Performance Marketing & Paid Search Restructure
LNR - Rebuilding Paid Search for the AI Era
How Tribera rebuilt LNR's paid search account from the ground up using semantic AI, taking 150+ campaigns down to 64, growing revenue and lifting ROAS from 6.4 to 8.0.
BACKGROUND
LNR connects the West Midlands, the North West and London. Like every rail operator, it's not just competing with other train companies, it's competing with cars, buses and coaches, and with affordability now one of the biggest factors in how people choose to travel. By 2025, LNR's paid search account hadn't kept pace with how passengers actually search, plan and book. We were brought in to rebuild it from the ground up, giving LNR a platform that could grow into FY25/26 and beyond.
CLIENT CHALLENGE
Four things were holding the account back. Spend was over-indexed on brand and uncontested commuter keywords while under-invested in routes where market share could genuinely be won. 150+ campaigns and 1,200+ ad groups made strategic precision impossible. Acquisition was passive, built around passengers already committed to rail rather than the much bigger modal-shift opportunity. And the landing pages hadn't been touched since 2019, undermining paid investment on what's now a mobile-first booking journey. All of this on a live account already performing well, so transformation had to happen without disrupting existing revenue.
TRIBERA'S SOLUTION
Before changing anything, we tested our assumptions. We paused brand bidding for a controlled period and found the leisure revenue impact was minimal. Most people searching LNR by name weren't booking, they were trying to claim Delay Repay. That single insight gave us the confidence to rebuild.
We consolidated the account into a focused three-pillar architecture built around how passengers actually search. Cluster campaigns grouped smaller stations thematically so low-volume routes could access keyword inventory they couldn't sustain alone. Hub campaigns targeted LNR's biggest interchanges (Euston, Birmingham New Street, Liverpool Lime Street), capturing high-volume generic demand. Route campaigns targeted point-to-point searches at the exact moment someone was ready to book.
Underneath that, we ran thousands of historical search queries through a semantic AI model to group them by what passengers were actually trying to do, not just the words they typed. That surfaced spend going on informational queries Google's AI Overviews now answer directly, with no click and no booking. We cut it and redirected that spend into commercial intent.
Alongside the account rebuild, we worked with a specialist web partner on the landing pages. The biggest win was the simplest: replacing a five-tap booking widget with a one-tap version, lifting booking engine handoffs by 60%. The whole restructure was phased over three months so Smart Bidding had time to learn at each stage before the next went live. Performance never dropped.