I switched my US hosting for my 100% UK targeted .com to the UK and predictably saw a huge improvement overnight. (I put too much confidence in webmastertools geotargeting) e.g. SERP for a competitive 2 word keyword leaped from mid 2nd page to 5 for 'the web' and 4 for 'pages from the UK' searches
More surprisingly, SERP for the most competitive (1 word) keyword jumped from +100 to top 5 in 'the web' and to 2 in 'pages from the UK'. then next day, the keyword disappears entirly from 'the web' search, but it remains steady at 2 in 'pages from the UK'. All other keywords held steady. I made onsite changes in the mean time, but the suddenness of the lift makes me think that the drop of this single term from 'web' only is not a 100% onsite issue.
which brings me on to my main question, with such a massive improvement, and instability in 'the web' rankings, should i migrate the whole site from the com to the co.uk version? (it currently redirects to the com) or do i leave things to settle first. or does it make no difference having a com now i am hosted in the UK?
If i do switch, do I need to actually move the files to the co.uk in any sense, or will one line of 301 do the trick?
'Test by migrating a single directory to the co.uk', says Cutts, but i'm not sure i understand what this means. Do I move just a single page and monitor a single term? Can a singel page exist on a different domain in isolation like this for purpose of testing? And how can i isolate the effect of the co.uk with terms flying up and down all over the place?
all advice appreciated, big and small.