Case Study: Improving customer contact for a retail client during peak trading period

Summary
In summer 2018, we worked with a retail client to transform its forecasting & planning capability. As a result, the abandonment rate on all inbound calls fell from 9.1% (peak period 2017) to 3.4% (peak period 2018). This enabled customers to gain service more reliably and quickly than in the previous year, and contributed to an improved commercial performance.

Client
The client is a well-established multi-channel home shopping retail business. It provides a service to over 1 million customers each year through a combination of direct marketing and online. Shoppers can choose to pay for their purchase within 28 days, or take advantage of a monthly credit programme. The business has contact centre operations in the UK and overseas - taking inbound calls, handling emails, web chatting and dealing with more complex customer correspondence.

Background
In the 2017 peak period (September to mid-December), customers often had to wait a long time for a call to be answered. The abandonment rate on all calls was a high 9.1%, with Customer Service calls in particular having an abandonment rate of 14.5%. It was very difficult for customers to get through to the contact centre, leading to enormous customer frustration, and this was harming revenue.

Further, rather than having a consistent email response throughout the day, advisors were sporadically deployed on and off emails due to poor within-day match between required resource and actual resource. This led to a poor customer experience, and too much management time was spent moving the advisors around.

It’s well known that competition is fierce in the retail sector. It was essential that customer calls were answered in a speedier manner, and improvements were required in time for the 2018 peak period.

Identification
The first stage of the project was to study existing capability, using our best practice forecasting and planning methodology. We identified many improvement opportunities, focussing on:

  • The forecasting models themselves
  • The process and tools used to review the forecasts with internal stakeholders to gain consensus
  • The visualisations within the weekly planning model
  • The in-house shifts compared with the customer demand profile
  • The interval profiling of the resource requirements given to the outsourcers

Improvement
The first phase took place from July to September 2018, with the aim of ensuring there was enough resource in place for peak period 2018. A new forecasting tool was created, including weekly forecasting models for each worktype. Some of these were explanative models, where the variation is explained by the business’s sales forecasts. Other worktypes were forecast using time series models – using seasonality and trend to predict future activity volumes.

Charts were created, and meetings were held with stakeholders to ensure that consensus was built. Further, a tool was created to track forecasting errors – this was a feedback loop to ensure that the business learned from the errors. Forecasting accuracy improved enormously, with relative improvements to the Mean Absolute Error of 57% for Orders and 26% for Customer Services.

A new weekly resource planning tool was created to take the forecasts – and other assumptions – to determine a resource requirement, against which resource could be planned. A large resource deficit was identified with the new model, but there was just enough time to review the in-house recruitment plan, and reshape the weekly outsourcer resource requirements before the busy weeks came. Both in-house and outsource centres went on to deliver the required resource to deliver the much-improved service levels. The net result was that overall inbound peak period abandonment rate fell from 9.1% to 3.4% year-on-year.


With peak trading out of the way, the second phase of the project took place in February to May 2019.

We built a scheduling tool, which enabled alternate scheduling models to be tested against within-day requirements. Optimum shifts were designed, which were then tweaked by management and during consultation, to ensure the optimum mix of employee goals and customer goals was met.

Also in this second phase, a method and process were devised to ensure that the within day outsourcer requirements reflected what was actually needed to maximise overall interval fit. We wrote these new planning methods into new contract schedules for the outsourcers.

The result of these second phase changes was that the within-day resource at the client more closely matched the customer requirement. This further improved the ability to answer calls promptly, and ensured that work on emails was spread more smoothly across the day. It also meant that Customer Services management spent less time on within-day workflow matters.

Summary
We helped the client transform capability in forecasting, weekly resource planning, scheduling and outsourcer planning, which contributed to an improved commercial performance for the business.

The Customer Services Director of the business said, "We have gained significant customer benefit from deploying best-practice methodologies, particularly focussed around improved service levels across the phone and email channels. Our structure and role profiles within the Planning and Forecasting function were also reviewed to enable us to adopt best practice across the contact centre estate"

This project was performed by Philip Stubbs for Drakelow Consulting, which is now incorporated within Atlantic Insight.

Contact Us

Please contact us if you would like to partner with us to identify and deliver improvements to your operation.

Email hello@atlanticinsight.com

Call +44 (0) 203 086 8004

Office: 130 Old Street, London EC1V 9BD

Click here to use the webform