Should Hypertext Links Be Blue And Purple

According to Gerry

McGovern, hypertext links should be treated like traffic lights, and always

use the same colors in a pre-established order: blue, followed by purple.

Obviously, for

reasons of efficiency and, above all, safety, no one would ever dream of replacing

the three colors of the traffic light system with colored neon lights, even

if it did make them look better (Gerry McGovern "Web

navigation: traffic light, not neon light design").

Both Gerry McGovern

and Jakob Nielsen believe

that it would be a mistake to use any colors other than blue for unvisited links

and purple for visited links.

Although I agree

with McGovern that the primary aim of site navigation features should be efficiency

and that making them more flashy usually only confuses visitors, I don't believe

that the colors chosen are all that important.

In my opinion,

the main way of characterizing a link is not by using a particular color, but

by underlining the text itself.

If the color was

the main characteristic of a hypertext link, all we would have to do would be

to make certain parts of the texts on our sites blue to let users know that

they were clickable. Allow me to doubt the efficiency of this rather simple

tactic.

In my opinion,

the color means nothing if the text is not underlined. The color of a link can

make users more aware of it, but it is only of secondary importance. Moreover,

putting underlined links in bold can work just as well.

It is interesting

to note that, on McGovern's own site, Nua.com, the links in the center of the

page are indeed blue and then purple once they have been visited, but that blue

is also the color of the site's left- and right-hand columns. Choosing the color

blue for its links is therefore totally in harmony with the site's color palette,

and works well esthetically.

However, we also

noticed that, for readability purposes, the hypertext links in the left-hand

menu are white on blue! What's more, they do not change color once they have

been visited.

Like the menus

on McGovern's site, the great majority of e-Commerce sites have now given up

using the standard blue/purple colors for links.

This was one of the findings of our study of the homepages of the 100 top American

eRetail sites ("Homepages

that Sell").

The study sample

is composed of sites with browser-to-buyer conversion rates of between 2% and

30.3%. These sites are visited by 120 million users every month and are therefore

totally representative of Internet traffic.

The results of

our study show that, as far as hypertext links are concerned, only 27% of the

sites still use the "standard" blue color for unvisited links. In

other words, 73% of these e-Commerce sites now use links that they have customized

in their own colors.

Example

of customized links on oneHanesPlace.com

(old and new version of the site):

We can therefore

conclude that blue/purple "standard" no longer exists - in any case

as far as e-Retail sites are concerned.

My second reflection

concerns the question of whether visited links should change color or not.

61% of the sites

in our study do not use different colors according to whether the links have

been visited or not. Worse, only 13% of them still use the color purple for

visited links.

I'm not trying

to say that these figures demonstrate that the standards championed by Nielsen

and McGovern should be given up for good, but simply to point out that it is

quite clear that users today are not disorientated and/or disturbed if sites

don't use them.

Moreover, when a user visits a site selling clothes, for example, he probably

does not need to be reminded whether he has already visited the menswear, sweater

or tie sections: he is perfectly well aware which categories he has already

looked at.

The same reasoning

is true for many of the main "repetitive" pages habitually found in

the navigation structure of a retail site.

This leads us

to question whether we are offering users a real service by changing the color

of visited links or whether we are not, in fact, making navigation more awkward

for them.

In fact, I would

go even further than this and say that it can sometimes be counter-productive

for a site to change the color of its links for certain categories (Best sellers,

for example): the user may not revisit links if their color has changed, although

the site's commercial success may well depend on these categories being regularly

revisited.

Admittedly, my

approach goes further than the usability aspect alone, since it is mainly focused

on the eShopability capacities of e-Commerce sites.

Choosing the color of links on a retail site and changing the color of visited

links is no longer merely a question of adhering to standards, but of deciding

whether or not they fit in with the site's commercial goals.

I'm not saying that Nielsen and McGovern are totally right or wrong, but quite

simply that we should be specifically approaching the subject of retail site

usability from a new angle: eShopability.

On an information, a press or a university site, for example, the function of

a link is to let readers know that an underlined text will give them access

to another resource, either on the site itself or on an external site.

This type of link corresponded perfectly to the needs of the first sites that

appeared on the Internet. However, the aims of these sites were quite different

from the aims of e-Commerce sites today.

In the "informative"

context, changing the color of a link is totally justified: it effectively reminds

the visitor that he has already accessed the information connected to that particular

link.

This is particularly

useful for searching the site's archives, for example.

However, the things

that make this type of site easy to use do not necessarily work so well on e-Commerce

sites.

That said, each approach has its own merits and the two can sometimes be used

together in a complementary way.

There is no point,

then, in discarding one system or the other; it is much better to use them appropriately,

not automatically, according to the aim of each particular site.

I believe that

Web standards should evolve as users' experience grows. If we try and freeze

certain criteria, by likening them to traffic lights, I don't believe we will

be doing much of a favor either to retail sites or to users.

Luc Carton - eShopability.com