Sources reveal how UK political parties spent ~£3.2M on online ads during 2017 general election, including efforts to coordinate ad targeting on- and off-line
Jim Waterson / BuzzFeed : Tweets: @ruskin147 , @g_price , @jsavagetweets , @stevenkellow , @josephmdurso , @andydangerfield , @wallaceme , @kadhimshubber , and @jimwaterson Tweets: Rory Cellan-Jones / @ruskin147 : “We'd find that you as an individual were using this Mac and that phone and we were targeting those devices” Really? http://www.buzzfeed.com/... Gareth Price / @g_price : Interesting justification for otherwise barmy FB like buying: give friends the impression they endorse the message https://www.buzzfeed.com/... Jonathan Savage / @jsavagetweets : Labour spent £500,000 on Facebook ads, just £20,000 on Twitter https://www.buzzfeed.com/... Steven Kellow / @stevenkellow : Interesting stuff, political campaigning rapidly heading online here in the UK https://www.buzzfeed.com/... Joey D'Urso / @josephmdurso : Startling detail on Labour's online campaign operation. Under the radar, like the Tories in 2015. via @jimwaterson. https://www.buzzfeed.com/... pic.twitter.com/QV3tyYscAX Andy Dangerfield / @andydangerfield : Here's how Labour targeted voters on Facebook and unleashed a Snapchat ad blitz that was seen by 8m people in a day https://www.buzzfeed.com/... pic.twitter.com/XbT4qZm2TJ Mark Wallace / @wallaceme : We @ConHome warned in 2015 that CCHQ had to build on its microtargeting success fast as Labour would learn and copy https://www.buzzfeed.com/... Kadhim / @kadhimshubber : Wait, how specific does Facebook's targeted advertising get? https://www.buzzfeed.com/... pic.twitter.com/Xu1XRxzMYB Jim Waterson / @jimwaterson : Excl: How Labour spent £1.2m targeting voters online during the election - while media worried about Tory dark ads. https://www.buzzfeed.com/...
Context & Ripple Effects
BuzzFeed's sourcing opens up the 2017 UK digital ground war: roughly £3.2M spread across parties, with Labour alone putting £500K into Facebook, £20K into Twitter and running a Snapchat blitz inside ~£1.2M of online spend — and targeting run at the level of 'this Mac and that phone', deliberately synced with offline campaigning.
The toolkit on display here is the same one later reporting documented at scale in US politics: the Trump campaign's Custom Audiences and Lookalike Audiences playbook, the divisive 2016 Facebook ad ecosystem, and eventually a $20M, 218K-ad conservative-targeting operation. The UK figures are an early, well-sourced data point in that same arc.
First-order effects
- Labour and the Tories make platform ad tools — Facebook, Snapchat, Twitter — core election infrastructure, buying device-level targeting fused with offline voter contact rather than treating digital as a broadcast supplement.
- The disclosure shows parties were purchasing cross-device identity matching during the 2017 campaign before any UK rule required them to disclose how that money was targeted.
Second-order effects
- Rival campaigns and third-party groups get a proven template: the same audience-matching tools that let the Trump campaign scale its ad purchasing power become cheap enough for any funded actor to replicate.
- Platforms inherit a transparency problem they did not price in — a gap later exposed when reporters got Facebook to approve fake ads purporting to come from all 100 US senators.
Third-order effects
- If the pattern holds, election finance regimes built around broadcast spend caps give way to fights over data-driven microtargeting — the trajectory that ends in hundred-thousand-ad, multi-million-dollar operations like the Trump campaign's 2019 effort.
- The structural question shifts from how much parties spend online to who verifies political ad buyers and whether on/offline targeting coordination is disclosable at all — a regulatory gap every major democracy now has to close.
The trend: Political campaigning is consolidating around platform microtargeting that fuses online ad buys with offline voter files, consistently running ahead of disclosure and verification rules.