|Search Engine Rankings to Click Through Ratios|
How much traffic will be generated by position 1 relative to 5 etc?
| 4:45 pm on Oct 23, 2001 (gmt 0)|
Does anyone know where I would find info about the differences in traffic generated by a number one search engine placement, relative to rankings 2,3,5,10 etc?
Many thanks for any input!
| 5:05 pm on Oct 23, 2001 (gmt 0)|
I would assume that one can only estimate these things.
| 5:11 pm on Oct 23, 2001 (gmt 0)|
From Overtures's FAQ:
'...our research shows that listings in the top three positions get double the number of clicks than positions 4-240 do combined!'
Doubt you'll ever get any more accurate stats than these.
| 5:53 pm on Oct 23, 2001 (gmt 0)|
I think there are two factors that trump rankings: screen position and click'ability.
It's not just the number, it is also where the user has to scroll. Below the scroll, near the scroll, and top of the scroll historically do worse than those that end in the middle of the page.
Put your ear out the window. Hear that whoosing sound? That's click through rates at Hotbot falling like a rock. Users now have to scroll for most search results.
2 weeks ago we averaged 550 referrals a day from hb across 52 high ranked top-ten moderately competitive phrases (most in the 3-7 range). Fairly stable rankings for a long period because most are direct hit influenced pages and in the core ink db.
By yesterday, nearly the same rankings (higher actually) were generating only 250 referrals a day.
That's more than a 50% hit in referrals simply because the user has to scroll now.
Click'ability is the click me now factor of any listing. Being listed is just the beginning. Top five rankings wont produce anything without click-me-now factor. I'm sure we've all seen top 10 rankings that just didn't seem to produce. We've also seen top 10 rankings that seem to out perform themselves.
Title makeup and descriptions are such an art form, that most seo's all but refuse to talk about it. Some of the most productive titles and descriptions would make you scratch your head as to why.
Those that produce have a few features in common:
- no hype. Powerful "call to action" statements just don't work like they used to do. They are counter productive. Three examples. I pulled similar to the extra descriptions straight from altavista searches.
a) WebmasterWorld, the best and greatest webmaster forum on the net.
b) WebmasterWorld, click here to visit now.
c) New and improved with daily updates. 100% fat free.
d) WebmasterWorld, news and discussion groups for the independent web professional.
Which one will produce the most clicks?
Which one will produce the most targeted traffic?
Lets look at them in order.
A: has the word "best" in the description. Adjectives and adverbs are getting "sub ranked" when they analyze the description. These type of "hype" words that try to make something sound better than it is, are getting beat down in some engines. They don't want hype in the descriptions. Go compare top ten rankings in alta with say 200-250. You'll start to form a word list of words that hurt you on there.
b: almost the same as A. Descriptions with the word "click" in them are history. Users generally ignore them now and engines also push them down.
Some commonly used description words you want to avoid. (new, fresh, updated, click, best, largest, cool, awesome, top.. I'm sure you can think of more and their derivatives)
c: It's interesting. It has curiosity and a touch of humor factor (humor is always good if you can slip it by). However it contains no keywords. It would actually have a fairly high click rate if you could get the page listed in decent position (highly unlikely without kw's in the title or description and because of the word "new"). The targeting would also be blown since users would have no idea what they were clicking onward to see.
d: Will out produce the other three by a factor of five to ten. Low hype, no poison words and pre-filters the users. Those types of descriptions and low fuss titles are really coming into vogue. They really produce and get you up in the rankings via click factor engines like direct hit.
| 6:14 pm on Oct 23, 2001 (gmt 0)|
Good post Brett.
I wonder if target market has influence on your post. Since you are targeting SEO's that are "programmer" types and also a field that is flooded with hype.
The programmer types are of the "only the facts" type (as in another thread they prefer text only email).
Have you found similiar response in markets more consumer oriented?
| 8:48 pm on Oct 23, 2001 (gmt 0)|
The top 2 listings at Overture are posted on all of their affiliates, and the top 3-5 on some affiliates. This skews the number of impressions, and therefor clicks for top biddings.
| 11:10 pm on Oct 23, 2001 (gmt 0)|
Nice post Brett! It would seem that D would be the best strategy to go with whether targeting SEPRs text link placements and maybe even banners.
Flashy banners tend to get more clicks, but often the higher conversions come from the more straight forward placements where in the words of GoTo, a clear path is present from the banne through the desired outcome.
| 11:36 pm on Oct 23, 2001 (gmt 0)|
Very well said Brett.
I like your mention of page positions. When I used to bid a lot at what was GoTo.com, when #1 was out of reach, I found #20 had better sales than #10 due to the position of the listing. #10 was off the first frame, but when the user scrolled to the bottom to hit the next button, their eye caught my listing. Turned into a much better ROI.
"Will out produce the other three by a factor of five to ten" <-- definitely not hype.
btw, thanks for the tip on those hype words, I had not put much thought into that.
| 10:57 am on Oct 24, 2001 (gmt 0)|
Hi Tabbie and welcome to WMW :)
You wont find a better answer to your question then the one provided by BT!
Echoing the others, excellent, excellent post!
There are no tried and tested formulas because of the variables.
| 12:11 pm on Oct 24, 2001 (gmt 0)|
Just out of curiosity, how do the SEs come up with their algorithms? I guess I am used to dealing with "definites" such as HTML tags rather than the "unknown" of SE placement! Aren't there definitive resources out there?
| 12:39 pm on Oct 24, 2001 (gmt 0)|
Brett, as always a clear glorious day.
wangdy, if anyone new the answer to that then it would be child play, but unfortunatley, like life, its complex, and experimentation and errors, are the best teachers, just keep at it and slowly a euclidaen like mental image will start to form, which will grow ever increasing, concerning design methodologies, right sort of submission routines, my favorite SE backdoors for free, and brillant ranking.
Drawing back to Bretts point, as a user, it is the title that grabs my attention, when the serps come, if the title is uninspiring and uninformative, then it could be #1,2 or 3, still i will drill to what i am asking for.
| 12:41 pm on Oct 24, 2001 (gmt 0)|
Excelent post Brett!
We experimented it on GoTO.
The possibility to change title and description is an excellent tool to find the best compromise about keywords and good copywriters.
| 1:49 pm on Oct 24, 2001 (gmt 0)|
"I would assume that one can only estimate these things."
... you don't know enough spammers:) roflhehehehe
I agree with everything said about quality and clickability, but if you are sales driven, the real issue is the conversion rates as you go from one to ten... granted, if you are third on the page but the first relevant listing, you might do better than #1.... BUT if you could somehow level the playing field, you would see that conversions go way down, quick, as you go down the top 10.
Imagine you had 10 listings that were exactly like BT's 'd' example (above) except they were all different:) (.... you know what I mean....) they were all relevant, low hype listings.... if that serp was live, and showed to a 1000 surfers, the first question is 'how many clicks?'.... the second should be, 'what is the conversion rate?'
I say #1 rank converts 3 sales for every 2 sales generated by #2.... given:
- equal amounts of traffic
- the term is something 'the masses' are looking for, and are willing to buy online if presented correctly.
...in other words, #1 is the place to be, not only in terms of quanity, but in terms of quality.... together you get a 'double whammy' effect. (That is the technical term:))
| 1:50 pm on Oct 24, 2001 (gmt 0)|
Great post, Brett.
One other factor I'd add to the clicks vs. position debate is the nebulous area of "quality of results vs. user expectations". Put simply, if the first 5 listings are dead-on for what the searcher is looking for, he may get no farther. If the results contain a lot of irrelevant (from the searcher's viewpoint) stuff, a listing in the teens, twenties, or beyond may get clicked.
This isn't something you can design for, although as Brett notes an on-target description will help you get the clicks you want. Many people are poor searchers, and use terms that generate a result set that has a low density of useful results. If someone searches on "webmaster", who knows what they are looking for? Do they want to hire a webmaster, get a definition of the term, learn about the career, find job openings, get info about a science fiction movie, etc.? Assuming they did have something specific in mind, they may scan through pages of results and only click on a couple of listings. On the other hand, a well-targeted search might produce the same or better information in the first few listings.
I think this variablity is a major reason (in addition to screen geography and other reasons mentioned by Brett) why clicks vs. rank defies accurate statistical analysis, and why tables that claim to predict "clicks per hundred exposures for a given ranking" should carry the warning, "your results may vary - a lot!".
| 11:45 am on Oct 29, 2001 (gmt 0)|
I forgot one: all cap words in titles and descriptions. There is some sort of algo at work on it with one engine, but a couple others don't mine. On one engine all cap words in the description are as good as asking to be buried in the results. It's certainly on the out list now.
Also, urls in titles are out. I used that one in 98 and part of 99, but Alta started filtering for it.