Training Video

This support page has been built as a guide to help you understand the OI classification of customer contact.

You will learn about:

OI trains custom models to analyze the unstructured customer language from your service interactions. The models are trained to identify why customers are contacting, in their own words.

Think of it like automated ‘dispositions’ or ‘wrap up codes’ except far more granular and accurate. The technology analyzes the words of the customer and then automatically labels the reason for contact.

Models can be trained for different functional areas and channels, including voice (phone calls), email, chat, messaging, social media etc.

The benefit of this approach is that you can hear from every single customer and understand, automatically and at scale, why customers are reaching out to you.

OI trains the models to automatically identify a Level 1 (L1) - Level 3 (L3) understanding of why customers are contacting. L1 is the most aggregated view of why customers are contacting and L3 is the most granular view of why customers are contacting. In OI, the levels map to the following pages:

For every Root Cause, there are associated Inquiries and Sub-Inquiries. Below is a graphical representation of the relationship between the levels.

Within the OI platform there is a page to view:

Within the Root Causes page, you can see the Root Cause reasons why customers are contacting. This is the most aggregated view of ‘like’ inquiries and provides the context around what’s driving the customer to reach out.

These categories are determined based only on the language that customers use when they’re contacting.

In the above example, the single biggest Root Cause reason why customers are contacting is because ‘I’m having an issue with’ something. When you drill down on this category you can see all of the specific issues, however this view shows the impact of these contacts at scale.

Total Interaction Cost: note, this is quantified based on the average frontline salary cost, it typically represents agent costs, which is cost to service.

Use Cases:

When you click on a Root Cause, you can then see all of the associated Inquiries on the next page:

In this example, we can see the single biggest ‘issue’ is ‘logging into my internet banking/app’.

The labels are designed to be read as if a customer is saying it, in their own words.

When you click on an Inquiry in this view, you can then see the Sub-Inquiries, which are the most granular reasons why customers are contacting.

If you click on a sub inquiry, you can then take a step to look at actual interactions for that contact type and view the actual data.

If you click ‘view data sample’ you can then see actual interactions that have this classification:

If you click on one of the boxes, you can see the interaction and automated labeling by OI:

You have now learned how to navigate Root Causes and drill all the way down to individual interactions.

When you click on the Inquiries page you can see all of the individual contact drivers. This is what contact centers are typically most used to looking at in understanding contact reasons.

When you hover over an Inquiry, OI will also show you the Root Cause associated with that contact, so you never lose that context.

You can also click the ‘See more’ button to get additional data and metrics about that specific inquiry, including the Avg Cost, AHT and Customer Sentiment.

Click on the filters bar on the left hand side of the screen.

Click on the icon with 3 horizontal lines.

The results are then filtered (below is showing a Top 10 filter)

You can click on any Inquiry to see the associated Sub-Inquiries, which are all the different ways customers are articulating their needs or issues.

To see additional information about Inquiries, select the ‘detailed’ view by clicking on the following box via the left hand filters pane.

You will now see the page expand:

Note: use the Time Series at the top of the page if you want to change the dates or time period of the interactions that have been analyzed.

If you scroll down the page, you will see how Inquiries are changing over time, the following view has the ‘weekly’ view enabled.

Now you can see exactly how Inquiries are changing over time, which is super helpful for identifying emerging trends or quantifying impacts to the contact center.

You can also filter to an individual or group of Inquiries, using the filters section.

By selecting a single or multiple Inquiries, the view will then filter to show you that result. In the below example, we’re now looking at how customers calling to ‘Process a payment’ changes.

Finally, at the bottom of the page, you can see the volume of Inquiries based on a different Queue or Location.

You now know how to navigate the Inquiries page.

The Sub-Inquiries page has the same functionality as the Inquiries page, however it is showing you all of the L3 Sub-Inquiries, which is the most granular view of why customers are contacting.

Note: for any given user of OI, there will be hundreds of Sub-Inquiries, so we recommend filtering the view to see the Top 10.

When you hover over a Sub-Inquiry, you will see which Inquiry and Root Cause is associated with it, so that you don’t miss out on that valuable context.

Note: if you click on ‘Export CSV’ you can download all of the data in that specific view. For more information on exporting chart data see here.