Techpresident ran a good summary of how the Obama campaign in California used various web-based tools to connect offline with online — and get out the vote.
This is crucial stuff for those seeking to convince UK and other parties and politicians to invest more in online but we don't yet have real studies or much data on offline effects, i.e. how many extra votes, new voters, convinced late deciders or new organisers the online campaigns have generated over previous tools, such as direct-mail, or traditional shoe-leather methods. This will undoubtedly happen in the post mortems but the evidence already points to a real effect, particularly in generating momentum.
As I noted earlier, Obama's online edge obviously hasn't pushed him over the top but traditional negative campaigning has been seen to hold him up. Similarly, Ron Paul's massed online supporters couldn't translate that into votes, although they did work alongside a central operation which didn't properly harness that energy.
As web campaigning guru Patrick Ruffini puts it: "Ultimately, it’s all about fundamentals. If a candidate doesn’t have mainstream appeal and isn’t ready for prime-time, Internet activism isn’t going to make a difference." There are a lot of competing, complex factors in actual vote generation which will only be unpicked and properly analysed later.
Online to offline
The two main tools which Obama has used are:
- a social networking tool that helps self-organisation
- a tool which translates that into a get-out-the-vote operation
Social networking is centred on my.barackobama.com using a Blue State Digital toolset which progressive organizations like moveon.org have used and developed.
Get-out-the-vote for Obama used a - crucially - distributed deployment of the Voter Activation Network (VAN). It can generate what's called 'precinct walk and call sheets' as well as a virtual phonebank and lots more.
Republicans have used the
Voter Vault database from the late nineties and very effectively in the 2000 and
2004 Presidential, but in a closed and centralised way (the Tories also have this). This was enhanced by using commercially available data
right down to pizza toping choices to profile potential supporters.
VAN was deployed by the Obama campaign using volunteers who went through a
Camp Obama training session, which could either cover a weekend or ten minutes training via video and written tutorials. Much of the work in developing the tools was done by volunteers. What it utilised was in-group, including
Knitters for Obama, and local contacts contained in VAN.
Texas Obama Precinct Captains homepage
In California this led to:
- An initial 105,000 contacts statewide through the social networking tool
- Of those, more than 10,000 signed up to be precinct (local) captains
- 7,000 or so VAN login accounts were deployed
- Of those, four to five thousand became active
Those people then managed more than 60% of the average 100,000 contact attempts per day, 40% came from more traditional phonebanks. In the whole state more than 10% of eventual Obama voters had been contacted by a neighbour — which is unprecedented.
Campaign geek volunteers also developed a predictive analytic model using live feedback from those calls to refocus pro-actively on more specific demographic pockets.
The Clinton campaign in California focused on absentee, early voters using the Catalist toolbox.
Other impacts
What the better use of volunteer energy in California meant for the Obama campaign was more central resource spend available for the other Primaries on Super Tuesday and the following eleven straight wins. Plus the bottom-up drive led to Obama's one million donors — by some counts already as many individual donors than the entire Bush campaign of 2000, and almost as much as Kerry 2004.
Some of the issues which remain unresolved or in early development are:
- scaling up and strengthening capacity
- can these volunteer networks be used to repel smears?
- earlier use to target mail-in voters (as Clinton has done)
- retaining and using volunteer energy post-campaign
- what will the diversification of fundraising away from big money interests mean?
- what will happen when all campaigns are using the same toolboxes and tactics?
Marketing missesWhat's been missing thus far is much use of online marketing energy and tools, in particular
geo-targeting. This is the method of determining the physical location of a website visitor and delivering different content to that visitor based on location, such as country, region/state, city, postcode, organisation, ISP or other criteria.
Email marketing
still dominates expenditures (62%) with display, search and video ads just 11% of online budgets (the balance is for web development, with many more local candidate sites). Only $78m is expected for total online spend by all candidates in the primary campaign. Which could be less than the Clinton campaign spends
on doughnuts and pizza runs.
One reason why online ad spends are so low is that
consultants cannot make as much money from it.
What web advertising there has been has actually been dominated by
McCain and
Romney with the
Obama campaign spending very little, although they did run "Have you tried the convenience of early voting? Find your early-vote location" ads in Texas. These drove users to landing pages featuring a video message from Obama but apparently didn't work so well on the day for potential volunteers. The Clinton campaign has run virtually no web ads.
All these tools and techniques can also be employed elsewhere.
In the UK internet use
is already by a majority, is growing over other media use and is only going one way - up. I would imagine that the Tories are ahead of the game on this (my impression, though I'm advised it may well be the Libdems - it's definitely
not Labour) but once the real facts have been unpacked it would be a huge mistake for the other parties to just think 'fundraising' and not recognise that - as well as having a compelling candidate - running from the bottom-up, empowering supporters and making use of the Web's power is really what's behind Obama's success.