Telecom Giant Solves Checkout Promo Code Error

Challenge

Their analytics team knew they needed to act fast when customers started reporting, via a Qualtrics survey, that they were unable to use promo codes they had received.

This was during peak season and each day the issue went unresolved it impacted NPS scores, decreased customer satisfaction and jeopardized potential sales. But with survey data alone, they lacked the contextual information they needed to validate, escalate and prioritize the issue and then locate, reproduce and resolve it.

Missing details

Was this a user error (i.e. they were simply putting in the wrong code)? Or a technical error (i.e the promo code system was generating invalid codes or there was a faulty API validation)? This lack of information risked increasing the time spent diagnosing, reproducing and resolving the issue.

Missing context

It wasn’t clear how many promo codes and product checkouts were affected, how many customers were experiencing the same struggle or its effect on revenue. This hindered their ability to route the case to the right team and provide them objective reasons to prioritize it.

Missing visibility

There was no simple way to track user flow and measure the effectiveness of any fix they would implement. Without the ability to monitor the fix, their VoC analysts couldn’t be confident they had solved the root cause of this issue.

Solution

The analytics team leveraged the deep link integration between Glassbox and Qualtrics to answer three key questions quickly and accurately: why were users struggling, what was the impact and how effective was the resolution?

Understand the struggle

Using the deep link, they easily located and reproduced the error by replaying the sessions with customer feedback. By searching the replay activity tree for text in the customer message, such as “expired,” they were able to jump directly to the struggle. After watching a few replays, the VoC team reproduced the issue. Customers were receiving promotional codes from a specific campaign that were already expired. By sharing this critical information with the appropriate team the time to resolution was dramatically reduced.

Prioritize by impact

They then created an ad hoc funnel showing the journeys featuring this error. Knowing the average order amount, how many customers were affected and the increase in error rate allowed them to measure the true impact—over $10,000 in lost sales in one week alone. Armed with these measurements, they were able to assign a high priority to the case.

Monitoring changes

Monitoring the user flow showed an increase in customers successfully completing this step. Furthermore, to ensure they had visibility into a potential recurrence of this issue they implemented a threshold alert The alert triggers when the error rate goes above their 3% baseline, giving them the confidence that they will know in real-time if the issue resurfaced.