The dynamics of Software Development Centres closely resemble complex queuing systems such as traffic or airports. Using our everyday experience to prompt questions on the economics of why these systems are set up the way they are can give us insight into how best to organise and fund standing delivery organisations.
Read the scenario and then deeply explore the questions that follow.
Picture yourself arriving at an airport terminal.
Your goal is primarily to catch your flight. Miss the gate’s closing point, and life has just suddenly taken a downturn. You’re facing rebooking fees, and maybe net higher ticket costs. And you may miss the entire value of your travel. So you want to get through the airport – check in, bag drop, security (often in multiple layers), gate – quickly.
But even without this dramatic drop in value of travel, standing in queues is still a pain. It’s always better, more comfortable and less stressful to spend your airport time sitting in the lounge than shuffling your stuff forward a few centimeters at a time.
So getting through the successive processes of air travel quickly is important, and has quantifiable economic value to passengers. It also has economic value to airports and airlines to deliver this to their passengers. As a frequent flier between Scotland and the South of England, I will fly almost anywhere in preference to LHR and LGW, providing additional revenue to SOU, LTN, LCY and FlyBe/EasyJet over British Airways. Simply because those airports are easy and quick to pass through – more than once I’ve arrived in the SOU terminal as my flight is being Last Called and still made it.
Airports and airlines recognise this by offering short cuts to high priority passengers: frequent fliers, business class passengers, and those prepared to pay ‘pre screening’ fees. They fund dedicated and exclusive capacity that enables the selected people to pass through much shorter queues. Sometimes you can self-select into the valued customer category by choosing to pay a premium at the airport once you see the length of queue and make a quick value check assessment.
So, picture yourself arriving at check in. It’s moderately busy but there are plenty of desks open for your airline, so the queues are short, and you know you’re going to get checked in fairly quickly. You’re a Business Class passenger, so you walk right up to a desk and are served right away.
As you head for security, you look over at a super budget airline next to you. To keep costs and fares low, they only have two desks open, and the check in agents are run off their feet trying to get everyone checked in in time for their flights. Advertised check in closes 40 minutes before departure time, but the queue is 20 minutes long, so they have a supervisor (who could be on a desk) walking the queue expediting people on the next flight to close. If someone’s flight isn’t leaving soon, they don’t seem to be moving at all.
Use your experience of the scenario to think through the questions below, and then find parallels in the delivery centre environment.
- Which of these models serves passengers best? Which serves the airline best? What factors impact the economic case?
- What happens to each airline’s queue if there’s a passenger who takes an extraordinary length of time to check in?
- Many airlines bunch flights together to minimise waiting for connecting passengers. What will happen to check in (and security) demand and queues as the number of imminent departures varies through the day? And what happens on the first day of the holidays?
- Airlines and airports mitigate the worst of these changes by increasing their capacity – the number of desks. But they can’t simply recruit people for these on demand. Even given a base level of skills (eg checkin agent), the specific training time for each airline’s systems and culture far exceeds the time available in ramp up. So how do airlines maintain a rapidly deployable capacity that can respond to demand variation?
- What would happen to the queue if the first check in question were the passenger demanding an exact estimate of the time that all airport processes to boarding are going to take?
- Given budget airlines’ strategy of separating out every cost and reflecting it as a marginal fee, why doesn’t Ryanair charge sliding scale fees for every airport facility, estimated in detail at booking or check in?
You ponder these questions during your flight. As you depart the airport at the other end, and stand in the taxi queue, you ask yourself some more questions:
- Is it better for the passengers to wait for the taxis, or the taxis to wait for the passengers? What would you need to enable the latter?
- How much effort and cost would be needed to plan the exact number of taxis needed by an airport at any exact time? How far in the future could you plan this? What would the consequences of error be?
- If you spent the planning money on taxi driver capacity instead, how much organisation would you need? What would the impact on queues be?
- How much of a surcharge would you pay for the extra capacity to provide responsiveness?
- Would you need to pay this all the time, or just as a premium service when you’re in a hurry and the queue is long?
- If you’re getting good service at a price you’re willing to pay, how upset are you if a driver fills in the crossword every day?
Latest posts by Martin Burns (see all)
- Colocated Teams, Distributed Organisations - January 20, 2018
- Christmas Reading: 6 Books That Will Change Your Thinking - December 22, 2017
- Now Is Not The Time for #NoProjects - December 14, 2017