That's the final thing to consider. Having the data displayed in a way that can help you decide is great. But, what if the dashboard could tell you what the decision should be instead?

If you can write down your decision process, what gets into it and what rules you follow, then having the decision made for you by the dashboard gets trivial. It already has all the data. It can answer your question without you having to think at all.

You save time, mental energy and money by focusing on activities that may be more lucrative or require more attention.

Which leads to the final point, if your dashboard can decide for you, can it also act on your behalf ?

Getting rid of the dashboard

It's the ultimate stage, can having the full process handled by the dashboard make the dashboard obsolete ?

After all, if it can do everything by itself, the last part remaining is you checking it and performing some action as a result.

There are ways to optimize both part of the process:

  • Have the dashboard notify you (push notification) instead of you having to go look at it. (We'll have an article on the subject). You just need to decide what is the notification trigger based on the dashboard goal and then have a process that will reevaluate the answer and tell you when you need to act.
  • Delegate the action either to someone else or automate it as well (you could remove being notified and not think about it for a while). Depending on what you need the dashboard for,it is worth exploring if you can avoid being in the process at all. You could then focus on other more important tasks or enjoy the freedom you created. Each option, delegating to someone else, automating or an in-between has its pros and cons, but a great step in improving your business and making it more efficient and less reliant on yourself.

Dashboards are great to assist in decision making. They can also be a great first step to automating or delegating processes that rely on you and your knowledge.

But, as great as they can be, make sure to have a sensible approach to building them. Decide what to do and how deep to go based on the work involved to get there and the expected ROI.

Some things are just not worth improving