Successful back office planning teams are a brilliant source of expertise on workflow, process times, managing skills and reducing failure demand (see box) and there are abundant opportunities to spread this best practice across all kinds of end-to-end planning as well as existing back office operations where resource planning is still relatively undeveloped. It’s notable that in several areas of the planning cycle, the dedicated back office planning team at Santander Operations stands out as Best-in-Class. Looking back 20 years, the back office was an operational culture very distant from the contact centre and not always attractive to up-and-coming planners. Looking ahead however, the rapid pace of digital transformation beckons in a very different future, supported by growing adoption of digital journeys by all kinds of customer, especially post-COVID-19 (see pg 68). It’s time to take a fresh look at how we plan for processing operations!
Understand the math
Best practice in planning always starts with ‘understanding the math’ as they say in the USA. In call centres and face-to-face operations, Erlang and Queuing Theory help you allow for random variability around predictable workload patterns. For planners used to this, the maths of the back office can feel very alien, based as it is on the fact that backlogs can spiral out of control very quickly. If you don’t understand statistical and mathematical concepts like exponential growth or compound interest, you will struggle in planning or managing back office and processing operations. You also need to analyse process times and variability (see box) and your capacity planning will be as much supply-led as demand-led, still less common in the contact centre.
Transformation & capacity control
A processing operation needs clear capacity control. To plan for a process, you must know exactly what steps need completing and how this is undertaken. Famous from our earliest training workshops 15+ years ago, the Waterfall Model shows how to prevent backlogs spiralling out of control. With process times, some variability is predictable, some is not: you really need to understand this! Rarely do we stumble on this the first time, but we need it going forwards and there’s huge potential in the digital world. Fortunately, there’s a wealth of best practice to learn from (see box) and, at The Forum, we can help via learning modules, workshops or consultancy and our end-to-end network for planning pioneers.
Many back office case studies focus on the initial transformation that creates eye-catching benefits, as backlogs come under control, and some organisations outsource legacy backlogs to give a clean start for the mainstream teams. Even more crucial is the business-as-usual cycle that prevents future backlogs spiralling. Both stages can be seen at Legal & General (pg 38), where the blending of skills from transformation, planning and operations teams has been key. This ‘Customer Service Triangle’ has created an impressive capacity for continuing change and developed agile working, a mindset that says this is no longer ‘office work’. The same challenges occur in other types of operation, as seen in field and branch operations at Capita PIP (2019) and increasingly with digital channels and robotics.
Process optimisation starts with process improvement, often including third party or outsource providers. At npower (2020), the new back office quality model uncovered £1.5m efficiencies in year 1, raising customer experience scores by a third, by engaging offshore partners in India and adding key questions to the QA form, to flag up processes not fit for purpose. At FSCS (2018), initial transformation led to completing 99% of claims end-to-end within SLA and going digital for 97% of claims within 12 months. With paperless mortgages at RBS (2019), a complex end-to-end process moved from 100% paper to 92% digital in 12 months, enabling same-day mortgage offers. This was a WOW factor for customers, the only one of our Innovation Award finalists so far to feature in nationwide TV adverts! The growth of digital channels is creating new tasks and processes, as automation and BOTs proliferate. They generate exceptions or escalations, but without all the initial questioning or checks, of inbound calls or emails. Increasingly, customer questions from a website or mobile app are specific and in context, with all the customer history. More choice usually means more interaction, despite automation, and we need to plan what can be delivered via AI & Robotics or here the human touch is required. Crucially, spot pinch points, as people do when optimising a factory line or a project. A critical path is created from understanding dependencies.
Resource optimisation is not all about process, however. Understand what drives demand, sometimes complex to forecast, and evaluate operating models to optimise how resource supply matches demand. Model your work types, from case work (eg claims, loans or complex queries) to cyclical tasks (eg billing or renewal) or portfolios (client accounts). Work side-to-side, across organisational siloes, on processes that may require many touchpoints, with calls, visits or evidence gathering. Workers may be skilled professionals, eg medical, underwriting or engineering. They may be in short supply, require qualifications or have a long speed to competency (6 months+). This needs long term planning.
Capacity control: best practice in the back office
- Visibility. A live, accurate view of all work will show: new work (‘in-tray’), work-in-progress (WIP), pending (awaiting action externally) and late (or in danger of this), with metrics for both operation and customers.
- The Waterfall Model shows when queues may go into backlog and start generating their own work (chasers). Understand tipping points and reassign work priorities to prevent backlogs spiralling out of control.
- Workflow processes. Know exactly what steps need completing and how these are undertaken, both tasks (worked by a person/BOT in one session) and end-to-end journeys (many tasks to complete a case). Show elapsed times (end-to-end) and process times.
- Process times. Analyse actual times for every task, from a workflow system, to predict averages, and potential variability, for planning. Some variability is predictable, some is not, you really need to understand this!
- Failure demand. Often there’s a huge amount of failure, but it’s invisible because it’s not measured, and no-one sits down to talk about why it is failing.
- Managing skills. Words to follow, words to follow. Forecasting and managing speed to competency is key. Knowledge or process management systems can really impact process times.
This article was first published in the 2020 Best Practice Guide - 2020 Vision: Crystallising your knowledge
To download a full digital copy of the Best Practice Guide, click here