# Domain Investing & Website Development

## Understanding Search Volume

Michael Sumner • Nov 9, 2009 • 15 Comments

Every day you see sales pitches for domain names that include exact match search volume, but how many of you know how to use that number effectively in your purchase decision?  Many domainers subscribe to the “more is better” philosophy, while others prefer going after less competitive niche markets and shoot for monthly search volume in the 2k – 6k range.  Some just use it as a guide to estimate type-in traffic, although I’m not aware of a generally agreed-upon multiplier (I usually use 0.02).  But what does this valuable piece of data really mean when it comes to organic search traffic?

Back in 2006 there was a tremendous amount of AOL search data that was leaked, it included information on slightly more than 9 million searches and almost 5 million clicks.  A member of EarnersForum.com, Breakpoint, scrubbed the data to determine how many clicks each position can expect to receive.  This is fairly well-known in the SEO community, but I haven’t heard domainers talking about it, hence why I wrote this post.

To begin analyzing search volume, we first need to know what percentage of searches will likely result in a click.  Obviously this will vary, but because this is such a large and broad sample, we can assume the data will apply fairly reliably across multiple verticals.  So lets get started…

The AOL search data included 9,038,794 searches which resulted in 4,926,623 clicks.  From that we can assume that 54.5% of searches will actually result in clicks.  If the exact monthly search volume for the keywords in your domain is 25,000, you can probably expect there to be 13,625 clicks (25,000 * 0.545) on the organic search results across all positions.  The next important question is how those clicks are distributed across the positions.  Obviously a site ranking #1 in Google will get more clicks than a site ranking #7, but how much more?

Here’s the breakdown of the 4,926,623 clicks:

Position 1: 2,075,765 clicks (42.13% of clicks)
Position 2: 586,100 clicks (11.9% of clicks)
Position 3: 418,643 clicks (8.5% of clicks)
Position 4: 298,532 clicks (6.1% of clicks)
Position 5: 242,169 clicks (4.92% of clicks)
Position 6: 199,541 clicks (4.05% of clicks)
Position 7: 168,080 clicks (3.41% of clicks)
Position 8: 148,489 clicks (3.01% of clicks)
Position 9: 140,356 clicks (2.85% of clicks)
Position 10: 147,551 clicks (2.99% of clicks)

Positions 11 and beyond had 501,397 total clicks combined, or 10.18%.  That means ranking #2 would be better than having every position from 11 and on to yourself.  I find it interesting that ranking #1 gets you 3.5 times more clicks than #2, and ranking #10 is better than ranking #9 since it is the last thing someone sees before clicking to the next page of results.

TRAFFIC $\approx \!\,$ (EXACT SEARCH VOLUME  *  0.545)  *  PERCENT OF CLICKS AS DECIMAL

Lets go back to the example where the exact match search volume in Google is 25,000 per month, from which you could expect around 13,625 total clicks across all positions (remember we did 25k * 0.545).  If you can rank #1 for the term, you could expect to get 5,740 visitors (13,625 * 0.4213) per month from Google.  If you ranked #2, you could expect 1,621 visitors (13,625 * 0.119) per month from Google.  All you have to do is multiply the total expected number of clicks by the fraction of clicks a certain position gets.  So ranking #10 should get you about 407 visitors (13,625 * 0.0299) per month from Google.  Of course if you really wanted to do a thorough analysis of the domain’s potential you would add in search volume for long tail terms that you could likely rank for.  An example would be if you had CarInsurance.com, you might want to analyze other keywords such as “car insurance quotes” that you would likely be able to rank for.

If you have time, I would be interested to hear from the community about this.  If you have a site ranking on the first page of Google, run the numbers based on your position and the exact match search volume, then check your analytics and see how it compares to the real-world numbers.

Of course this leads to other questions such as how difficult it would be to rank for each position based on the competition, how much money you could expect to earn from PPC ads or affiliate programs based on the traffic you are expecting (and CTR/RPC), etc. but I will leave that for another discussion.  The first question is very tricky to come up with a guesstimate for, so I like to choose the position for the calculations conservatively.  There are many variables involved, most of which are a closely guarded trade secret.  You can make an educated guess by looking at the sites ranking on the first page.  How many backlinks do they have?  What is their page rank?  How long has the site been up?  How many established players are there (Wikipedia, Fortune 500, etc)?  Are there any .edu or .gov sites (these are hard to outrank)?  The earnings question is easy to ballpark once you have a traffic estimate though.

### 15 Responses to Understanding Search Volume

1. Bruce Marler says:

Mike,

Absolutely brilliant post. There are some sites I have put together that rank number one for their terms. I will do some analysis to determine how they match up to their search volume based on the formulas above.

I had read some data around this topic but never to the detail of this post or all in one place.

Thanks!

Bruce

2. Michael Sumner says:

Thanks Bruce, glad you liked it. I look forward to seeing what your results are.

3. Ross says:

I love this post and sheds light on some things for domainers but like you said it is nothing new. There are numerous tools out there that group all of those things you mentioned in the last paragraph together along with the estimated clicks ect. My favorite one that i have been using is Market Samurai. They have some awesome tools, although not the best in the world it is working great for a the couple of projects i have going.

Just a shame how little most domainers know about actually developing a site, i.e. SEO.

4. Alan says:

Great post !

5. Ty says:

Hi Michael,

Excellent post, and this is the sort of stuff that domainers should definitely be talking about.

There are a couple of other things that I think are worth mentioning about exact match searches:

1) As far as I’m aware, every separately viewed page of results is logged as a single exact match search, so hard to find niche areas have inflated figures – ie people searching look at multiple SERPs pages in an effort to find what they want, generating multiple exacts match searches from one session.

2) Often terms have more than one entry in the exact match search data, so this should always be checked. For example, as you pointed out, [mini sites] gets 1,300 Global exacts, but this needs to be added to [minisites] which gets 720 Global exacts – giving a total of 2,020 Global exacts matched searches for the term overall.

3) The percentage of clicks returned per position is heavily altered by Google sponsored links, shopping links, video links etc. This can mean a number 1 Google organic position can return as little as 5-10% of the available traffic – as these other entries grab the traffic equivalent of the first couple of slots.

4) Global exact match search data often varies from country to country. I’ve often Twittered Global Exacts to be told that the Global figure is appearing differently in different countries. (This obviously points to the fact that the figures are not absolute and should be treated with caution).

5) Finally. Exacts match search figures are very susceptible to ramping – any automated query for a term will appear in the figures (and this can be done deliberately), so it’s worth checking the keyword tool data against related terms, Google Trends and common sense before purchasing a domain off the back of them, especially if it’s a substantial purchase. I’ve seen a **lot** of very unlikely terms with enormous exact match volumes, often where the increase appears in the Volume Trends for no reason a month or two before a domain is offered for sale.

I’ve been developing based on Exact Match for a while and have collected quite a bit of comparative data between traffic and Exact Matches… so apologies for the long reply – but wanted to expand on some of the points naturally arising from your post.

Keep up the good work!

6. terek says:

for rectum.net I am ranked at position 4 sometimes 5 for rectum, rectum gets 200,000 exact searches but my stats say I only get about 500 clicks a month ?

7. Michael Sumner says:

Ty, great points, thanks for your contribution. I wasn’t aware that the global search volume can change based on country. I was playing around with GAKT a bit tonight and noticed something interesting. If you select all languages and all countries, the local doesn’t always match the global, which surprised me.

@terek – the Google Adwords Keyword Tool is reporting 18,100 local (English >> United States) and 40,500 global exact monthly search volume for me, not 200k. Also, I’m seeing you as #5 in G. If you take (40,500 * 0.545) * 0.0492 you get 1,086 expected clicks. If you do the local you get (18,100 * 0.545) * 0.0492 you get 485 expected clicks.

I’m thinking it’s probably better to go off of English, United States local volume since global seems to be a bit inflated and flaky. I’ve tried quite a few after Ty’s comments, and doing it that way seems to produce more accurate results.

8. Alan says:

As a FYI – its imperative to log out of your Google account if manually searching to see your site rankings by doing a keyword search. Google collects data based on your user experience and is known to show search results which are reflective of your past searching history if you are logged in.

Log out – then search to see how good your SEO skills really are.

9. Michael Sumner says:

Nice tip Alan, thanks. Another option is to search in a different browser where you aren’t logged in, that way you don’t have to keep logging in and out. Also, a great tool which lets you check your rankings across multiple search engines really quickly is below:

If anyone knows of any other good tools feel free to post below.

10. Alan says:

Great link and I respect anyone who can develop a double hyphen to get 40k visitors a month 🙂

11. Michael says:

Great post. I like quantitative analysis like this. This just provides another measure to apply to the level of a sites success other than simple pagerank, SERP’s etc.

Great stuff!

Michael

12. David says:

Interesting post. There is definitely a scale and diminishing results factor when you get into 2.2M search terms.

Also, local results along with Google’s News and other items affect things. A lot of that wasn’t a factor when the AOL data leaked.

–David

13. terek says:

@michael – Thank you, seems I was doing a broad search. pretty accurate formula.

14. Ryan K says:

This is an amazing analysis of not only figuring out what to expect an existing site to bring in when referring to visitors, but also when researching niches and where to concentrate your focus.

I’m going to be whipping up a spreadsheet based on these values you put out there to quickly plug in the search competition and the # of searches to quickly return a ballpark value. If nothing else you could use these values relative to one another to help you decide between what site to market next, etc… but my guess is this is a lot more accurate than many would believe.

Thanks for the excellent post!

15. John says:

Hi,
The distribution you describe is called a Zipf distribution and is found wherever people
are given a list of choices that are otherwise indistinguishable.
http://en.wikipedia.org/wiki/Zipf's_law

Best