More Proof That Users Don't Like Ads That Smack Them In The Face, Marketing Experiments Study Options

[Scottie]

This study confirms something many people have known or suspected for quite a while- ad blindness is very widespread and things that look like ads aren't clicked nearly as often as things that look like part of the page:

http://www.marketingexperiments.com/improv...online-ads.html

This is really critical for e-commerce sites who think they need big, flashing calls to action to get people to buy- I thought the second part of the study that showed higher clickthrough rates on a text link vs an image button were really interesting.

[jehochman]

Thanks! This is exciting stuff. Some clients like to hang ornaments all over their sites. Every time they take a hot shower they get a new bright idea. I am very glad to have this contrary evidence.

[Scottie]

It's pretty compelling too- doubling the clickthrough rate is substantial.

[lisety]

I'm signed up to the same newsletter and also found it quite interesting as it also proves true for Adsense. The more it blends into your site, the better...not that I've really figured out Adsense since the new rule about no text near images but it's slowly improving since I changed my ads to reflect the colors and theme of my site.

[mcanerin]

This seems disturbingly close to misdirection, if you ask me.

Basically, my understanding is that the more you disguise the ad so that it looks like the content that the readers trust, the more likely they are to click on it.

I see that as a violation of trust, not smart marketing, sorry. Ad Cloaking, if you will.

If I went to a site that I thought was trustworthy and began to find myself on ads (and being redirected off the site in the process) I'd stop going back.

Perhaps that's just me, but I'd like to see some other data, like:

1)What was the conversion rate after the click? If you've just been tricked there, I suspect it's low.
2)Did the visitor return to the site again after clicking on the ad, and if so, did their behaviour change afterward?
3)What is the visitors perceived trust of the site before and after the advertsment click? I doubt it changes much for visible ads, I'd like to see the data on camoflaged ones.

I've watched my father surf the web looking for something. He types in a search (he can't tell the difference between an address bar and a search bar - or doesn't care, I'm not sure which), then chooses a page.

Once he gets to that page, he seems to think that everything that looks like content is part of the site, and everything that looks like a banner isn't. So time after time he clicks on those adwords that people put in the middle of the page, looking like content with colors that match the site, on made for adsense sites. I'm sure someone somewhere thinks this is great marketing.

Wasn't Googles big thing over Yahoo and Ask being that Google actually made it clear (or at least, more clear) the difference between ads and content? Didn't that increase their credibility, since they gave up some effective marketing in order to do so?

There is no doubt that this kind of thing works, especially on vulnerable or inexperienced surfers. I just wonder if the ends justify the means. It rarely does.

My personal opinion,

Ian

[lisety]

Good points, Ian (IMG:style_emoticons/default/appl.gif)

[Scottie]

What interests me in this isn't the ad blending (although you make good points there, Ian) but the small elements you may put on your site to attract attention may actually be "invisible" to users. The text link vs image button test is very eye-opening, IMO.

[projectphp]

Dude sounds like Dubya! Actually, he sounds like John Stewart's impression of Dubya.

I'm with Ian: there is missing information here. Reallistically, if the goal is more than just clicks, then this is advice needs one more step: what the people did post click. If you pay CPC, then that fact becomes even more important.

Some questions I have:
Do the same people click the ads in both case, i.e. is it all the same people + more, or different people?
Which group converted better?

Normally, I really like these experiments, but in this case, there was a lot of missing info, such thatI dont really know how to use the findings.

[Jbrookins]

I would agree that there is a line that you need to be careful not to cross. Deceptive marketing practices are not something we condone, though where the line on that actually is tends to vary even within our own ranks.

It's important to note that the test was not "make the ad look like an article", but rather "does flashy/stand out work?" and the answer seems to be "no". At no point was any ad created to be deceptive. We just simply removed the flashiness, or rather designed the ad with the advertising site in mind. What we determined is that users seem to automatically ignore anything that appears out of place, so by removing that barrier, we got a greater amount of consideration.

We'd love to have more data as far as user reactions, conversions and such, but the advertiser sites were not research partners and the product itself was low-conversion by nature. I think we plan on expanding this test to areas that may overcome those obstacles in the future, but I'm not part of the test-planning team, so I really couldn't say.

[smc_online]

Hmm... both sides make a very good point!

To me every page on a site (regardless of audience or purpose) is a landing page if its planed and architected appropriately as each page should have a reason for existence (inform, sell, persuade, etc.) with a call to action (sign up to be informed, buy, bookmark it, send to friend, etc.). I'm sure there are exceptions out there that don't immediately come to mind, but for the most part, all pages should take the user to the next step what ever it may be.

So, I wonder, can this concept of "blending" be applied in a site's information architecture of in general?