A good Black Friday period can make a retailers year but it can also cause significant brand damage if things go wrong says Andrew Vermes, Senior Consultant at Kepner-Tregoe
Black Friday has become an increasingly important indicator for a retailer’s sales performance with 30% of consumer spending taking place between Black Friday and Christmas alone.
On the day itself, a colossal £1.4 billion is estimated to be spent by bargain-hungry shoppers and unlike many other seasonal sales statistics, which tend to fluctuate, this popular trading day just keeps on growing.
Black Friday is also the first true test for retailers this year in terms of managing heavy traffic and meeting customer service expectations.
Reputation is at stake for high profile brands, so there is immense pressure for everything to work and for websites to run smoothly and seamlessly without interruption. But with immense pressure comes risk and when something goes wrong people get flustered and sometimes they make mistakes.
Pressures on retailers over the course of Black Friday are huge and systems and processes simply have to work. Consumers can often purchase the same goods from different retailers at the same discounted prices so if a system fails or stalls and prevents the buyer from getting what they want at the moment they want it, in an instant they’ll hit another website.
Customer expectations are the same as any other shopping day; they don’t take into account or make allowances for the fact that the systems are being struck by ten times the normal volume of traffic.
They expect to search for what they want and complete their transaction without delay. There is no room for error at the point of sale particularly on a day when consumers are up against other bargain hunters looking at the same items.
There are a number of common big risks for retailers on Black Friday. Connectivity is always a major concern when a website is overloaded. The processing of customer purchases rests on an interconnected system where a single small error can result in a failed transaction.
The real issues however, usually start when a critical problem occurs and a response is made. Every large retailer has a major incident team who are ready to respond and react when things don’t work. Sometimes issues are spotted and corrected before the consumer is affected but in many cases the reverse is true. But very occasionally something big happens, for instance a road contractor cuts through a fibre optic cable at the exact moment that the backup channel has had to reboot following a software malfunction.
Then you can get millions of transactions queuing and some even get lost. An effective response depends on everyone involved (internal IT staff, contractor, vendors and third party suppliers) all following a consistent method of troubleshooting. Without that, the outage can last hours or even days, as we’ve seen at some UK banks recently.
Retailers have a lot to lose in this kind of scenario and on a day like Black Friday it’s a problem they cannot afford to have. Teams must work closely together to solve problems, consistent communication is vital.
Despite this, there are a number of key mistakes that retailers make when dealing with problems. One ineffective process involves the IT teams checking that their part of the infrastructure is working.
Often they all run the usual diagnostics and report “my section is fine”. More effective organisations would have assembled all of the available data quickly as a team and used that to make good judgements about where exactly to look for the fault.
A common failing is also jumping to false assumptions, but this is tricky to overcome when the time sensitivity is so great. This kind of urgency can lead to panic and that is never a good place for clear thinking. What makes it worse for IT support is when senior leaders are constantly questioning, “when will it be fixed?” This simply exacerbates an already stressful situation.
Jumping into false causes can also make problems worse simply because it was tackled in the wrong way at the outset. To give an example, one large electrical retailer started to experience issues with point of sale systems at the start of the Saturday peak traffic. In short, they weren’t able to take any payments at all.
They quickly traced the issue to a lack of connectivity with one of two data centres. They decided to reboot the network switches concerned, which took 30 minutes alone. During that time nobody focused on the reason why the traffic didn’t simply fail over to the other data centre. The reboot restored traffic, but they still didn’t know why it had stopped in the first place.
Since things were running ok, some key technical staff went home. Within 15 minutes, it was failing again (sounds familiar?) Eventually the cause was traced and corrected, but only after three hours of lost trading at half of the stores. This was a clear illustration of the importance of complimenting workarounds with good root cause investigation.
Most retailers should have already started to prepare for potential risks associated with Black Friday such as performing resilience tests, if not they are at risk since no one can predict if and when a problem will occur and with the best will in the world you can’t prepare for every eventuality. So, if a critical problem does arise, there are a few steps you should consider first.
• Make sure you understand the impact of the problem within the first five minutes of contact.
• Look for as many workarounds as possible to restore function as you can and choose the best one. Avoid going for the first plausible suggestion, as this could well be a mistake that makes things worse.
• If locating the cause is needed to fix the problem, look for any contrast between the elements of the system that are down and those functioning normally.
• Don’t allow people to do things without asking yourself first “what could go wrong if we do this?”
• Ensure your critical incident team is staffed with mature people who can take the heat and manage panic under pressure.
• Let the leadership team know that they must comply with any or all of the requests from the critical incident team without question.
• Focus on the data itself; understanding what’s happening is crucial when finding the solution to a problem.
• “Who?” questions are not helpful to anyone so avoid at all costs.
• Black Friday may be very close by but it’s not too late to get a virtual, visual tool to share key information with the team. Use Skype, Onenote, Webex or any other similar tool to make sure everyone involved is looking at the same picture and in real time.