Can I rely on Keyword Difficulty tool?

[Darroch]

I just ran into a problem that I hadn't expected. Testing the Keyword Difficulty I saw the results contained a result for a page that has Domain Authority=1 and Page Authority=1.

As a result, Keyword Difficulty was reduced (compared to last month), which may actually be reversed if the site is crawled. Sadly, I didn't run the report on the figures as it was a small project.

Questions:

1. Can I rely on results shown by Keyword Difficulty?
2. Are results where Domain Authority =1 are used to calculate Keyword Difficulty? If so, why is that?
3. Is there any difference between a page that has received no links and a page that OSE/Mozscape has no link data for?

The problem

Using the Keyword Difficulty tool, I found swings of up to 14% in Keyword Difficulty (between Oct -Nov). Dr Pete may suggest that this is because of changes in Google's index ( http://www.seomoz.org/blog/a-week-in-the-life-of-3-keywords ). However, It would be helpful to have a figure for Keyword Difficulty that isn't affected by the gaps in the Mozscape data.

The (bad) solution

You can mirror something close to Keyword Difficulty using:

=(Sum of Page Authorities + Sum of Domain Authorities )/20

Right now, I have resorted to manually calculating keyword difficulty. I use the SEOMoz Page Authority & Domain Authority figures and a quick splash of Excel SUMIF and COUNTIF.

I find the results don't look as 'easy' when I can ignore results where the data is unknown (PageAuthority=1 & DomainAuthority = 1).

Background Info

One result I still have a report on is for the phrase [fixing your business puzzle] using US Results on Google. For the specific result, I found the additional information about the site:

DNS lookup shows the domain was registered in 2010
Archive.org shows no records
OSE shows no data for the site
Site uses https
Google showing No links
Robots.txt file seems fine
No Sitemap.xml

[JoelDay]

Hi Darroch,

Thanks for reaching out.  The Keyword Difficulty score is actually put together using a proprietary algorithm from many different metrics, not just the PA & DA. However, I'm glad you're finding a benefit from your formula as well. I definitely encourage folks to develop their own techniques and formulas like that.

As for the reliability, the short answer is: Yes. We are confident that our difficulty scores are a decently accurate representation of the competitiveness of the keyword.

The question about how the KWD score is affected by unindexed pages, I unfortunately don't have that information for you. They are going to be factored in, as they should be, but the weight will obviously be less. It's unlikely that we are not going to index a high authority page so the actual effect is likely little to none.

I hope that helps.
Cheers,
Joel.

[Darroch]

Thanks Joel,

In terms of the expected swing month-on-month, is there any kind of ballpark figure to compare against? 14% seemed pretty high.

Thanks

Darroch