Controversial Conservative peer Michael Ashcroft has done campaigners in all parties a service with the frankness of his book on the 2010 general election, Minority Verdict.

Though a short book, it contains some details of how the Conservative Party went about targeting swing voters in swing seats for last year’s contest. Aside from the subject’s inherent interest, this particularly caught my eye because part of what he recounts is the Conservative equivalent of what I was doing when working for the Liberal Democrats for much of the last Parliament.

The Conservatives took Mosaic data and then ran a detailed and expensive polling operation to work out the typical levels of party support in each Mosaic category as well as how likely people were to switch their support. Up to that point it sounded very similar to the work I did for the party – save that the Tory data set was less detailed and more expensive to produce than the data I compiled via other means.

A general election is not, however, a beauty contest between spreadsheets and it is what the Conservatives did with their data which provides lessons for our future campaigning – and where they met and overtook the Liberal Democrats.

The Conservative party key seat operation, heavily funded and influenced by Ashcroft, made skilful use of the data benefitting from two advantages the Liberal Democrats did not have: good integration in both its computer systems and between its policy and campaigning teams.

Having a powerful, widely-used and easy to update national database (of the sort that political campaigns outside the UK have now had for a decade or more, both Labour and the Tories had for 2010 and the Liberal Democrats are now intending to get) allowed swift, effective analysis to power an effective large scale direct mail campaign. Central efficiencies of scale were married up with large amounts of local variation as tens of millions of target letters were produced.

Having closely integrated policy and campaigning teams meant policy plans (and analysis of opponents’s plans) was boiled down to knowledge about who would benefit or lose the most, and then linked to data about where such people were most likely to live in different target seats across the country. For example Ashcroft writes,

When the 10p tax band was abolished in the 2008 Budget, we used Mosaic to identify the people most likely to be directly affected, and who fell within one of our target tiers. Battleground Directors pinpointed the polling districts where these people were likely to be found, and within a matter of days volunteers were delivery half a million leaflets explaining the effect of the Chancellor’s decision and setting out Conservative plans.

The Conservative Party was so focused on getting its campaigning messages varying effectively across different groups of people that its target seats were grouped by demography rather than by geography – e.g. New Towns or Seaside Towns rather than East Midlands or South Central. Each cluster than had its own dedicated staff person who, because all their seats were of a similar demographic type, could really get under the skin of those sorts of areas and tailor campaigning to suit them.

Even if not driven by the example of Ashcroft and his target seat operation, clearly many in the party are rightly drawing similar lessons for the future as those you get from a close reading of Ashcroft’s book gives. The locating of the party’s press, policy and campaigns staff at federal HQ in the same open plan office, for example, is an excellent move.

What Ashcroft’s book highlights is how important it is that the party learns from these lessons and continues to adapt its approaches. And I would make the book required reading for anyone intending to be involved in running an election campaign in the next few years.

Michael Ashcroft’s book can be purchased from Amazon at http://amzn.to/MichaelAshcroft.

A different version of this piece first appeared in Liberal Democrat News, the party’s newspaper. You can get a subscription to Liberal Democrat News here.