Was Finish Line’s New Site Disaster The Latest Cloud Casualty?
Written by Frank Hayes and Evan SchumanFinish Line is giving itself 90 days to decide if it will resurrect a cloud site that almost wrecked the chain’s holiday season with extensive functionality problems, all seemingly due to a major site refresh handled by Demandware.
“We believe the new site, which came online Nov. 19, 2012, cost us approximately $3 million in lost sales for the third quarter. Following the launch, it became apparent that the customer experience was negatively impacted, evidenced by a decline in several key performance factors,” said Glenn Lyon, CEO of the $1.4 billion sports apparel chain with more than 640 stores in 47 U.S. states, during an earnings call on Friday (Jan. 4). “We made a strategic decision on December 6 to transition back to our previous site, given the importance of the selling season. As part of our contingency plan, we had kept our previous platform up and running, parallel to the new site, so we were able to swiftly engineer a smooth return to the original site. This has generated improved results versus what we experienced during the three-week period the new site was live.”
A few weeks earlier, Finish Line was singing a very different tune when it said the chain’s “aggressive omni-channel growth initiatives drove its decision to move away from legacy in-house-managed software to Demandware’s flexible and scalable cloud-based solution.”
Chris Ladd, the chain’s EVP and chief digital officer, gave a news release quote that he may now regret: “Demandware reduces operational complexity and puts the power in the hands of our users to create world-class brand experiences. It provides customers a better and faster online experience, rather than having technology get in the way.”
But in the earnings call, CEO Lyon spoke very differently. “Well, we’re back to the drawing board in terms of the legacy system that we ran and the upgrade of that system, which was an option, the utilization of Demandware and their ability to perform. Obviously, we were disappointed by the performance back at the end of November,” he said. “We have some options here, and we’re going to get those out and more diligence over the next 90 days. There is a lot of learning to be done, including outside advisers and so on. There’s our confidence level, and that has been shaken a little bit. And as I said to you, reality reared its ugly head to us for the first time in a number of years in terms of initiatives that we’ve put out into our business. This was a tough one.”
Finish Line executives didn’t go into detail with what they called “customer experience and functionality problems.” But that language suggests the site was simply not responsive enough, a problem some other major site revamps have hit in the past. And the most likely culprit is the cloud architecture.
-Marc
January 11th, 2013 at 5:33 am
With what happened last holiday and with so many people greatly affected, I guess they will have a hard time building the credibility and gaining people’s trust again.
January 11th, 2013 at 10:37 am
Is this really Demandware’s fault or the Finish Line CIO/CEO’s fault? Why would anybody release new technology prior to their biggest activity season? The Finish Line made a risky bet and they lost. Retailers should stop the risky “just in time for the holiday season” mantra when implementing technology solutions. I’ve seen this over and over in Retail IT sector and they seem to end up with more eggs on their face than success stories lately.
January 15th, 2013 at 12:04 pm
Avid Finish Line (Online Shopper) & let me say first hand experience that site was absolutely horrendous…every link failed and timed out it was hellacious…The old site is just fine no need to change it.
January 15th, 2013 at 3:25 pm
Wow, the Demandware story is interesting. You rarely get to compare things apples-to-apples like this.
Demandware is OK, but their base functionality in terms of information retrieval quality is mediocre in my opinion. Their spell check is 2-step. I’m not impressed with their overall keyword-search quality, either. SEO — I won’t even get started. That and I’m not sure why they’d do this when Demandware typically takes a percentage of your sales.
It’s hard to know what they mean by “performance.” I think they kept it vague on purpose.
I’m frankly not surprised they switched back. I’m not sure which eCommerce platform it is (could be custom and I haven’t investigated deeply), but they look to be using Endeca for their search and navigation. My analysis is simple. This was a foolish move, and the fool that decided to switch from what they had to Demandware should be fired (or hire me to evaluate his next questionable decision). Even if I’m wrong about all of the above, they then had no control over their application, as their entire application is SaaS!
Strike 2. No control over your destiny. SaaS has a habit of causing that problem. For a large B&M like FinishLine, this is an unacceptable decision. Demandware has crocs.com in its portfolio, I see — which is no doubt a bullet point that Demandware used to seal the deal, but Crocs isn’t FinishLine. Crocs is a manufacturer that dabbles in selling online. Finishline is a retailer with serious merchandising needs.
Walmart switched away from Endeca, for example, but it wasn’t for any other reason (in my opinion) than to have even more control over their destiny. So they also obviously didn’t move to a SaaS platform.
They moved to an even more custom platform where they can write their own algorithms based on technology they acquired from a company founded by the same engineers that built Amazon’s A9 search platform. In other words, Walmart has the capital to invest in a completely custom merchandising/search/nav platform and probably didn’t like that their other customers had the same access to the same technology as they do if they made the same purchases. I don’t think they were _unhappy_, they just wanted more control.
Finishline isn’t Walmart. They were doing fine with their previous decisions and in my opinion this was a step backwards in terms of control and quality. Kudos to them for at least admitting they made the mistake and taking the fix to the finish line . . .
January 16th, 2013 at 6:02 pm
“And the most likely culprit is the cloud architecture, because Finish Line in November bragged about how the new site stitched together CRM, inventory and other databases with customer shopping carts. That means a lot of round trips between Demandware’s cloud, Finish Line’s datacenter and customers’ browsers. It may be possible to tune a site that ambitious to overcome all that inevitable network latency, but it sounds like Finish Line’s site didn’t manage to achieve it.”
How can you possibly say “the most likely culprit”? The FL President himself said:
“It was really about conversion that led us to make the strategic decision to move back to our legacy site and to ensure that we could preserve our important holiday selling season.”
Conversion rate is driven by so many factors that it’s impossible for an outsider to comment on what the problems were. Finish Line did a complete site redesign at the same time they launched a new eCommerce platform. If their design firm did a poor job with the usability of the site, that could cause conversion rate to plummet, regardless of what platform you’re moving to.
Calling this the problem of the SaaS infrastructure without any details of the problems doesn’t make sense.
January 16th, 2013 at 7:13 pm
When you have a conventional E-Commerce site that’s working fine, then replace it with a cloud-based site and conversion rates drop, you don’t just blame site design. The site’s new design didn’t drive traffic away — traffic went up slightly. It was just conversions that dropped.
And in fact Samuel Sato, Finish Line’s brand president, said (in a quote we didn’t use, responding to a Credit Suisse analyst’s question about whether the conversion problem was from the front-end creative or the systems part): “So yes, it was both. We believe that conversion, customer experience and conversion being the metric is largely impacted by both site design as well as functionality of the back end.”
The fact that Finish Line didn’t make the decision to simply fix the creative, but instead is taking a total of four months to assess whether the new site can be workable, suggests this isn’t just the front end, but a more intractable technical problem. As least, that’s what it suggested to us. Your analysis may vary.
January 18th, 2013 at 3:33 pm
No, this isn’t a “Cloud Casualty.” This is clearly just another example of a rushed launch. If you are re-platforming your entire e-commerce business, you should launch at least two months before the holiday season. Some pre-launch performance testing would help. It’s that simple. Anyone with any decent experience at all with e-commerce platforms who doesn’t have something to sell, knows I’m right. We’ve all been there.
March 6th, 2013 at 9:45 pm
After working in ecommerce for over 15 years I would bet there are several factors for this failure.
But this is usually the most common –
Performance usually takes a back seat to features and customizations. No one will step up and say to a retailer… you probably don’t want to implement this feature because the very nature of it will kill your performance and possibly the site itself. Its a game of Quantity and not Quality. Do not implement anything until you have confidence it can perform well.
Also, there are no perfect ‘platforms’. I’ll bet the Finish Line also has issues with their legacy system too, its just that over years of use… they are used to its shortcomings.
April 1st, 2013 at 3:18 am
It is always better not to make major changes before and during the holiday season. Launching a new technology or feature during this period is very risky and should be avoided at all costs.
April 2nd, 2013 at 12:36 pm
The fault seems to lie with both of them.
Finish Line for requesting to rush a platform and website redesign launched the week of Black Friday & Cyber Monday. (I mean, how stupid do you have to be? They must be the laughing stock of the retail community. You don’t make major IT changes in Q4 period, let alone Thanksgiving week.)
And Demandware for agreeing to go along with it. Foolish. Maybe greedy. Sometimes you’ve got to say no to a customer.
As for the conversion issues? Of course conversion is going to go down when the website is redesigned (website design is not done by Demandware) and consumers are unfamiliar with it, don’t know where things are, need to learn to navigate it, etc.
And performance kinks on Demandware’s platform side? Probably. But those get worked out over time (except when you’re launching and freaking out on Cyber Monday). I don’t hear Demandware’s hundreds of other customers having any problems. Hmm…
Oh, and by the way, Finish Line still paid Demandware for the contract (they took a charge to write it off; it’s public record) which basically admits that Demandware did not fail to perform under the contract.
But would I expect Finish Line management to try to save face and point fingers? Of course. It’s human nature.