Analyses are at the heart of Ipiphany. They store, organise, and present the insights that are found within your data. 

Analyses come in all shapes and sizes but a popular type in Ipiphany is called a Painpoints Analysis. A painpoints analysis is simply a list of specific problems that customers are experiencing. When painpoint analyses are done well and frequently they are very powerful tools for business improvement.

The following is a step by step process for creating a simple Painpoints analysis.

2.1 - Create the analysis

From the Ipiphany home screen click the 'New' button and select 'Analysis'

Choose the 'Painpoints and Highlights Analysis' type 

Select a dataset - for this example select the 'Airline Reviews' dataset which contains more than 8000 airline customer reviews

Give your analysis a name and click CREATE ANALYSIS. Don't worry about changing the metric, filter, or project at this stage.

2.2 - Find and add painpoints and highlights to the analysis

When the analysis is created you are displayed the Insights > Edit screen. This screen is the easiest and fastest way to build up a list of insights (painpoints & highlights in this example) but remember that this is just the start. Ipiphany contains many other powerful tools for extracting insight.

On the right side of the screen are all the things that Ipiphany has found in your text data. For this example, we will look at the default 'CONCEPTS' which are colored by 'Difference'. You should see the following:

'Concepts' are buckets of records that all talk about a specific idea. By coloring them by Difference (a measure of the difference between the concepts score and the overall score) you can quickly identify painpoints colored red or highlights colored green.

Now all you need to do is drag the bright red or bright green concepts into the analysis list area to create them as insights. If two or more are the same just drag them onto an existing insight and Ipiphany will intelligently combine the results.

2.3 - Tidy up your insights

At this stage, you have discovered some insights but they are poor quality. It is important to do the work to tidy up your insights so they can be quickly and efficiently used by business stakeholders.

A good quality insight needs to have a few things:

  • To be well named

  • To have a detailed description

  • To have some aspects

Let's take our highlight 'Good Flight' and see if we can tidy it up.

Click on 'Good Flight' and two things will happen, it will expand to show its available aspect categories and the insight 'Good Flight' will be previewed in the right hand side preview panel.

As you can see, we chose to group a range of concepts all under the insight of 'Good Flight'. Some of these concepts are better added as Benefits, Causes, or Improvements so to fix this we simply drag the concepts down into the aspect sections.

It is also a good time to remove any concepts that are better represented by other related concepts, in this case we removed 'Food' as it was more accurately represented in the highlight as 'Good Food'.

Aspects that are created as 'Benefits', 'Causes', or 'Problems' are automatically combined to create an insight description so leave the descriptions for now.

Now go ahead and do this for the other insights in your analysis.

Done? Good, now on to:
Step 3 - View and share analyses

Did this answer your question?