60 replies to this topic

[haylieplum]

Hello,
When someone is doing a search on google or yahoo, let's say, does anyone know if statistically they check out the organic listings before the paid ones? Or is it vice versa?

[SearchRank]

Most studies I have seen say that 75-80% click on organic listings while the remaining 20-25% click on sponsored or paid listings.

[madams]

Yeh... and 90% of statistics are wrong!

[ChipJohns]

Overall statistics can prove just about anything. For the most part we will do better with organic placements. However, circumstances can vary.

What would happen if we broke down the 20 percent that click on sponsered adds? Will they prove to have a higher conversion rate? Probably not in general. But how about in YOUR industry?

The best thing to do is TEST. It is really the best way to answer the question. Try running a campaign and see what the answer is for you, your product, and your Web site.

[bluenote]

The best thing to do is TEST. It is really the best way to answer the question. Try running a campaign and see what the answer is for you, your product, and your Web site.

I agree about the testing. And what do you get as a result of testing? Statistics! :-) However, at least these statistics would be more applicable to your specific situation and therefore easier to draw some actionable conclusions on. The sampling size will likely be too small to consider on its own, but it's amazing the insights you can gain from testing. This is one thing that helps us base decisions more on the users and less on ourselves.

[SearchRank]

Yeh... and 90% of statistics are wrong!

Do you have proof that 90% of studies are wrong or are you just stating your opinion? There is a big difference between the two.

[Selie]

It's a saying, sort of a joke... since it's so easy to make statistics up on the spot, and even researched ones can be heavily biased (or just inaccurate from too many uncontrollable factors).

Kind of a self-referencing joke.

Unless you were already familiar with the joke and now I'm the one missing out?

[mpea]

From personal experience I’ve always found that I would prefer to click on organic results, the reasons are:

1) I know the sites that pay for advertising are obviously after your money for some reason and not just there to help you.
2) The more money they have paid for adds probably means the more you are likely to see them, but also to me means the more i will have to pay extra for that product!
3) Imo, top listed organic sites are probably more likely to have been around alot longer (depending on the field), since new pay sites can spring to the top by just paying more. The longer they have been around the more I trust them.

However, now almost everyone has paid adds on their site I’m not sure what the average user would prefer to do?

This is just my opinion

[ChipJohns]

Actually, testing usually produces data. It is this data that is used to formulate statistics.

Statistics in themselves can be considered data, but in reference to statistical analysis Statistics is more the classification, analysis, and interpretation of numerical facts or data and more specific the use of mathematical theories of probability on such data.

The problem with data is that it can be interpreted in many different different ways.

For instance to ask 100 or even 1,000 people a question and see what the results yield can produce completely different results when a different group of 100 or 1,000 people are asked the same question.

Why does someone do something? The answer usually lies beneath the surface. Many things can be quantified, but everything cannot.

Knowing what data is valuable and what data is not and then performing the proper analysis on the data provides the most accurate statistics...

The best proof of how valuable statistics are is in baseball. The best managers use statistics and they win due to knowing how to use the statistics in the right way.

[Clintorius]

Hmm, I remember from my class that most statistics are made so that only 5% of it is wrong

I read recently (can't find the source at this moment) a paper about eye-tracking analysis (that is very hard to fake) that shows people are actually looking at sponsored links at the top of the organic listings BUT they proceed down the list in the majority of cases and click a link in the organic department.

So the adds are not ignored, they are just not selected for. If you got to pay to come to the top (like doping in sports) you can not be trusted, the searchers seem to think (there is no evidence for searches actually thinking BTW).

C.

[Randy]

Like a lot of things it's almost impossible to generalize. There are simply too many factors one would have to rule out. Probably the largest being what the Title and Snippet say in the PPC vs Organic listings.

This ad copy --that's what it is after all in both cases-- is very easy to control with PPC for any search term. More difficult to control with organic listings. The only way to really test it would be to have a PPC ad that gave exactly the same title and snippet as what came up in the organic listings, at the same level of the page. Then see if one got clicked on more than another.

No test I've ever seen have even attempted to make sure the two say the same thing. Thus any conclusions reached pretty much have to be dubious at best.

[Tom Philo]

When you do any testing / surveys the "domain" of the source is important!
Usually you see most surveys state that 1000+ people were surveyed and the margin of error is +- 3%. Unless they state what "domain" they picked it can be perfectly valid statistic and still totally wrong.

Survey of 1000 people but all the people were women at Boston University would yield a total different result than giving the same survey to 1000 men at the Engineer department of University of Seattle.

The domain has to be disclosed and referenced so you can see the built in bias of the survey.

Note: most "public" surveys never disclose or publish the raw data so they should be distrusted as "mostly meaningless" and should be ignored.

[ewo]

For what it's worth....

[adybee]

got these figures from Iprospect 2005 Survey

which are considered the most relevant results by SE users:


SE Organic Paid
Google 72.3% 27.7%
Yahoo 60.8% 39.2%
Msn 28.8% 71.2%
AOL 50% 50%

MSN no surprise

[ngupta]

its all depends on what type of user is surfing the net the general user sees paid results on the first position, and clicks them,

but i think organic results are more clicked.......