In its second-largest acquisition, Amazon buys the company for $970 million.
When it comes to coloration, the chameleon has nothing on the web. Near-infinite variations in the hardware and software combinations that deliver color images to millions of computer screens can turn the yellow shirt into orange and the red hat into pink. Web shoppers may be in the dark-until the package arrives from the online store.
Even the tech-savviest among them aren’t immune to problems. “I ordered a coat over the Internet and it turned out to be a totally different shade of green than what I saw on the monitor,” says Lia Schubert, an analyst at Boston-based InfoTrends Research Group, a market research firm. Back went the coat, and down went Schubert’s level of trust in color representation on the web. “Even with my awareness of this issue, I was still shocked,” she says.
Early shoppers didn’t buy much apparel or many cosmetics online, so sharpening up color reproduction has until lately ranked low in the priorities of e-retailers. “They’ve had other issues to deal with, like getting their product images up online at all,” says Schubert. But as more web shoppers venture beyond books and gadgets, and more have Schubert’s experience of sending for-and sending back-color-sensitive products that flunk the front-door test, online color correcting is taking on a new urgency for Internet sellers of apparel, furniture, cosmetics and art.
Research firms are tackling the color issue in increasing numbers. And as they pile up data, the dotted line connecting web color representation to online shopping behavior starts to look more like forged-steel links.
Here’s a sample: in a study by InfoTrends last December, 81% of those surveyed said concerns about inaccurate color depiction affect their decisions to purchase color-dependent products via the web. A BizRate study in March found that 50% of consumers would not make future purchases from an online merchant that delivered an item in a color that didn’t match their expectations. And 60% of online consumers in a recent study from Cyber Dialogue don’t trust the colors viewed on their monitors while 30% bailed out of online purchases due to uncertainty about the color.
Much of the research generated so far to demonstrate shoppers’ concerns about color depiction online has been sponsored by color solutions technology developers. The studies above were commissioned by Imation Corp. and E-Color Inc., two leaders in the growing field. But that doesn’t negate the importance that consumers attach to color representation-in fact, most retailers probably know intuitively that how a color is depicted on a shopper’s computer is important. Independent research shows that e-retailers aren’t waiting to hear from customers about color concerns; the issue is already rising on their radar screens of its own accord. In a May report from Forrester Research on online fashion marketing, e-retailers ranked color correction at 3.5 on a scale of 5 in importance to sales.
Internet color 101
To understand why some e-retailers are now using color correction technology, consider that the image of a red sweater delivered to a shopper’s computer screen may look nothing like the real thing by the time it gets there. Multiple factors affect the color viewers see online. The model and make of the monitor, its age, variances in circuitry, video cards, and operating systems are some of the factors that cause color shifts-and that’s just at the user end.
At the front end, where product images are captured on camera and color depiction gets digital, transferring color from one imaging format to another constitutes yet another set of opportunities for color to stray off the mark. “The fundamental problem is the physics of color,” says Stephen Mace, CEO of Praxisoft LLC, a Sterling, Va.-based provider of digital color matching software. “To try to network color when it’s first acquired through an image being scanned or photographed by a digital camera you have to bring it through different color gamuts.” A gamut is a device’s range of reproducible colors. “Each device has its own gamut.”
By the time the online image of the red sweater reaches the color monitor, it may look very different to different viewers. The situation is comparable to that of walking into an electronics store and seeing a dozen TVs all tuned to the same channel. All the TV receivers are getting the same signal, but they present it as something slightly different on each screen. For similar reasons, web shoppers may log onto a common web site but see different color shades of the same images.
How different? Color changes on the web are measured in units called delta Es-one dE being equal to one perceptible color shift. When E-Color conducted a baseline study last year with client Bloomingdales.com prior to deploying its technology on Bloomingdale’s site, it looked for what percentage of site visitors needed some level of color correction on their computer screens, and how much. The study found 87% of visitors were viewing monitors that needed significant color correction. On average, getting to an accurate color representation of the product required a change of 16dE. What that means is that the monitors displayed color 16 shades away from the real thing, a degree of variation roughly equal to the difference between, say, red and pink.
Server-side color correction systems, such as E-Color’s, gear color adjustments to the needs of individual monitors based on data from a cookie attached to the user’s computer. The cookie is generated when the shopper follows online prompts to calibrate color-an opportunity typically presented to users when they install a new system or an upgrade of equipment from a manufacturer with which E-Color has an OEM deal.
Users who don’t have new equipment or upgrades can still download the prompts and install the cookie through links on the web sites of merchants that feature E-Color correction. Once installed, the cookie automatically alerts the server whenever the user calls up any site that uses E-Color technology, and the server dishes up images color-corrected specifically for that monitor.