Prototyping Data Products: Types and Technologies

As I mentioned in the previous article on data prototypes and development methodology , the best way to validate an idea is usually to make a prototype. It is the best way for the end user to get an idea of what he/she will have as a final deliverable.

This can explain the product better than any user analysis, research, data analysis, etc. and help you sell the product to your stakeholders.

In this article, I want to explain what types of prototypes there are, the difference between a prototype and an MVP, and tools that exist to develop them.

It’s a continuation of the previous article related to the methodology to implement data products, and part of the serie of how to implement a Data Product Strategy.

What is a prototype?

Let’s first review what a prototype can be:

• They can be documents, designs or schematics of the solution that give us an idea of what will be the final product.
• They simulate or have implemented parts of the final design, in order to see how it works.
• They are implementations or simulations that allow to identify the acceptance of the characteristics of use and design by the client/end user.
• They help to involve the user/customer in the evaluation of the product from the early stages of development.

These are the objectives that are usually covered to a greater or lesser extent by prototypes. This will also depend on the level of complexity of the prototypes.

Types

At the level of complexity, prototypes can be classified into two categories:

Non functional

This category includes prototypes that cannot be used by the user, but which cover some of the points described above. This type of prototypes:

• They are drawn or designed to show a representation of the final product.
• Graphics or mock-ups are usually used to represent to a certain extent what is going to be developed.
• It is possible to simulate certain behavior through different designs, trying to simulate the behavior that the products will have, but always limited to a static approach.

Functional

In this category we have the prototypes with which the user can interact, and that can implement one or several functionalities of the product:

• It allows to visualize a test version, which does not have all the functionalities implemented.
• Part of it can be reused to implement the productive version of the product.
• It is recommended that it is made up of the same technology of the final version, in order to be able to reuse most of what has been done.
• It must have at least the most important part that we want to highlight of the final product to be released.

Technology

In order to develop prototypes, there are several technologies on which we can rely to do so. Listed below are 3 of these, ordered from least to most complex and value-adding:

MockUps

This would be the basic tool. It is used to quickly “draw” what our data product will look like. It can be used to implement products that have a graphical representation (those we mentioned in the initial article as Decission Support), but it will be difficult to implement other types of products, such as “Decission Automation” products):

Source: https://balsamiq.com/learn/ui-control-guidelines/charts/

There is a very simple, mature and widely used tool on the market called Balsamiq that can help you a lot in this area.

Wireframes

The next level would be the so-called Wireframes. These can show the navigation and part of the product’s behavior. For the data area, the usefulness is limited, since the products are usually data representation or analytical models. Even so, if the product you are developing has a high navigation component, it can be quite useful.

Source: https://www.figma.com/community/file/966554326385625740/user-interview-report-model

An example of a tool to be used in this area, and a very widespread one, is Figma.

Functional Prototypes

Finally, we have the functional type, in which certain functionality is offered to the user. For this, tools are used with which an application can be easily prototyped.

Source: https://blog.streamlit.io/how-delta-dental-uses-streamlit-to-make-lightning-fast-decisions/

In the Python world, Streamlit is consolidating as the default tool for this. Easy to assemble visual interfaces, workflows with the user and also has connectors with any API or DB existing in the market.

Conclusion

In this article we have reviewed the concept of prototyping, how it can help us in our data products strategy and the existing types and technologies for it.

If you are interested in more information about Data Products, feel free to take a look at the articles I have about it.

If you want to give me your opinion about it, feel free to comment!