Initial insights on the impact of COVID-19 on global call volumes
This page is designed to provide data on trends we are seeing across our ecosystem. We will continue to update this page as new data become available.
We are happy for this data to be used to support research, and we ask that data is accredited to Afiniti.
In this analysis, we examine trends in call volumes in response to the closures and work from home policies implemented by governments in the US, UK, and China.
The data shows that companies have seen downward trends in sales and retention calls, with retention calls dropping more steeply than sales. These moves are highly correlated to increasing stringency of shelter in place policies. Despite this, abandon rates and wait times have spiked, as agent numbers drop by as much as 50% due to a mix of technical issues caused by working from home, transfers into other skills, and agent absences.
The drop in sales and retention call volumes is likely caused in part by customers being put-off by extended wait times. However, the significant reduction in ports and delta between sales and retention calls suggests that customers are more interested in securing their connectivity than risking being disconnected. Further, evidence from China and emerging trends in Italy suggest that these trajectories may reverse; preemptive sales made to secure connectivity will unwind and latent churn (especially due to cancellations in truck rolls) will return. This suggests that optimizing customer interactions will be key to retaining and up-selling customers, as the initial reaction to shelter in place policies abates.
Future analyses will focus on the predictivity of historical churn and propensity models, and how to reduce the rebound effect, as latent churn and temporary sales unwind.
Stabilization Phase Hypotheses:
Predictive Decay – predictive models are unlikely to be as effective due to step changes in historical behaviour, meaning new features, models, and potentially new techniques will have to be used.
“Tidying” Phase – anecdotal evidence from Italy, in which customers are beginning to optimize their contracts, leading to gradual increase in volumes as the initial collapse subsided.
Backbook Activation – previously inactive customers are calling in due to increased time and communication with relatives.
New Norm Hypotheses:
“Unwinding” Hypothesis – Customers upgraded their services ahead of isolation to ensure connectivity. This may be wound-down as the reaction phase diminishes.
“Latency Effect” – Customer retention has been suppressed due to:
Reduction in truck rolls, fear of moving, new deals, and moratorium on involuntary churn (see behavior of high churn risk customers).
Unwillingness to wait or frustration at wait times in retention queues.
Abandons and Repeat Call Features – repeat calls, high speed of answer, and abandons are likely to be leading features of future churn or lost revenue.
Removal of Temporary Discounts - as companies responded to the COVID-19 crisis, many removed caps or offered steep discounts for short periods to support their customers during the initial reaction phase. These will have to be removed as countries move on to stabilization and new norm phases, which can have a highly divergent impact on the customer base, with more affluent customers being 'hooked' on the higher data capacities, while other customers struggle to pay for full price contracts.
US sales volumes have ranged from 110% to 85% of normalized values and retention volumes have dropped between 2% and 46%, with abandon rates inversely correlated, gradually increased to 27% and 15% respectively.
Average speed of answer has increased by over 1000% for sales and 500% for retention, correlating inversely with agent counts.
COVID-19 Insights USA
Sales and Retention volumes recovered after an initial drop, but volumes have started dropping again.
Abandoned calls have been creeping up since the first restrictions on travel.
Agent count seems to be the leading factor in volume changes as well as spikes in customer wait times.
Calls = Calls received aggregated across our clients.
Total Agents = Agents that were logged on and received at least one call on the skill that day.
Daily Call Volume and Agent Counts for each day in March 2020 are normalized to each client’s weekday average in the period from January 1 2020 to February 29 2020.
Calls Received, Abandonment Rate and Total Agents are presented as the mean across clients.
Source: Government Stringency
Hale, Thomas and Samuel Webster (2020). Oxford COVID-19 Government Response Tracker.
Data use policy: Creative Commons Attribution CC BY standard.
Sources: Local governments; The Center for Systems Science and Engineering at Johns Hopkins University
This is the first in a series of COVID-19 Impact Insights that Afiniti’s Advanced Analytics team will be releasing to help our clients and global community navigate the changing pandemic environment and enable them to make data-driven decisions using emerging trends.
The Afiniti Advanced Analytics team is available to support our clients with any ad-hoc analyses, as well as best-practices on hypothesis testing, developing new strategies and models, and implementation and performance measurement.