Guided vs Discovery vs Blending Information Products
by Shane GibsonMay 3, 2016
I have seen a massive change in the business intelligence, data and analytics tools our customers are buying and using over the last couple of years.
The advent of data wrangling tools such as Trifacta, open-source analytics tools such as R and visual exploration tools such as QlikSense.
Customers often struggle to decide what their current requirements are and what to buy. Once they have made the hard decision and selected a tool they struggle with what to deliver with the tool first.
For the visual exploration tools, I talk about three specific information product types they can deliver:
- Guided Information Products
- Discovery Information Products
- Blending Information Products
GUIDED INFORMATION PRODUCTS
Guided Information Products are designed to answer a small number of questions with ease. For example, who are my top customers, where are they located and how long have they been with us.
I categorise Guided Information Products as:
- They contain a small subset of data
- They provide minimal interactions
- They will provide one or two possible filters
- They are targeted at a large user audience
- They are targeted at non-data savvy users
These are the Information Products that you would typically hear people define as a “dashboard”. Although, the subject of what a dashboard can and cannot be these days is another longer blog.
DISCOVERY INFORMATION PRODUCTS
Discovery Information Products are designed to allow users to answer any question with ease, within a known data domain. For example, Who are my top customers by dollar value and margin grouped by gender and by regional location?
I categorise Discovery Information Products by:
- They contain a large subset of data
- They provide a large number of possible interactions
- They will provide a large number of possible filters
- They are targeted at a smaller user audience
- They are targeted at data-savvy users
These are the Information Products that you would typically hear people define as a “pivoting” or “slicing and dicing”.
BLENDING INFORMATION PRODUCTS
Blending Information Products are designed to allow users to answer any question with some effort, via accessing any available data.
I categorise Blending Information Products by:
- They contain no data, users access multiple sources of data themselves with the tool
- They might contain templates that help users access the data easier
- They do not provide any prebuilt interaction, users define the interaction themselves
- They do not provide any prebuilt filters, users define the filters themselves
- They are targeted at a tiny user audience
- They are targeted at advanced data-savvy users
These are the Information Products that you would typically hear people define as a “mashing data up” or “querying data”.
CHOICES, CHOICES, CHOICES
My recommendation is to pick one of these as your primary focus for delivery, it will help you pick the best tool for you for now. Most tools allow you to expand into the other style of Information Products, but will probably be limited, for now. But the tools vendors provide are moving faster than we can implement so there is a good chance they will provide the extra capability you need before you need it.
Change, learn or fade away, it’s your choice – Shane
Shane blogs about all of the things data and business intelligence.
You can read all of Shane’s blogs here.
Don’t forget, we can train your team in the art of agile business intelligence at any time!