2025 Gold AI

Agency: Merkle

Client: Compare the Market

Entry Name: Making AutoSergei real with Meer-tech

Executive Summary

Research shows ‘comparing’ is too much effort for a third of us. The brand was in a fight for convenience and needed to change - fast.

Strategy

Brand and CX effectively remained quite different: one inherently emotional, the other inherently functional. The team identified an opportunity - AutoSergei, more advertising conceit than customer experience at the time, could bridge the gap. He had both the pull of the Meerkat ,and the potential credibility and expertise to speak directly to customers.

Importantly, he would allow the brand to humanise the experience in a way other anonymous AI-powered services simply couldn’t; he could enrich the experience without becoming a distraction, making it more emotional while maintaining all the functionality.

Creativity

In making AutoSergei feel real the aim was to elevate his role from a ‘quoter’ to a more personal assistant that felt human. The first step was to reduce 576 rigid communication templates down to 165 AI-powered dynamic templates, using 28 centralised modules. In car insurance alone, the team went from 2 journeys/38 templates to 1 journey/14 templates, simplifying script from three steps and 750 lines of code, to 1 step and 4 lines of code.

The brand’s proprietary AI engine for insurance pricing automatically reached out to its panel of more than 200 partners to generate 120 million pricing combinations, all aimed at providing personalised one-to-one insurance savings for 25 million people across the UK.

Jacquard and MovableInk’s AI expertise dynamically created all copy, imagery and pricing in 25 million communications, adapting and optimising it in real-time across all email, app, SMS and web push activity using Natural Language Processing (NLP).

Results

AutoSergei was no longer a TV mascot, but the growth driver of the business. 120 million personalised AI-powered quotes, with 120 million personalised pricing recommendations (99.9% accuracy), increased CRM’s overall revenue contribution by 33% in the first three months.

This value continued to improve to a peak of 55% within nine months. All this was achieved with 25% fewer communications through more channels.

Advanced machine learning algorithms drove a 44% uplift in clicks in car renewals and a 31% higher conversion rate, generating an extra £2.5m in annualised revenue for car insurance alone.