LADbible Group Onboards ArcSpan AMS in 28 Days and Generates 30%+ Increase in Audience Cohort Sizes

The Challenge

LADbible Group — publisher of LADbible, SPORTbible, GAMINGbible, Tyla and its wider network — reaches around 69 million users and 155 million sessions a month. Its entire commercial model depends on turning that scale into addressable, sellable audiences.

The group was running a legacy DMP with more than a thousand segments accumulated over years of campaign work, and an audience function that was, in practice, one person’s remit. Replatforming at that size is the project publishers postpone indefinitely: the segment library has to be rebuilt, every activation path re-plumbed, and any gap in coverage shows up directly in revenue.

So the bar was not “is ArcSpan better?” It was: can we move without breaking anything — and come out the other side with more sellable audience, not less?

The Solution

ArcSpan deployed the ArcTag across all LADbible properties with no custom scripting from LADbible’s developers, and it was collecting data from day one. Thanks to No-Code Onboarding and automated segment translation, ArcSpan did the heavy lifting allowing LADbible to operate their day to day as normal.

Inside 28 days, ArcSpan:

• Translated 400 cohorts into AMS in roughly two hours — exported straight from the previous DMP UI in its native query format, with no reformatting and no manual schema mapping.

• Widened the addressable footprint of every segment by leveraging ArcSpan’s built in LLM to classify and target keywords and IAB categories automatically, while preserving boolean conditions, nested rules and operators intact.

• Built LADbible’s priority audiences before the SOW start date, at no cost, so the trial opened with live, populated segments to evaluate rather than an empty platform and a backlog.

• Stood up 2 SSP activation paths, Index and Xandr, inside the trial window — one of which had consumed several months of integration work under the previous DMP. LADbible granted Xandr permissions one day and segments were flowing the next.

Results

Bigger audiences, and 30% increase in audience size. Because ArcSpan’s classification widens each cohort rather than simply recreating it, average segment size and audience cohort size rose by 30%, with some segments up to 2X their previous size. Every additional impression that falls inside a targetable audience is inventory LADbible can price and sell, so a 30% increase in audience size leads to a direct increase in revenue.

Cohorts delivered to GAM 5x faster than the previous platform. Audiences reach the ad server while intent is still fresh, so campaigns ramp faster, delivery is smoother and less of the addressable window is lost to latency.

Live in 28 days with nothing broken. Zero technical issues at ArcTag deployment; zero revenue-disrupting incidents at cutover.

A platform that keeps up with the business. New data sources and activation partners onboard in days, without engineering involvement.

Proven on Live BBC Campaigns

To launch a slate of drama titles, LADbible built Metropolitan Professionals in AMS from first-party signal collected by ArcSpan: urban renters in demanding, well-paid roles with sustained engagement across crime and drama content. The BBC ran it against four titles aimed at that viewer — Lord of the Flies, Small Prophets, AI Confidential and Murder the Case — with the incumbent DMP’s closest equivalent segment running alongside it on the same campaigns, inventory and dates.

Across the full partnership — 39 line items, three audiences and nine BBC titles — ArcSpan delivered 3.88 million impressions against the incumbent’s 1.70 million and an improvement in CTR, winning on scale in 37 of 39 line items. ArcSpan’s post-campaign report then closed the loop, verifying that the users actually exposed to the campaign matched the target cohort.

The campaign also left assets behind. Users who engaged with the creative were captured into a retargeting segment of 4,000, letting LADbible sell Havas a warm second flight with no additional tagging work. Overlap analysis on the cohort then surfaced a 5× index against betting and gaming interest — turning an audience built for a broadcaster brief into a repeatable product for an entirely different premium category.