I think it would be too much of a risk, personally. You'd never know if the links you were to disavow were the ones that Google were specifically looking for.

Of course any effect would be delayed anyway, and you would probably not suffer from it, but I'd not take the chance.

This would be an interesting test. And I agree, a big risk.

If I was Google, I would have some kind of flag alerting me when the subsequent disavow file has dropped links reported in a previous one. But they may or may not have this alert.

If they do, there may be a manual penalty applied if they think you are trying to manipulate the ranking in this way.

If they don't, I would guess that when the second disavow file is processed then you *may* see maybe increase in ranking at first, and perhaps followed by fall in ranking when the Penguin is re-run, if you get caught by Penguin with cut-down disavow file.

If anyone has tested this, please share.

I doubt Google would penalize a website for undisavowing links because there are too many ignorant webmasters who get scared, overreact and disavow 100% of all links (good & bad).

I do think that undisavowing spammy links is not the smartest idea but I haven't tested it myself.

goodroi, you make a good point but Google looks for patterns. If you disavowed a bunch of sites Google rates as okay and later UNdisavowed them, I'm guessing Google would be fine with that (and may even have ignored it in the initial disavow request). But if you UNdisavow links from sites Google rates as spam, I would expect trouble.

I doubt Google would penalize a website for undisavowing links because there are too many ignorant webmasters who get scared, overreact and disavow 100% of all links (good & bad).

Thinking from the other end of the scope, this could well be the reason to suggest that Google won't really take the disavow data as seriously as many would have expected.

Thanks for all these replies everyone, really helping me out. I take it there has been no official statement regarding this from Google? I have been through an un-holy amount of videos and guides

Update: Matt Cutts tweeted, no, you should not remove the disavow file. [seroundtable.com...]

So what do you do if you make a mistake?

It's OK to overwrite it with more links, but not remove some links?

Thoughts?

Thats exactly what im thinking Whitey! I don't mean remove the entire disavow, of corse you don't want that. This is not a 'lets get around the system' post, purely what if you disavowed a link which was genuine, could you regain value from that link by taking it off the disavow file or does it mean it will never pass value again?

I guess the only safe way to use it is to be as sure as you can to only add spammy links to the disavow file.

After the event, how would you know which links were the good ones anyway?

What happens if disavow records are lost ( for whatever reason ) on the site-owners side ; an old webmaster won't co operate with a new webmaster ( for whatever reason ). How is the site owner protected?

I think WMT should hold a downloadable record of what's been submitted. Thoughts?

You are able to just download the current list from the Disavow page itself.

On the removing links from the file subject, there is absolutely no problem with removing a few links here and there if they are found to be natural. We have done so in the past and although I cant claim to demonstrate a rise in traffic resulting from the 'dis-dis-avowing' (TM applied :)) of the links as there was too much noise in all the cases we observed to record a clear cause and effect. What MC is referring to there is a direct question about removing the whole of the file, which would be a very silly thing to do.

Thanks Shai, this is the sort of answer I was looking for!

We submitted a disavaow file with some old spammy directory links in July - then added some article sites in August- Our site began to fall dramatically. In September we removed some of the less spammy article sites from the disavow file in an attempt to reverse the fall and we are still plummeting. Our search terms have fallen from 3245 in August to 536 now. So the whole exercise has proven disastrous. I wish we had never touched it but couldn't work out exactly hat was wrong with the site <snip>

[edited by: aakk9999 at 12:22 pm (utc) on Oct 28, 2013]
[edit reason] Please do not post URLs [/edit]

@niggle - how's your traffic? Google's keyword referrals are no longer shown in analytic reporting due to "not provided".

Hi Whitey

Pretty dire - we were getting 3000 uniques a day up to April 2011 - it all went downhill after thay - now around 110 - I was referring to the search terms report in Google webmasters which lists the keywords and phrases you are found for and the rough page rank - the terms have fallen 80% in 4 weeks !