Studies on influence of meta description on CTR

[Theo-NL]

After having answered quite a lot of questions here, I figured it was about time to ask one of my own.

Can anybody point me to decent (experimental) research articles or blogs that actually show variations in meta descriptions influence the Click Through Rate (CTR) of searchers?

On dozens of websites on the internet it is stated that 'meta descriptions affect CTR', but (good scientific) sources for those statements are nowhere to be found. The only research I can find that comes closest to providing any evidence is a translated study by dynamical.biz, which states that searchers LOOK a lot at the meta description, but this study (atleast in the translation) mentions nothing of searchers actually CLICKING it.

Any help would be greatly appreciated!

[DanDeceuster]

I wouldn't necessarily look for stats or studies about meta descriptions here, but look at research done on PPC copy as well. There is likely a whole lot more of it for one, and its a very similar concept. Which words get more clicks? That's what it boils down to. I'm sure if you look more for stats in the PPC side of things you can find more concrete sources on what language is most effective.

[kdaly100]

I havent seen any studies but you could create two pages with different meta and see the results. It is hard to be empirical with this I would think as Google now limits it to two lines but ther could be options for A/B studies that you could follow?

[Marcus_Miller]

That's a great point from Daniel - Thumbs up! When people are paying for those words, you can bet the data is out there.

Scientific data for this kind of thing is tricky though, have you done testing of your own on your own projects? I have seen improvements for clients of around 30% with improved meta tags but it's very much a test, tweak, test, tweak, test, tweak (ad infinitum) process.

[Theo-NL]

Thanks a lot for the responses so far people! What I would like to use this data for is to back up the argument 'that good meta descriptions boost organic CTR' during a presentation. While I'm almost certain I'll find evidence that this is the case for ppc, I can't just generalize that data to organic seo. As kdaly100 suggested I'd think this is actually testable in a lab study. Present people with manipulated SERPS varying just one element at a time (meta description, title, url, etc.) and see what result gets the highest CTR. Strangely enough there doesn't seem to be any (public) studies that investigated this?

[AlanMosley]

People were paying riduclus prices for houses here in Western Australia, now rthey wish they didnt. People tend to follow each other when spending money, if he is spending that much then its ok for me to do so, atitude.
This is how we get bubbles.
Having said that i dont know any other stats you could use, but I do think description is very important for conversion.

[Breakout]

I also gave that a thumbs up, but I think that some of the claims about PPC tweaking are over-blown.  I can see tremendous variation in CTR and conversions on my keywords WITHOUT making any changes.  I think that people spend a lot of time making changes based on what is essentially random variation:

1. Ad A is shown to 1000 people, 200 of whom are actually interested in buying the product
2. Ad B is shown to 1000 people, 50 of whom are actually interested in buying.

Ad A will show a big edge in CTR and conversions, but this will have nothing to do with the content.

[KeriMorgret]

For any type of test, remember that an engine doesn't necessarily show the meta description as the snippet in the SERP.

[KeriMorgret]

Theo, did you ever find any more information on this? Would love to see anything you did find, as it is a topic others have asked about. Thanks!