How Web Graphics Affect Conversions

http://blog.kissmetrics.com/shocking-truth-about-graphics/

by JayInt • 14 years ago

98 39 14 years ago

blog.kissmetrics.com

Another tip for increasing readership: don't have annoying popup social widgets that jump in front of the text.

I'm glad I wasn't the only person affected by this. I couldn't finish the article. At first I thought it must be some sort of commentary related to the article (i.e. web 2.0 and social clutter) but no.

You can minimize the "no spam" widget but apparently are held hostage by the other box until (I assume) you click one of the social +1's.

Assuming the rest of the article's content is well thought out, perhaps they should forward it to their own designer.

aw3c2• 14 years ago

And people laugh at me when I say I browse with Javascript disable as default (added an enable checkmark in the bottom status bar, yay Opera).

drostie3• 14 years ago

Seriously. I had to click an X in the top right corner and use Firebug to set "display: none;" for three boxes just to get the distractions out of the way. It's a site called "KISSmetrics" which cannot "keep it simple stupid" and is telling me about how images can distract people from your articles?

One cute statement about their final image example: whenever I have to pay a new web service, there are just an absolutely tremendous number of options for me to be dealing with. In order to narrow down the field to companies who actually care about my time, I have some peculiar general rules:

(1) Name must not be concatenated CamelCase words. This still eliminates about 50% of my options, while GitHub is among the only known exceptions where it might be worth paying them for something.

(2) No autoplaying flash. This no longer eliminates so many services as it once did, but it should still be standard.

(3) SSL is a must. You would seriously be surprised.

(4) May not contain clip art of a white woman wearing a headset. It can be an Asian woman, or a white man, or she can be holding a phone, or it could not be clip art (here is our cofounder Olivia...), whatever. This actually seems to further limit the field by over 50%. Web service hosts apparently have really odd stereotypes, or else they expect me to.

I give that list here, again, because their last example of ineffective image use is, naturally, clip art of a white woman wearing a headset.

simonbrown• 14 years ago

What's wrong with CamelCase names?

talmand• 14 years ago

What is wrong with a white woman wearing a headset? I believe this says more about you than it says about a company that just grabs a random image of a woman with a headset.

reinhardt• 14 years ago

What annoying popups?

I guess AdBlock and/or Ghostery did a good job once again.

SonicSoul• 14 years ago

seriously. who uses that crap? i'm supposed to spam all my facebook contacts with 20 second reads i went through? and even if i did want to share it, i could figure it out w/out having a a big expanded dialog. this is useful probably to 1% of readers and annoys the other 99%.

z92• 14 years ago

Finished reading it. But where is the "Shocking Truth"? Wondering if I am missing something.

swombat• 14 years ago

The shocking truth is obviously that this was a linkbait deadline. I'll grant you that that's not very shocking, really.

paraschopra• 14 years ago

Shocking Truth was that shocking headlines really do work.

omonra• 14 years ago

Did you convert? :)

ysjwang• 14 years ago

I feel like the bombard-the-reader-with-images approach could be an artifact of thinking from the time when HTML images were still a novelty, simply because no one else had done it before. (I have absolutely no data to back this up though).

It just goes to show that quality > quantity.

Thanks for the link, good read!

troels• 14 years ago

Images in web sites? That hasn't been novel since .. well .. before html

wildster• 14 years ago

It's a shame that the kissmetrics.com does not follow this philosophy.

revorad• 14 years ago

The shocking truth about conversions is that businesses, including startups, spend crazy amounts of time tricking people into "converting", instead of building products that, you know, actually improve people's lives.

toumhi• 14 years ago

Don't you think KissMetrics provides a valuable service? A startup doesn't need to be focused on saving baby seals to improve people's lives.

viscanti2• 14 years ago

I think you're conflating "tricking" with "doing a better job of communicating". For most businesses (including startups), there's a clear value proposition. Their product/service actually fills a need/want. Iterating design/marketing copy is an attempt to find the best (or at least a better) way to communicate that value proposition. In the vast majority of cases, the goal isn't to trick someone into converting, it's to properly communicate the benefits of converting.

Czarnian• 14 years ago

There's an argument to be made (not necessarily by me) that if you have to spend a lot of time communicating your value proposition to your market, there may be something wrong with your product. The value proposition of your product should be self-evident.

mistercow• 14 years ago

> Ogilvy found that, on average, headlines placed below an image are read by 10% more people than headlines above. Since reading the headline is a prerequisite of reading the body copy, you’re losing a potential 10% of your audience if you’re distracting them with an image in the wrong place.

Actually, you'd be losing at most about 9% of your audience. I know that's a bit of a nitpick, but it bugs me when people equate "not X% more" with "X% less".

Also, to actually lose 9% of your readers to your headline position, you have to make the very optimistic assumption that you're not losing readers to any other factor.

scriptproof• 14 years ago

These Ogilvy's stats are about printed documents. I am sure it is false on the Web because we are so used with images. All major websites put the title above the image. See the Huff or Ars for example. And they know their business. For the rest, it may be true or partially true.

deepakprakash• 14 years ago

Most of it applies to photosharing websites as well - especially the bit about keeping the title/headline below the image. I'm irritated every time the title is at the top and I need to scroll up to see the title (The initial response for me is to inspect the photo - exactly as mentioned in the article).

Building one myself, I instinctively felt this should be below the image, but my designer disagreed and that's the way it is right now. We will definitely push this down - evidence always wins :)

dustingetz• 14 years ago

Not sure who kiss metrics is and I'm on my iPhone so not gonna check before posting, but this looks like a shitty Jason calacanis, mahalo style content farm article. There probably is something to the thesis "graphic design doesn't necessarily equal higher revenue" (I mean duh, see digg) but this post didn't support the thesis in a meaningful way.

JangoSteve2• 14 years ago

This being an article by KISSmetrics, I was disappointed there were no actual numbers or research done for the article. Unless I missed something, not a single statement was backed up with actual data. The only study mentioned in the entire article was in the context of ad copy and was done over 30 years ago (I'm assuming they're referring to a study from Ogilvy on Advertising [1] but I could be wrong).

For example:

If the images you’re using are not clearly tied to your value proposition, or to the central theme of your page, then they will only confuse your readers. At best, they’ll be pointless distractions. At worst, they’ll give the wrong impression and lead readers to feel tricked or disappointed.

In my experience, the blog posts I've written gain much more traction when they have images, even pointless stock images (I have experimented with this). So while I agree, that ideally, images should provide actual value to the story, I've found that a couple pointless stock images are better than no images. This is why I would have really liked to see actual evidence to back up their claims.

EDIT: Also, while we're on the topic, I also hate including all those social sharing widgets, but viewership takes a dramatic hit when they're not there.

[1] http://www.amazon.com/Ogilvy-Advertising-David/dp/039472903X

nhebb• 14 years ago

Blogs and landing pages should be treated separately. For a blog, a pointless stock image still serves the purpose of visually breaking up the article so it doesn't look like a big wall of text.

I agree with you though that I would have liked to see some numbers, but otherwise it was a good article that should at least give you some ideas on what to test next.

JangoSteve2• 14 years ago

I completely agree, which is why I was also confused that the referenced study was on offline ad design and copy, their first screenshot was a landing page, their second screenshot was an online news article, and their third screenshot was a blog post.

Kiro• 14 years ago

I thought it was more or less proven that stock photos (like headset girls) do in fact work, how dull it may seem.

Pent• 14 years ago

The annoying popups on the bottom of the article actually get in the way of reading it, to a point where I just close the page. They are definitely not following their own advice.

micaeked• 14 years ago

Here is a design tip for you: don't take up my screen space with useless shit I can't scroll away from

ootachi• 14 years ago

I thought the reason annoying generic stock images of salespeople smiling and wearing headsets were so common is that they did increase conversions, a lot.

billpatrianakos• 14 years ago

That is the reason, you are correct. What I've learned in the past two years is that the web is chock full of blogs that repeat tired old "tips" that are either outdated or just plain wrong and that other blogs that give fresh insights and new tips but wrongly give readers the impression that those tips are gospel (ie set in stone).

The trick is to do your own homework, don't take any advice as absolute, fit everything to your unique circumstances, and always remember that your mileage will vary. I've spent enough time reading a lot of crap and had to learn to be extra judicious the hard way. I take most articles like this as very loose guidelines. It was definitely a great article with good advice but I know better than to just run out and apply it just because the experts said it. Experts contradict each other all the time.

How Web Graphics Affect Conversions

Related Stories