Selecting a Data Visualisation Tool: 10 Key Considerations
For organisations looking to implement a data visualisation tool, this usually means locating a tool capable of dealing with large data sets, in multiple locations, that can be collaborated on by key team members.
If you’re currently traversing the process of selecting a data visualisation tool, the good news is there are a number of quick and easy checks you can put in place throughout your selection process to ensure you choose an appropriate tool.
In this article we’ll detail 10 key considerations when selecting a data visualisation tool. From assessing the capabilities of your tool, to understanding your data, and future-proofing your selection for scalability; by the end of this article you’ll be well on your way to making an informed and cost-effective decision when it comes to your data visualisation needs.
1. Understanding your data volumes
These days everybody loves to throw around the term ‘Big Data’ like it’s common knowledge. However, when it comes to classifying your data, this can prove to be a complex task.
For example, commonly you’ll see estimates of what ‘Small, Mid-Sized and Big Data’ are based on the number of database rows your service has. i.e.
- Small Data: Less than 1,000,000 rows
- Mid-Sized Data: Greater than 1,000,000 rows
- Big Data: Greater than 20,000,000 rows
Estimates such as these aren’t particularly useful when selecting a data visualisation tool or planning how to store your data. Why? Because it’s simply not possible to compare apples with apples using this methodology.
For example, a database table with 100 columns will almost always contain more data than a table with 20 columns. Even if the larger table is perfectly optimised and uses advanced storage, classifying data based on row-sizes is an inaccurate approach.
If you’re looking for a more thorough means of classifying your data, we recommend taking into consideration:
- Your current and expected data size in gigabytes
- Whether your data is stored in a columnar or row-based format
- The query types you’re going to need to implement (i.e. Multi-Table Joins), and
- Any data compression you’re able to take advantage of
Sure the number of rows your data table have is important, but the volume of data in size and the way you use your data are two vital checks you can't ignore when it comes to speedy data access and product selection.
2. Check that your tool supports ‘Columnar’ data
If you’re new to database design, columnar database structures may be a new concept for you. Basically this data storage mechanism allows your queries to access smaller sectors of your database tables resulting in much quicker response times.
In the world of data analytics and visualisation, columnar database design is a must when setting up the basis of your platform. Our simple chart below shows the difference in required table cells required to be read in a columnar layout vs a row-based layout. As you can see using our simple 5x5 layout, if you’re looking to access just two columns of data, you can save 60% of resources right out-of-the-box which is certainly not to be sneezed at!
3. Evaluate the list of supported data formats
These days’ data comes in many different formats. From CSV, to SQL, JSON, Google Sheets, and Excel, the list of supported data formats that data visualisation tools support can vary greatly.
For example, during recent research undertaken by our team into the data visualisation space, we’ve seen tools which support as many as 40 different data types, whilst some only support a handful. This of course can be a major strength of the tool you select, or a serious hindrance. So next time you’re selecting a visualisation tool, double-check the support of common data types.
4. Make the most of your trial period
It’s surprising how many organisations select data visualisation tools without actually utilising their full trial period.
Each of the tools on the market vary greatly and offer a very different user experience, set of features and to put it simply either feel right or don’t. By taking the time to test a handful of data visualisation tools and read through the reviews of others who have used the tools you’re considering, you’re going to notice the areas of your tool you can either live with or without.
Free trial periods are also a great way to save a few dollars. Now this may not be much of a consideration to you if the product you’re considering offers month-by-month payment with a customer-friendly cancellation clause, but if you’re forced to pay up-front for a full year and find yourself with a solution which can't deliver the kinds of insights you’re looking for, this can become a very costly exercise.
5. Integrations with external systems
In software development circles, connectivity between systems is big business. Why? Because this capability allows you to use purpose-built tools to share data between systems without the duplication of data or having to export data from one system to another and re-invent the wheel.
Data sharing API layers such as those which allow JSON format data to be shared using RESTful API’s have for some time now become the industry-standard, so if you feel this may be a requirement for you, make sure you evaluate whether your tool can deliver this capability.
If you’re new to custom integrations, API’s (Application Programming Interface), or think you may have a need to connect your database records securely with another system or open them up to the public, feel free to get in contact with the OSE web development team.
6. Find a level of chart customisation you’re comfortable with
Depending on your audience, the level of customisation of your chart data (i.e. output) may shape which data visualisation tool you choose. For example, some customers require highly visual data with imagery, interactive features, and the ability to make dynamic comments; whereas others require nothing more than a simple chart you may see in Microsoft Excel.
This of course can drastically affect the price you pay, and there’s not point over paying for something you simply don’t need. However, ensuring you pick a tool which is future-proof can be tough in this case. As a guide, consider your industry. For example, if you’re in marketing, greater control over your visualisations may be much more advantageous than if you’re in server administration where raw data visualisations are king.
7. Export options
If you’re planning on never exporting that data, you may want to skip this section. However, if you’re looking to export your data visualisations for use on the web, in presentations, or to be shared in raw formats, then you better make sure you check these capabilities of your tool!
Commonly in the data visualisation industry you’ll see tools provide the ability to export data to Excel, a CSV or a PDF, but beyond this is where things get interesting. For example, solutions that offer a shareable link regularly provide this in an iFrame which may or may not suit you if you have a firewall up within your organisation. That being said, if your website doesn’t support CORS (Cross-Origin Resource Sharing), importing data from a remote server may also be a challenge.
Taking the time to understand whether a given tool supports your content goals and internal IT policies is an important aspect to selecting a data visualisation tool. After all, it’s what you can do with the output of your data that can drive real change in your business.
8. Cloud vs Local Server vs Desktop
Wow, it suddenly feels like it’s 1999 again, but the age-old battle of ‘Cloud vs Local Server vs Desktop’ is still a major point of discussion in the data visualisation space.
When you begin to research market options, you’ll soon find that some organisations support all of these alternatives, whilst others purely offer cloud solutions.
This of course means that the product you receive, although they’ll tell you they’re identical in every way, are likely to be somewhat different. For example, a desktop solution may allow you to work without an Internet connection, whilst a cloud solution is solely reliant upon an active Internet service.
That being said, the size of your data can affect the performance of your application and you should test your data at scale across these platforms if you’re not committed to any particular service. Personally, I’m a huge fan of cloud solutions, but one-size certainly doesn’t fit all when it comes to data visualisation solutions.
9. Data storage and localisation
If you’ve ever worked for a Government organisation or an internationally owned private entity, you’ll know that their data and security policies can be quite detailed.
For example, some policies may dictate that all data needs to be stored in the country of your company/department, whilst others dictate that all access to information must occur over secure channels locally.
This can be confronting for those who are new to the rules and regulations around security of data, especially if that data contains confidential or personally identifiable information. So, if you’re working for a major organisation, make sure you take the time to investigate if any data storage or localisation restrictions exist around your data, as this can significantly affect your ability to make a data visualisation tool purchase.
10. Price and user counts
One of the really interesting (and most important) aspects to selecting a data visualisation tool is how providers in this space price their services.
For example, some companies will price their service on a price per user basis, whilst others will offer a higher price point for unlimited users. Contrast this with some other providers and the number of users you purchase gives you access to volume discounts. Welcome to SaaS.
Price is obviously a concern for all organisations as you don’t want to pay more than you have to, but be warned...read the fine-print when it comes to data visualisation tools. Often a single-user setup can be quite restrictive, so options with unlimited user access can look quite expensive yet ultimately prove affordable when you begin using your data visualisation tool in production.
The other consideration around cost is the cost of hosting your data. If you’re using a database such as MySQL this will be relatively low as your data volumes will be moderate. However if you’re moving into the mid-sized or big data ranges, you may be considering services such as DynamoDB or Amazon RedShift which do have an associated cost worth assessing. Our advice, do your numbers. When it comes to selecting a data visualisation tool it’s all about the numbers when it comes to avoiding unexpected bills at the end of the month.
Selecting a data visualisation tool is an exciting process, and if carefully undertaken can result in many fantastic outcomes for your business.
Whether you’re looking to identify potential avenues for new business, find bottle-necks in your operations and optimise the way you work, or simply visualise your data, adding a dynamic presentation layer that can interpret and share this data with key decision makers is a service any organisation can benefit from.
If you’re looking to enhance the way you work with data, talk to the team at OSE for a confidential discussion.