Sam Coates’ Post

Today marks the UK General Election 🗳️ But what can the shifts in consumer search behaviour tell us about what voters—and indeed consumers—have on their minds as they head to the polls? And what can brands learn from this? We’ve delved into our onsite search data to uncover the most significant trends observed in the run up to this election. Read Captify®'s newly released trends piece to find out more: https://lnkd.in/eHwe5-uA #consumerinsights #searchtrends #searchintelligence #shifthappens

Yesterday we released a piece of analysis (https://bit.ly/3W7wNGC) looking at consumer search, and how effective it could be at understanding and predicting the shifts in behaviour for this general election. We also took a punt at predicting the results - so how did we do? With two seats left to declare we can see that the Tories are currently down 67%, whereas the number of Labour seats has gone up by 105%. Yesterday, the voting data I examined showed a predicted Tory loss of 55%, and a Labour gain of 92%. Not too shabby. What accounted for the difference? A couple of thoughts: - I analysed data across the period from the election announcement, through to polling day. Looking at our timeline of search interest there was a last minute spike for Labour, which meant that on the day Labour performed a little better than expected. - While I could see the huge growth in tactical voting search (spiking 1200% after the release of Best For Britain's tactical voting guide) I was unable to see the outcome of this tool usage. I suspect tactical voting was one of the main drivers behind the Lib Dem's success, and explains why the Tory loss was greater than I predicted. Long story short - yet another striking example of the power of search data in understanding shifts in voter (and consumer) behaviour. If you'd like to know more about yesterday's work - or better yet, how we can turn these techniques towards understanding your audience and solving brand problems in a different context - drop us a note. Cheers, Sam

I'm intrigued by the recent election and the inaccuracy of predictions made by various polls. Consumer data analysis has always fascinated me. The discrepancy between what consumers claim they will do and their actual behaviour is stark. The final polls from Survation, YouGov, and IPSOS projected Labour to secure around 38% of the vote, but they only obtained 34%. Similarly, the Conservatives were predicted to receive 20%, yet they achieved 24%. These significant deviations highlight the challenge of accurately predicting outcomes. In an age driven by data, the inaccuracy of forecasting is thought-provoking. I'd guess that low turn out, constantly shifting consumer opinions and a tendency to provide socially desirable responses to pollsters may have contributed to these discrepancies. It's important for people involved predicting how people will behave, we need to understand that every forecast we make will probably be wrong! Danny Finkelstein's analysis in The Times puts this argument far more eloquently: [Link to the article](https://lnkd.in/dbU9t4d4)

Last night really highlighted the accuracy of YouGov's data! - We accurately predicted 92% of the seats correctly.  - Every party was within our bounds on both share of seats and share of votes. All of this is possible due to the unique relationship we have with our panel, which is the core of our business! #YouGov #MarketResearch #GeneralElection #Poll #Insights

Curious about the connection between elections and consumer spending? This new article from my APCO Insight colleague James Hazzard explains and offers advice to businesses on how to engage https://lnkd.in/eGxxDXtC #elections #consumerspending

In summation - incumbents are on notice - on all sides of the aisle. The numbers for younger voters, especially men, are a wake up call. it’s another reason to evaluate whether it’s the message, how we talk, or both. Voters want something different, and they want politicians to listen to them. Whether it’s the USA, United Kingdom,Japan or Romania, voters want governments to talk about the issues of voter, not policy elites. Parties and elected officials who care about trust in democracy, and want to move our values forward, need to become more sensitive to where voters are, and that scrutiny is higher than ever. For strategists, that means getting out of our silos and helping our stakeholders, elected officials, and partners talk to voters more broadly, not just those who will agree with us. Read more below, especially about the big trends for how younger voters are driving the conversation about jobs, cost of living and democracy. https://lnkd.in/dqzxPrbd

In collaboration with MediaVision, we used the Metis platform to predict the UK General Election results. Read up on our findings and what, if we're correct, this could mean for the world of Search. #SearchData #ElectionResults #ElectionPredictions #UKGeneralElection #UKElection

📈 What is MRP? MRP (multilevel regression with poststratification) is a method innovated in the late 1990s to produce public opinion estimates for small groups in the population. These could be hard-to-reach audiences that are too difficult to capture through traditional survey methods, or sub-regions like states or UK constituencies. 🕹 Famously, academics in the US used survey data of Xbox users to predict the outcome of the 2012 US presidential election. Focaldata has been at the forefront of deploying MRP to predict election outcomes. We were one of the first research companies to call the 2019 UK General Election, and the only firm to model down to electoral district level for Australia's The Voice Referendum. With lots of public MRPs forecasts already released, we've written a primer on what to watch out for... Read it below 👇

While I think we all know who will win the #UKElection, MediaVision used a combination of Metis data, YouGov data and machine learning to predict exactly how much each party will get in terms of share of vote. The recent rise of search demand around Reform is huge and while some will be about more information about a new party does this show the "Trump effect" where people are more likely to search something and tell the polls something else? Would love your perspective on it all. Its another way we are using consumer search data to predict outcomes. #opportunityeverywhere #ukelection #generalelection

Predicting who will win the UK General Election, and by how much, using search data. 👀 The things we Google are a far more honest reflection of our inner thoughts than the answers we give on polls and surveys - and this is especially interesting when it comes to politics. We decided to put this to the test, using a combo of Metis data, YouGov data and machine learning to predict each party's share of vote in next week's election. ✌ #UKelection #labour #politics #reformuk #conservatives #google #searchdata #seo #prediction #generalelection2024