Don't forget about the weather

One of the best parts of the work I do with Ulster University is the chance to supervise students on the BSc(Hons) Managing the Customer Contact programme when they are completing their dissertations. This is a major part of their studies, and as part-time students is often a huge achievement as they combine work, family and study pressures over the last 3-4 months of their course. The results are often impressive and the topics they research often stick in my mind and help me in the future – and I am only the person reading and discussing their work. As the person doing the research and reading – picking and becoming an expert in a subject they are passionate in must be a great boost to their careers.

A couple of years ago, Tracey Howe, then working for Thames Valley Police, looked at the impact weather spikes had on volume in their contact centre. Amassing several years of data, she was able to prove that when the temperature went above or below certain thresholds there would be a significant change in volume. Using control groups she was able to show that the concept was not unique to the policing sector and, although the thresholds may change, was also valid in many other sectors. I’ve talked about this in many courses over the past few years, but up until now I was basing my talks on her paper and didn’t have the chance to experience and test her theories in person.

For the past few months I have been helping a client in the insurance industry with their Resource Planning processes, including looking at forecasting. I would like to think my short tenure has been successful and amongst other things I have helped them improve their forecast accuracy immensely, often getting the week within 1-2%. As any planner would expect there were days that just didn’t conform to those accuracy levels and it left me looking for reasons. The usual factors came into play – the date within the month (people calling around pay day), the dates direct debits were taken and even things like holidays and school opening. But we recently had a couple of days where we received 20 – 30% more calls than anticipated on one day and 20-30% less than expected the next.

Of course this led to several post-mortems. There was nothing in the historic volumes to give a clue to these widespread deviations – and more worrying was the fact that the week was spot on – but we missed service targets because of the day we were way over forecast. Then it hit me. There was a yellow weather warning on the busy day. The only explanation was that people delayed their Christmas shopping, didn’t go out on the day of the storms, and then changed their plans for the following day. Tracey’s research came back to me. The weather has an impact even in centres where there is no logical or obvious link. Customers change their plans based on what is going on in the world around us, and just like people stop calling when big football matches, Royal weddings or other major events are taking place, they also change their plans based on the weather.

In my first ever job, one of my duties was to fax weather forecasts for the next 5 – 10 days to hospitals and ambulance stations to help them prepare. Back then to be honest forecasts were only reliable at the 2-3 day window – but weather forecasting has improved. 10 day forecasts are now very reliable and often are only inaccurate by a matter of hours. The Met Offices in UK and in Ireland are now criticised for giving notice so far in advance that it creates a media circus – and that has an impact on our volume. For each of us it may have a different impact but it is worth looking back. What do your customers do when the weather changes? Does good or bad weather have a bigger impact? It’s not just criminals who change their behaviours! The 10 day forecasts need to be another tool for planners when optimising schedules – but it can only happen once we look back and see what has happened in the past – then apply that knowledge into. our short term planning.