Ever wonder what an organic click might really be worth? For that matter, ever wonder what is the number of clicks you might gain by moving from position #5 to position #1 in the search results?
What if I told you that there was information out there that might be able to help you predict just how many organic clicks you may stand to gain by moving up in search engine rankings? Well, it’s your lucky day.
The Click Distribution Data
There are 3 data sets & studies currently in existence to help us understand traffic distribution by SERP position:
- AOL’s 2006 Data Leak – In 2006, AOL accidentally leaked over 30 million search queries from over 600,000 AOL search users. This has been discussed many times since then. See Richard Hearne and Bob Hodgson.
- Cornell University Eye-Tracking Study – In 2004, Laura A. Granka, Thorsten Joachims and Geri Cay conducted a user-behavoir study focused around search behavoir specifically on Google. They used an eye-tracking study of a sample of undergraduate students to determine clicks and attention distribution.
- BrandSoftech.com Study – New software recently developed by BrandSoftech allowed them to track and create statistics based on over 63 gambling sites, which received 5,357,519 clicks from 29,327 different key phrases typed into Google. The ingenious software looked up a site’s position for a key phrase the moment it received a click from Google.
The 2006 AOL Leaked Data:
The Cornell University Eye-Tracking Study Data:
The BrandSoftech.com Click Distribution Study Data:
By looking at each individual study, you can easily see that there is a great disparity in terms of the number of clicks the #1 result receives versus #6 or #10. This is to be expected. AOL’s data also showed that 89% of people clicked somewhere on page one, while Cornell’s users click page one results 98%, and BrandSoftech’s users click page one results 99% of the time.
By my reasoning, I believe that none of the 3 data sets referenced above is 100% accurate. In fact, each data set has it’s share of flaws. However, I believe that the real click distribution percentage for clean SERP pages (you’ll see muddy SERP pages referenced below) likely falls somewhere in between all three studies.
So, I took the liberty of average all 3 percentages together to come up with a new mark:
- Position #1: 45.46% of all clicks
- Position #2: 15.69% of all clicks
- Position #3: 10.09% of all clicks
- Position #4: 5.49% of all clicks
- Position #5: 5.00% of all clicks
- Position #6: 3.94% of all clicks
- Position #7: 2.51% of all clicks
- Position #8: 2.94% of all clicks
- Position #9: 1.97% of all clicks
- Position #10: 2.71% of all clicks
- Total: 95.91% of all clicks occur on Page #1 of SERPs
Please keep in mind, that this is NOT an exact science. This is only a way of attempting to estimate the percentage of click volume you may receive based upon SERP position.
What You Can Do With Click Distribution Data
The idea here is that you can utilize both your rankings data and keyword search volume data to determine:
- How many clicks you might receive my achieving Top 10 positioning for a particular keyword.
- How many clicks you can expect to gain/lose by moving position in the SERPs.
Okay, let’s show an example. For keyword research, I recommend using the Google Adwords Keyword Tool. I sometimes question the accuracy of this tool, but it’s free and easy to use – and also allows you to export data via excel. This data can show you monthly search volume trends on a keyword level.
So in our example, I chose the word “shopping.” According to the tool, the keyword “shopping” had:
- 24,900,000 Global Monthly Searches
- 5,000,000 Local Monthly Searches (i.e. In Your Country (United States is mine))
Lets go ahead an just knock off about 20% right now due to the click volume that will probably go to Pay-Per-Click ads. That leaves us with approximately:
- 19,920,000 Global Monthly Searches
- 4,000,000 Local Monthly Searches (i.e. In Your Country (United States is mine))
Based on the click distribution percentages above, you might expect approximately the number of clicks displayed below if you rank in the Top 10 SERPs for the keyword “shopping”:
Applied to Global Search Volume
- Position #1: 9,075,552 clicks (45.46% of all clicks)
- Position #2: 3,126,112 clicks (15.69% of all clicks)
- Position #3: 2,009,264 clicks (10.09% of all clicks)
- Position #4: 1,094,272 clicks (5.49% of all clicks)
- Position #5: 996,000 clicks (5.00% of all clicks)
- Position #6: 785,512 clicks (3.94% of all clicks)
- Position #7: 500,656 clicks (2.51% of all clicks)
- Position #8: 584,984 clicks (2.94% of all clicks)
- Position #9: 392,424 clicks (1.97% of all clicks)
- Position #10: 539,832 clicks (2.71% of all clicks)
Applied to Local Search Volume
- Position #1: 1,822,400 clicks (45.46% of all clicks)
- Position #2: 627,733 clicks (15.69% of all clicks)
- Position #3: 403,467 clicks (10.09% of all clicks)
- Position #4: 219,733 clicks (5.49% of all clicks)
- Position #5: 200,000 clicks (5.00% of all clicks)
- Position #6: 157,733 clicks (3.94% of all clicks)
- Position #7: 100,533 clicks (2.51% of all clicks)
- Position #8: 117,467 clicks (2.94% of all clicks)
- Position #9: 78,000 clicks (1.97% of all clicks)
- Position #10: 108,400 clicks (2.71% of all clicks)
Personally, I recommend using the Local Search Volume as your primary area of focus unless you’re a very well-known worldwide brand. So if we go by the above logic using the Local Search Volume numbers, you can expect that if your site moves from the #5 organic position (which receives 200,000 clicks monthly) to the #2 organic position, you should get a monthly increase in organic traffic of around 427,733 clicks.
What Is a Click Worth To You
Lets use the above example, and really roll with it into conversions. Let’s say your website’s conversation rate is 2%, and you’re currently ranking #5 for shopping which we think pulls 200,000 local clicks monthly. Each conversion for your company is worth $200.
At the #5 ranking, your site should be generating 4000 conversions per month which equates into $800,000 monthly.
A move from #5 to #2 increases your organic traffic, which increases your number of conversions (if your conversion rate holds) from 4000 to 12,555 conversions. At $200 per conversion, this new revenue number becomes $2,511,00 – a difference of $1,711,00!
Please, before you get excited I want you to keep in mind, all of this is approximate. This logic does not occur in a vacuum, and will most likely happen on a MUCH MUCH smaller scale for most examples. But in the end, this should serve as a good starting point from which to begin to understand how apply the data points that we are aware of to understand how shifts in your organic ranking may affect your organic traffic.
The Flaw in the Logic
As I said above, there are several flaws in the logic I stated above.
Different Types of Searches
Different types of search classifications is discussed in detail by Aaron Wall over at SEOBook. These types are Navigational, Transactional, and Informational.
- In general, for navigational searches people click the top result more often than they would on an informational search.
- In general, for informational searches people tend to click throughout the full set of search results at a more even distribution than they would for navigational or transactional searches.
- The only solid recently-shared publicly data on those breakdowns is from Dogpile [PDF], a meta search engine. But given how polluted meta search services tend to be (with ads mixed in their search results) those numbers were quite a bit off from what one might expect. And once more, they are aggregate numbers.
Different search contexts certainly impact the patterns in which people click around the SERPs, which in turn may affect the click distribution percentages.
The concepts discussed above occur in the vacuum of a clean search results page. However, we know different. SERPs are muddied with local listings, inject with news feeds, social media, pictures, media, etc.
This means that there aren’t always 10 static listings in the SERP, which in turn means that each listings click distribution gets watered down. In addition, you must consider the effect of personalized search results which may display different listings for different users – making the click distribution percentages even less reliable.
In the example above, there were actually more than 15 listings of one type or another on the search result page.
What To Do?
Probably the best bet is to take whatever numbers you come up with utilizing the math above and knock about 10-20% off of them due to the invariable flaws and differences in search results pages and search types.
Also, set your expectation as such so that you’re not recreating these experiments in a vaccuum and then coming to me later if they’re not exactly right. Remember, this isn’t an exact science, simply something that is meant to guide you along the path of understanding how search rankings impact traffic and potentially revenue.