I am not, by any real measure, a psephologist. However, it seems clear that the phenomona of the "Conference Bounce" in the UK is, these days at least, a general universal for the three main political parties. You have a media managed conference, and you see your poll ratings go up. It's unlikely, in today's world of PR that you will see your vote go down, even if you do engage in a bit of mild blood letting, especially if the monopoly broadcaster in the country is largely on your side.
A quick look at the polling data over September shows the effect. The Lib Dems had their conference and perked up. Now Labour have had their conference and it's neck and neck between them and the Tories on 36%. This time next week there will be another poll and it will show the Tory lead regained as the "conference bounce" takes it turn on them.
There's been much said in the past week about the Tory poll lead. I recall someone from Labour saying on Radio 4 that a 8-10 point lead is nothing like the lead Labour had over the Tories in the early 1990s. The argument goes that the polls are meaningless and "the Tories aren't doing as good as we did anyway". Of course, the Labour has, on election day, consistently polled less than the polls say they will, whilst the Tories have done the reverse.
Like I said at the beginning, I am no psephologist, but todays poll is not a particular bad one for the Tories or a good one for Labour. After all, during weekes where you effectively have guaranteed media coverage across the board everyday all day, you'd have to do something utterly silly to not get a bounce in the opinion polls. Arguably carrying out a poll during Conference season is pointless as it does not show the true popularity of any of the parties, it's more a snapshot of respondents short-term memory.