By Anne Holland, President

These past two weeks, I've been in endless debates and design meetings trying to choose new colors for our revamped and expanded Web site.

If you've ever had to choose site colors, you'll understand completely. The three biggest problems:

(1) Everything's really subjective. What a color "means" can be personal or cultural, but it's not the same for everyone. Did you know baby girls wore blue and boys wore pink 150 years ago? Did you consider the green-means-money rule doesn't work outside the US?

(2) Everything looks slightly different on differing computers. Non-dithering hues notwithstanding, most people's screens look a little different. Laptops vs desktop monitors, old vs new screens, Macs vs PCs, varying background glare ... the differences are not massive. But they are enough to make a pale brown appear to be pink.

(3) There's virtually no data on marketing and color.

I know because I checked our site's new Research Database, which has more than 1,800 records, for stats on color. Very little came up.

Turns out, you can find loads of articles on the Web about color choices. However, most are based on hearsay instead of lab tests, cultural associations and/or broad generalizations that don't help much when you've got a palate of hundreds of hues to choose from.

My next step was to check out our Case Study Library with nearly 750 Case Studies. Did anyone test color choices?

Well, yes, they did. However, results were disheartening to a marketer stuck in a design meeting. Aside from the twin factors of legibility and good taste (based on target demographic), color tests were *never* a big factor in improving conversion rates.

The important factors were invariably things such as:
o Traffic source
o Offer
o Specific words in copy
o Ease of navigation (including lack of distractions)
o Relevancy of images
o Trustworthiness
o Reading comprehension (type size, type color, background color)

The last item on the list nearly always resolved to fairly big type (11-12 points+), in black "ink" on a white background. So that's not color so much as eye-enablement.

So, you could say to yourself, well, since color doesn't matter like this other, far more important stuff, I'm not going to pay attention to it. Let's pick something quickly and end the debate.

Except for one thing: branding.

It's how I found myself in this pickle in the first place. If you check the Wayback Machine -- http://www.archive.org/index.php -- for 2000, you'll see Sherpa's first site colors were bright red and yellow. These were chosen non-scientifically because they are my absolute personal favorites.

However, nobody else liked them, so I was shanghaied a few years ago into picking new colors so MarketingSherpa's brand would feel more "corporate."

What's corporate? We ended up with red and gray. Which looked fine, if a bit boring to me. Unfortunately, red and gray also looked fine to a bunch of our competitors. If you're a color-sensitive person, Jupiter, eMarketer and MarketingSherpa all looked pretty much the same.

We needed a new color scheme to stand out from the fray. Our first choice, nicknamed 'Operation Desert Storm' was finally vetoed because sometimes a palette of khakis and dark red don't have thrilling associations.

Our second choice, nicknamed 'Kindergarten' was far more cheerful, but also vetoed because, well, you can guess why.

We actually went live last week with our third choice, 'Brownie.' And then too many people on staff complained the softer hues looked unpleasantly pink on older laptop monitors. (I got emails from folks with the words 'Pepto-Bismol' in the subject line.)

Anyway, all of this is to explain, if you're in color choice meetings yourself, I feel your pain. Deeply.

And also, if you've been confused by the odd changes our site colors have been going through over the past few weeks (and days to come), this is why. Please bear with us. It will all be over soon, and then we can concentrate on the
important stuff.