In retail -- and especially in e-commerce -- there's a nuanced distinction between having a very popular sale and arranging for far too little merchandise. It's like those hold recordings that say the lengthy hold time is because of high customer call volume, prompting most people to mumble, "That and the fact that you're too cheap to hire enough call center operators."
That's E-Commerce Lesson #1. E-Commerce Lesson #2 is that the distinction between a site being down -- having crashed -- and a site that is responding very slowly is usually lost on shoppers. IT may understand the difference, but the consumer who can't make the desired purchase won't. Target this week relearned both lessons.
On Sunday, it launched a much-hyped promotion for Lilly Pulitzer apparel. The sale was intended to last weeks. In reality, it lasted hours. And during those hours, visitors to the site experienced some impressively slow site response.
Kathee Tesija, Target's chief merchandising and supply chain officer, posted a Target-written Q&A explaining her reactions to the incident. (Must be nice to be able to control both the Q's and the A's. Makes the A's so much easier.)
"We anticipated having enough product to offer the collection for several weeks, but at the end of the day, our guests let us know that they didn't need that much time to decide that this was a collection they wanted to bring home," Tesija penned.
Suspicions among Target shoppers have been running high. How could Target -- which has extensive experience in such matters -- have misjudged demand so wildly? The top suspicion: That Target planned this as an elaborate bait-and-switch, knowing that most shoppers would be hit with out-of-stock notices and could be persuaded to buy something, potentially at a higher margin.
The reality is that the Web -- and comments made to Tesija's post -- spoke of a large number of copies of the Lilly Pulitzer products finding themselves on eBay and other resale sites, invariably at a huge markup. Tesija disputes this, saying that a tiny portion of products were being resold.
"When we look at the amount of Lilly Pulitzer for Target product being resold, it translates to roughly 1.5 percent of the collection," she said. "While we'd prefer that number to be zero, it tells us that the vast majority of guests who purchased the collection did so with the intent of enjoying it for themselves."
Part of the problem with that very precise 1.5% figure is that Target didn't say how it came up with it. Was it done through a series of Google searches? If so, when was it performed? How much of the product was already resold before they checked? And not every site -- especially not less reputable ones -- are so easily found in a Web search.
As someone posted on the Q&A site, using the name Michele Oresky Kusiak: "If you take a look on eBay, there are over 72,000 Lilly Pulitzer for Target items listed. That doesn't say that most people bought it to enjoy it themselves," Kusiak wrote.
Added another commenter, using the name Stephanie Carter Perry: "There are other secondary markets, such as PoshMark where 1,000's of these items are listed. The 1.5 percent is not a true or accurate number because it doesn't include other secondary purchasing sites OR what transactions have already taken place on eBay. Bottom line, Target doesn't care. They got theirs."
Target certainly made an effective use of social media to push this sale. "This is now officially the most talked-about collaboration on social media that we've ever had," Tesija wrote. "When we launched on Sunday, we saw heavy traffic to our website and lines outside of many of our stores across the country. Clearly, Target's guests had this launch marked on their calendars, and the product quite literally flew off of our shelves. By mid-morning, it was essentially sold out."
"The product quite literally flew off of our shelves"? No, it literally didn't, unless the garments sprouted wings.
Target underestimating -- and that's the most charitable interpretation -- sales is one thing. What about underestimating Web traffic?
There's no dispute that Target's site, during this sale, slowed to a crawl. How did Tesija reply?
"The website never crashed, but due to high traffic, it was slow and our guests had to wait longer than they should have to access the full collection. That's frustrating, both for them and for us," Tesija said. "We're taking a close look at what happened. We're committed to constant improvement, and are laser focused on providing our guests with a great, seamless experience."
When IT looks at a site, it doesn't declare it having crashed until there is zero activity. When I have talked with retailers about site troubles, the answer is invariably something like "Of course our site isn't down. I am seeing thousands of sales being processed." That's a reasonable conclusion, but it doesn't factor in how many shoppers could not get through. Would they have otherwise seen 100 times that number?
To a shopper who can't get through, the information that many others did get through provides little comfort.
Is Target getting the message? Well, this may be a coincidence (and given how long it takes IT at a place as large as Target to react, it quite likely is), but on Tuesday (April 21), Target's mobile site issued a patch. The iTunes description said that it was a "fix to increase stability during high-traffic events. A special thank you for your patience as we addressed this bug!"
(By the way, the Target mobile app also added support on Tuesday for iPhone fingerprint biometrics. "You can now sign in to your account using Touch ID with iPhone 5S or newer" as well as it "resolved an issue where some (shoppers) were repeatedly prompted to sign in." Thought you'd want to know.)
My favorite part of Tesija's message, though, was saved for the very end and it was a message for shoppers who were locked out of the sale. "We won't be restocking this collection: once it's gone, it's gone. However, if you missed out on the launch, you can still check back at your local Target store over the next few weeks. We have a 14 day return policy for programs like this, so there's always the chance that someone might return an item that you're looking to get your hands on!"
Yes, nothing says Target like waiting to pay full price for product rejected by someone else.
Having said all this, please allow to say a few words in Target's defense. Let's see. Target secured some product that its shoppers really wanted (good work, purchasing), it did a terrific job getting the word with aggressive marketing and social media (lots of points for marketing), the sale launched when it was supposed to (no brownie points, but there's nothing bad here) and the popular product sold out.
While it's fine form to Monday Morning Quarterback inventory decisions, it's not known how much product was even available to Target. More importantly, projecting sales is never easy. Granted, thinking you have enough for weeks and selling out in hours is compelling circumstantial evidence that somebody misjudged inventory levels or demand, but maybe marketing did its job too well. It's hard to fault Target for that.
The site slowdown? This is a bit trickier. Target.com should be able to handle tidal waves of customers when they knew it was coming. But no server farm's capacity is infinite. If marketing indeed did a better than expected job, it's reasonable that IT would have caught offguard. Although I said that there is no perceived distinction among shoppers between a slow site and a downed site, there is in fact a difference and the site--as far I can tell--indeed never crashed. When hit with an unexpectedly huge rush of traffic, a site slowdown is hard to avoid. Can't find much fault in Target's IT team, either. (For this, guys. Just for this.)
There are lessons to be learned and someone needs to come up with more accurate ways of forecasting demands for highly marketed campaigns. Until then, though, Target's loss is eBay's gain.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.