If you missed the IMPACT Virtual Conference 2021 then this briefing is for you. As a Tech Value Creation follower you might be surprised to learn about the silver lining found in a site failure.
Here’s three insights you’ll get.
- Have you found yourself sitting in a virtual queue thinking why does it have to be like this?
- Or you have been in a mad rush to try and get that once in a lifetime deal and the website crashes.
- What you might not know is that behind-the-scenes ones of these experiences delivers a better business outcome than the other.
Using case studies, we deconstruct website service outages and poor performance with some illuminating results. It turns out, oxymoronically, sometimes a total system failure at a moment of high-stakes online commerce can actually be good for business. There are various scenarios and strategies for replacing randomness with a concrete plan that maps to both technical and business requirements.
- One of the most prevalent online challenges: high-volume sale shoppers suffering long waits and crashes because of retail (or other) system failures.
- Loss of orders because frustrated consumers drop off and go elsewhere.
- Order processing breakdowns that result in blockbuster sales prices being revoked/unfulfilled, leaving disappointed shoppers.
Repeated glitches and crashes are terrible for business, right? Well … not necessarily. Here’s why:
Testing and measurement has shown that there can be an inverse impact to websites when they eliminate chronic spiking and failing. It’s an anomaly that makes sense because of a broader social phenomenon.
FOMO: Fear of Missing Out. Defined nearly two decades ago by a marketing strategist, this is a consumer driver based largely on appetites whetted by social media posts.
These “soft” market appetites can drive unforeseen consumer demand, even (or perhaps especially) when online transactions are impeded by technical difficulties.
CASE STUDY #1: Grand failure during the Grand National
A high-profile betting company experienced huge, uneven surges prior to the Grand National. As large number of fans rushed to place wagers, the system failed. Three years running, the company’s platform spiked and crashed repeatedly. Once diagnosed, the betting company set up a system of online queues to manage the onslaught. They fixed the problem, but their customers went to competitor sites. The urgency was gone, they weren’t waiting around in a virtual queue hoping to place a bet. They took their own chances elsewhere.
CASE STUDY #2: Alternative Black Friday, same results
A high-profile retailer staged its own “Black Friday” event alternative, making it into a big marketing event and drawing in unique visitors through a pre-registration process along with tiered access to the event. The first two years the company’s website crashed early and repeatedly, much to the panic of the entire IT department. The diagnosis was straightforward, an isolated print driver error. Once addressed, the following year went off smoothly, thus the volume of sales shot up, right? Wrong. All the press and drama associated with the website’s failure had actually drawn traffic to the website as buyers worried that they were missing something. They were motivated to buy as much as they could, as quickly as they could, in those pockets when the system was online.
CASE STUDY #3: Turn it off and on again
A well-known travel company announced its huge annual sale with much fanfare, but within hours of opening the window for online purchase the entire system crashed. Getting it rebooted and back online was laborious and took huge chunks of time. The issues were eventually identified and remediated, but the airline did not see an increased volume of sales. It turns out buyers responded with more frenzy to a disastrous, fitful online sale where every moment counts than to a smoothly optimised ticket purchasing process.
Ultimately, there are many FOMO variants, but the human instinct for the hunt and desire to be where the action is can significantly influence website traffic.
The trick is to control the randomness of a system IT breakdown and instead modulate it to best commercial advantage.
The key elements required to do this are:
- Performance – Do your customers’ expectations mirror the amount of time required for you to process on your platform?
- Efficiency – How do you choose to balance technical time and structure versus cost and other customer priorities?
- Stability – Do you want things to slow down? Are you having outages, why and at what point?
The strategy to identify and respond to bottlenecks is two-fold but based on data. You need access to data to inform your strategies and decisions.
Technical: The simpler the better.
- Start outside your APM or monitoring tool.
- Can you get access to CPU, Disk (I/O), Disk read/write, Memory?
- If you want defined granularity, can you get that?
- Once you have that data, do you have a process for interpreting it?
Business-side Operations: Factoring in Technical Realities
- Orders per second.
- Web logs that track customers on platform.
- Overall concurrent users.
A Strategy Going Forward: Once technical data and business metrics have been extracted and a process documented, they can be mapped across the enterprise. Data bears out: Identifying and harnessing the power of consumer FOMO can lead to significant shifts in online patterns, intensity and volume of business.
Key Quotes from the presentation:
• “Are big sites and platforms that can handle really high volumes and make it completely seamless for you customers a good thing? What we found is that in fact it had the inverse impact. People actually spent less money on websites that weren’t spikey, falling over or failing.”
• “The customer being happy doesn’t necessarily correlate to them spending more money with you.”
• “The act of something being sold out becomes a journey itself to go and try to find that thing somewhere else.”
• “Now that we’re all more connected and we can see everything that’s going on in people’s lives, (FOMO) has become much more a thing at the forefront of our everyday lives.”
• “You would think that having a super-stable, super-functional platform would be great but those two years that this retailer was on the news, having a bad day, caused more people to go on their website, more people to try to book stuff. They wanted to know what they were missing out on. Was it some big discount? Was it something they could never get again?”
• “They still talk about how we can do that again … find a process of almost engineering this feeling that what (the website) has got is so valuable that people start to fear missing out getting access to it.”
• “Efficiency. Performance. Stability. Once you have a handle on these three things you can start to choose which levers you need to pull in order to create that scenario where people almost feel like they are missing out. You stop it being a random situation.”
• “It’s not just a technical problem. It’s not just working out performance. There’s also an element of, ‘How do my customers expect this to work. How can I map this to my own internal and my customers’ expectations?’ ”
• “What you want to do is create this scenario where you can really make whatever change you want and you don’t have to worry about it. You understand the ins and the outs of the platform and it just becomes simple.”
• “If you can get the best of both worlds and capitalize on being able to control your platform in the right way but still create that fear of missing out, you’re onto a winner.”