;

Three e-Commerce Lessons From the Super Bowl

Superbowl BlackoutThe 2013 Super Bowl was more digital than ever. The 71,024 fans in attendance all had access to free wi-fi. Dozens of Super Bowl advertisers launched aggressive digital campaigns in the weeks prior to the game (including, in many cases, pre-releasing their commercials). And the game itself featured a record number of commercial spots with digital calls to action, and extended on-line experiences.

The Success of these various digital initiatives was mixed, but three clear lessons emerged that can be immediately applied to e-commerce sites.

Lesson 1: Prioritize your calls to action

Forty-two advertisers paid $3.5 to $4M each for 30 second advertisements during Super Bowl XLVII. What did those advertisers ask viewers to do?

In 2012, 68 percent of advertisers had a digital call to action. A year later in 2013, 83 percent featured a digital call to action.

2012

2013

Included a URL

57%

40%

Included a Hashtag

7%

48%

Included Facebook

11%

10%

(2013 data from marketingland.com; 2012 data from Altimeter Group)

Surprisingly, fewer ads in 2013 sent viewers to digital property owned by the brand (corporate websites), sending them instead to third party social sites. The big winner was Twitter, which saw hashtag use rise dramatically. Advertisers clearly banked on the 108 million Super Bowl viewers’ impressions being amplified by Twitter’s 200 million active users.

The conventional school of thought is that Twitter is now the best social network for realtime events. But some unconventional calls to action also appeared in 2013. Oreo sent fans to Instagram, helping Oreo grow its Instagram followers from 2,200 users to over 30,000 (at a cost of about $200 per follower, for those keeping score at home). Jack in the Box invited viewers to use Shazam to see an extended version of their video. Apparently they expected 108 million Super Bowl viewers would have Shazam pre-installed on their smartphones, and would race to activate the app before the end of the 30 second commercial.

Were these particular calls to action the best choices for these brands? It’s hard to imagine Jack in the Box wouldn’t have done better by sending viewers to a landing page on their main website to watch an extended video. And couldn’t Oreo have offered a better experience by running a voting contest on a site they owned?

In much the same way, e-commerce sites today face a dizzying array of potential calls to action. The best practice is to make an inventory of all the potential calls to action (CTAs) on your e-commerce pages, such as:

  • Add to Cart
  • Add to wish-list
  • Checkout
  • Create Account
  • Subscribe to E-Mail
  • Click to Chat
  • Find a Store
  • Call
  • Like/Follow us on a social network
  • Share our content with your social network
  • Write Reviews
  • Compare Products

Estimate the value of each CTA, taking into account the monetary value of each one and its role in helping your users complete their primary tasks. Next, determine which CTAs are being emphasized in your customer experience. Ideally, you’d like to curate your users’ attention by having a clear visual hierarchy of calls to action, with the most valuable CTAs getting the bulk of attention.

Lesson 2: Plan for things to go wrong

Coca-Cola had ambitious Super Bowl plans, running an interactive contest tied to the TV ad for their “Mirage” campaign. Unfortunately, as soon as the ad aired, traffic to their site spiked from 33,000 visitors to 1.3 million visitors. The Coke site slowed to a crawl and eventually crashed. Although Coke had dedicated 33 servers to the campaign and load tested the contest site for much higher volumes, the real-world traffic of 78K requests per second took them down. 2013 is the second year in row that the site crashed during the game.

Coke didn’t fare any better on Twitter, where they were also accepting “Mirage” votes. Because they thanked every voter with a reply from the official @CocaCola twitter account, they quickly hit Twitter’s 1000 tweet per day limit. Coke requested and received an increase in their maximum tweets, but they still maxed it out, leaving @CocaCola unable to tweet for almost 2 hours during the game.

Yet, because CocaCola had carefully rehearsed their big day a number of times, they had contingencies in place. When the website went down, they were able to use twitter voting; when their official account got locked, they were able to switch to hashtag voting. In the end they generated over 8.2 million interactions, including more than 900K votes (well above their 400K goal).

Coke had plenty of company. Thirteen of the 17 advertisers who featured web urls in their ads had outages during the game (based on data from yottaa). That’s a whopping 76 percent of advertisers who were unable to keep their site up for the game, after spending $3.5M to promote their site in an ad!

For e-commerce sites, our Super Bowl is usually Cyber Monday (the Monday after Thanksgiving). Many sites will do 40 percent of their annual revenue in the six weeks starting with Cyber Monday. Best in class e-commerce sites carefully rehearse the big shopping days, including implementing a full holiday readiness plan. July is the time to start executing that plan to ensure you don’t have an outage in November.

Lesson 3: A single point of failure can bring you down

Single Point of Failure

The Super Bowl is one of the most meticulously planned events in the world. Frank Supovitz, the NFL’s senior VP of events, leads a team of thousands, choreographing one of the most watched events of the year. The event represents billions of dollars of economic value (the rights to broadcast the game alone cost more than $700 million). So every element is carefully rehearsed, with contingencies in place for virtually any eventuality.

Yet despite all this attention from the best event-planning minds in the world, the Super Bowl went black for more than 30 minutes. Shortly after halftime, power to half of the New Orleans Superdome was lost, causing a 34 minute delay in the game. Entergy, (which supplies electricity to the Superdome) immediately disavowed any responsibility, announcing via Twitter that the problem was on “the customer’s side,” essentially throwing its client (the NFL) under the bus. A later joint statement from Entergy and Superdome managers said they were jointly looking into the problem.

Yet, when the power first went out it’s unlikely Frank Supovitz or anyone in the control room, knew who Entergy was, much less had developed an emergency response plan with them. Even though they were a critical third party that clearly had the ability to disrupt the Super Bowl.

Similarly, every e-commerce site is dependent on numerous (often dozens) of third parties to deliver its customer experience. A sub-set of those vendors are “Single Points of Failure” (SPOF), meaning if their service is interrupted it will be impossible for users to complete their tasks. Third Party javascripts (libraries, analytics, social, reviews, recommendations, etc…) that block the page from loading until they are successfully executed, are particularly common and dangerous. It’s critical that e-commerce leaders know what those SPOFs are, and have a plan in place to respond when there is a problem.

SPOF-O-Matic is a great browser extension for Chrome that will help identify technical SPOFs on e-commerce pages.

WebPageTest includes an option to simulate all (or select) third parties failing, so you can see the effect on pageload speed, and even create a video of the page loading with the third parties failing.

Additional Reading:

  • Steve Sounders (one of the founding fathers of web performance) explains Front End SPOF.
  • Patrick Meenan (another member of my personal web performance hall of fame) writes about how to test for Front End SPOF.

What have we learned?

Even the best laid plans of the world’s largest digital marketers can go wrong under the stress of huge traffic spikes. E-commerce leaders are faced with similar conditions each year around key shopping dates like Cyber Monday and Black Friday. It’s imperative that we not only minimize the risks wherever possible, but that we understand which risks can’t be eliminated and rehearse what we’ll do when the worst does happen.

How are your 2013 holiday readiness plans going?

You can still dunk in the dark

 

 

Speak Your Mind

*