It has been pretty well documented in the online community just how many retailers, many of them major brands, managed to suffer a rather spectacular collapse on Black Friday. Since the Thanksgiving spending spree first migrated from the US last year, it has seemingly become an unstoppable force, assaulting the UK market with a truly ferocious wave of consumer expectations. But did the online shopping sector hold up to the barrage of traffic assaulting their websites on the sales day?
The general consensus seems to be mixed – plenty of high profile crashes and queuing situations arose – while some retailers managed to really nail it. The key, as with so many short timeframe events, seems to have been a combination of planning and flexibility; ensuring that site infrastructure could handle the sheer volumes of traffic and user interactions that would be occurring, while also giving active marketing teams the freedom to make promotional adjustments on the fly.
It is hardly surprising that Amazon’s lightning deals system worked so well. The UK site received over 80% more website traffic year on year for the week, as the behemoth of online retailers has had plenty of practice preparing for Black Friday deals in its US incarnation. Rolling out preparations more globally was likely more a case of logical expansion rather than revolutionary thinking.
Some other major brands held up with only minor or no snags for the whole twenty-four hour shebang – notables included Marks and Spencer, Debenhams and tech giant, Apple. A few, like John Lewis, toppled in the rush but battled back to fighting fitness to scrounge what they could from the rest of the day.
Econsultancy even featured a delightful example of agile marketing at AO.com with some cheeky PPC brand bids diving them into the competitive mix, offering ad messages like “The Lights Are Still On” or “No Queues, Just Amazing Deals”, allowing them to steal traffic from less fortunate brands. Agile indeed, and a brilliantly inventive use of what was no doubt a cunningly allocated flexible PPC budget!
The biggest losers are of course those whose sites went down altogether, of which Argos, Game and River Island were just a few. Other users reported general problems with sites crashing midway through browsing or (attempted) purchase journeys, naming some pretty prestigious names that really ought to have known better – Nike, Topshop and Net-A-Porter were just some of the offenders.
More interestingly, some brands chose to put in defensive measures rather than looking at scaling infrastructure, implementing auto-refresh or “queuing” options for their sites to only allow so many active users at a time. PC World and Tesco both opted for this in an effort to retain visitors – and one could perhaps argue that there’s a form of scarcity marketing there. With the lightly burning server farms aside, with hordes of deal-hungry consumers anxiously hanging around on their browsers in the hope of snagging just the right offer when the chance arises.
Plenty of brands threw tons of paid search (and other advertising) budget at Black Friday too, which if anything makes the outages worse. Adwords and other paid search solutions will continue quite happily charging for clicks even if a site is offline, and with bid prices skyrocketing during those critical shopping hours, it hurts to think how much media spend has been wasted on clicks pushing users to queued or otherwise inaccessible sites.
UX, SEO & Google’s Viewpoint
Does that really offer a good experience for the instant gratification user culture that pervades digital, though? Is a modern web user, accustomed to modern web speeds, next or same day delivery and instant availability, going to be willing to sit through such pedantry, especially when a mere click can send them to a competitor for more rapid gratification?
And – here’s an interesting thought – will the unavailability of some of these major brands’ websites, some of which were still struggling to get back on their feet after 24 hours, impact their search visibility? Just how much downtime is Google willing to accept before it starts slashing visibility in favour of the more stable? If user experience is the name of the game, then it seems a logical assumption that a site which goes offline or “unavailable” for prolonged periods will see their organic traffic suffer as search engines decide they aren’t the best solution for searchers.
What To Do
If a site is down for a planned (or otherwise “anticipated”) outage, avoid a plain ol’ 404 which will be poison for your homepage for anything over a very small handful of hours. Similarly, don’t put up a placeholder and serve a 200 response – while there may be something sensible displaying, everything is most definitely not OK! If you’re using queue systems, stop and think before you throw in redirects willy-nilly – a day or so of minor borks probably won’t impact most websites, but all it takes is for a search robot to hit your site for a large crawl at precisely the wrong time and suddenly you’re sending all the wrong visibility signals.
Fortunately, there’s a solution to all this – the 503 response. This HTTP response code sends a clear and definite message to all comers – “Service Temporarily Unavailable” – which is exactly what is in fact happening. It therefore allows search engines to see the same sort of message as your users can, which of course is the best idea whenever dealing with thorny search marketing issues.
The best part is that you can even include an instruction to crawlers to come back later when you anticipate normal services will have been resumed – this could be a date and time or a number of seconds, so you could for example ask the crawler to try again the following day at 8am or some other suitable hour (probably a worthwhile idea in the case of Black Friday)…
Retry-After: Sat, 29 Nov 2014 08:00:00 GMT
Or, if you’ve got a technical team beavering away with fire extinguishers, you could ask the crawler to come back in two hours’ time (which needs to be specified in decimal seconds)
Don’t of course forget to put a nice page up for users too (regardless of whether you’re offering queuing or just an “oopsie daisies please come back later” message), and remember that a 503 is not a long term solution, and Google in fact specifically advises that “lasting 503s can eventually be seen as a sign that the server is now permanently unavailable and can result in us removing URLs from Google’s index.”
Not all PR is good, contrary to the old saying. So the next time your marketing goes into overdrive, make sure you don’t get so focused on those juicy forecast figures that you forget the long-term health of your user base and how they perceive you – and the same for the search engines!
4Ps isn’t just another London SEO agency. To discuss how SEO, PR and paid advertising are evolving together in order to keep pace with new developments in user interaction and brand demands, give us a call on +44 (0)207 607 5650 for a no-obligation coffee and chat about data, marketing and user behaviour across all inbound channels.