Corvidae AI Press Pack

For journalists, editors and analysts. Everything here is on the record and may be quoted without further permission.

As Trusted By...

Corvidae AI Press Pack

For journalists, editors and analysts. Everything here is on the record and may be quoted without further permission. Figures are drawn from client-approved case studies. Patent details are public record and independently verifiable.

Last updated: September 2026 | corvidae.ai

Corvidae AI At a Glance
What is Corvidae AI?An enterprise-grade analytics and attribution platform which also automates improved ad performance.
What does Corvidae do?Corvidae increases the revenue generated by existing Google and Meta ad budgets, and gives transformationally accurate attribution across all marketing activity.
How does Corvidae work?Patented AI reconstructs customer journeys from raw first-party data, with no cookies or third party identifiers for unified attribution view, then feeds that data back into ad platforms' bidding algorithms.
Founded2008
BasedEdinburgh, Scotland
Parent companyQueryClick Ltd
Funding£5m, led by Archangels with the Scottish Investment Bank
PatentsGranted in the United States, Europe, the United Kingdom, Canada and Australia
Clients includeRS Group PLC, QVC, Menkind, Macmillan, Groupe SEB
Entry price£1,200 per month
Chris Liversidge: CEO & Founder @ Corvidae AI

We are very proud of our Corvidae platform, which is the result of over eight years of incubation, £5m in investment and a two year collaboration with AI researchers at the University of Edinburgh's School of Informatics, the largest informatics department in Europe.

Having taken it to market as the only cookieless analytics platform in the last year, we are excited to see the reaction and explosion of interest in effective, compliant marketing measurement.

Google may have backed away from removing third-party cookies in Chrome, but strict consent models had already broken measurement - there has never been a better time to join the AI revolution and embrace modern marketing analytics.

— Chris Liversidge: Founder & CEO @ Corvidae AI

Solving Advertiser Pain One Customer at a Time

The Corvidae Story in One Page

Every advertiser spending on Google and Meta is bidding on incomplete data.

These platforms can only see their own data. They cannot see what happened at the top of the marketing funnel - before someone arrived in front of their tracking pixel - and they cannot see the channels they do not own.

The tools sold to solve that problem - Adobe Analytics, LiveRamp, GA4, and increasingly expensive in-house analytics builds - all depend on data controlled by somebody else. Third-party cookies for some. Third-party entity maps for others. Platform-side modelling for the rest.

Consent rules have broken all of it. When an advertiser adopts a strict "reject all" cookie policy, cookie-based measurement does not degrade gracefully. It collapses. At RS Group PLC, a FTSE 250 distributor, Corvidae tracked £193m more revenue than Adobe Analytics across a four-month test of all their global properties.

The conventional response is to buy a better report. But a report is only as good as the underlying data it uses. And a report cannot correct a bid that has already been placed on a guess.

Corvidae takes a different route. Its patented AI reconstructs customer journeys from raw first-party eventstream data - no cookies, no persistent identifiers, no third-party dependency - and then feeds those reconstructed journeys straight back into the advertising platforms' own bidding algorithms via API, and into the client's own reporting stack.

The result is not better reporting. It is more revenue from the same budget, or the same revenue from less. At RS Group PLC, a four-month controlled pilot in Italy increased Paid Search revenue by 6% while reducing ad spend by 5%, and improved return on ad spend by 12%. It is now being rolled out globally.

The technology took eight years and £5m to build for enterprise clients. It is now available from £1,200 a month.

Clarity is Golden

What Corvidae is Not

  • Not an attribution dashboard. Corvidae produces attribution, but does so by completely rebuilding the way data is joined together. All attribution models benefit. Yes, there is a fully featured SAAS front end, and AI derived attribution models - plural - and an API for reporting automation and data feeds for further AI modelling. But, the most powerful use of Corvidae is automation of AdTech using the same stitched journeys with AdTech identifiers stitched in.
  • Not a cookie replacement. It does not swap the third-party cookie for another identifier. It uses no identifier at all.
  • Not a walled garden. Data is fed back into the client's own systems as well as the ad platforms. Clients own their reconstructed data. Corvidae is open.
  • Not another SAAS login to add to the list. Corvidae data goes where you are. Its API allows integration with existing visualisation layer tech like Tableau, PowerBI or Data Studio - or use in Frontier AI CLIs. At RS Group PLC, that is Power BI, alongside the client's own commercial data - and back into Google Stack itself.

Logos & Assets

Corvidae's logo is available for use as represented here:

Products and Pricing

Corvidae is available direct or via agency partners for preferential pricing options.

Corvidae is available direct or via agency partners for preferential pricing options.
ProductWhat it includesPrice
Ad OptimiserThe automation layer. Reconstructed journey values fed back to Google and Meta bidding via APIFrom £1,200 per month direct.
Corvidae EssentialAd Optimiser plus Corvidae AttributionOn application
Corvidae AttributionFull journey attribution with consultancy, custom reporting and strategic support, for enterpriseOn application

A Deeper Dive into Corvidae AI Technology

How the technology works

The non-technical explanation

When someone visits a website, they leave behind a trail of raw interaction data. Traditionally, tools stitch that trail into a journey using an identifier - a cookie, a login, a device ID - which acts like a name tag. Remove the name tag, as consent rules now routinely do, and the journey falls apart into disconnected fragments.

Corvidae does not look for a name tag. It reconstructs which fragments belong to the same journey by learning the patterns of how real sessions behave, using neural network models, and rebuilding the sequence predictively. The journey is reassembled without ever identifying the person.

That reconstructed journey is then converted into a value against each advertising click and sent back to the platform, so the bidding algorithm is optimising against what actually happened rather than what it could see.

The technical explanation

Corvidae ingests raw first-party eventstream data via a first-party pixel.

A data cleansing layer normalises it, a re-evaluation engine recovers session relationships within non-session data, and a path-building engine links interaction sessions into complete journeys. Long short-term memory neural network models handle the sequential reconstruction using attention layers.

Reconstructed multi-touch journey values are exported via API against GCLIDs for Google, and equivalent identifiers for Meta, and to the client's own data and BI layer.

No cookies and no persistent personal identifiers are used at any stage, which makes the architecture GDPR and CCPA compliant by design rather than by configuration.

Why Corvidae cannot simply be copied

Competing approaches replace the third-party cookie with a different deterministic signal - a logged-in identity, an entity map, a device graph. That inherits the same fragility: when the signal is absent, the journey breaks. Corvidae's approach uses no identifier, which is what the granted patent covers.

In addition to the patent, there is a significant technical moat to overcome to avoid over-fitting which requires extensive domain expertise and technical development to resolve over and above replicating the patented approach.

The patents

All independently verifiable. Held by parent company QueryClick Ltd.

Title:A System and Method of Reconstructing Browser Interaction from Session Data Having Incomplete Tracking Data

Corvidae Patents have been granted globally and in strategic regions.
United StatesUS 12,461,984, granted 4 November 2025
EuropeEP 3888325, granted 20 November 2024. Unitary effect registered 10 January 2025 across 18 EU member states, with separate validation in Ireland, Spain and the United Kingdom
OppositionThe European opposition period expired with no Notice of Opposition filed. The patent can no longer be challenged through central EPO opposition
OriginUK application 1819613.9, filed 30 November 2018
InternationalPCT/GB2019/053404, with national filings in Canada and Australia
Inventor of recordChristopher Liversidge
Unitary patent effectUnitary effect covers Austria, Belgium, Bulgaria, Denmark, Estonia, Finland, France, Germany, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Portugal, Romania, Slovenia and Sweden.

Customer Results: RS Group PLC

All figures quoted here are from client-approved case studies. The full RS Case Study is available here.

RS Group PLC Case Study Results with Corvidae AI
MetricResult
Revenue tracked versus Adobe Analytics£193m more, across all RS Group PLC properties worldwide over a four-month test
Customer journeys trackedOver 95% of the active customer base, under a strict "reject all" consent model, with no cookies
Italy Paid Search pilot, four monthsRevenue +6% on 5% lower ad spend
Return on ad spend+12%
Journey length versus platform reporting+46% UK, +68% EU
Channel overlap revealed58% of all sales touched a Paid Search ad; 72% of those also had an Organic Search interaction
StatusRolling out across global marketing operations

Further Case Studies: Gift Universe // Menkind

Menkind (Gift Universe)

Gift and gadget retailer. Read the full case study.

After building trust in Corvidae automation, Gift Universe increased their ad spend by 68%. Usually, this level of scaling leads to diminishing returns and rising acquisition costs.

But with Corvidae’s journeys feeding Google Ads’ algorithm, Google’s own AI targeted auctions earlier in the customer journeys, and avoided overspending in last touch auctions.

This automation drove massive volume: an increase of +143% in revenue, while lowering CPA by 11%.

Corvidae supercharged Menkind’s ability to compete in online ad auctions and revamp affiliates leading to an improvement of +45% in ROAS.

QVC Germany: Corvidae AI Impact

QVC DE

A multi-platform retailer. QVC stands for "Quality, Value, Convenience" and is a major televised home shopping network and digital commerce company. Recently extended a multi-year deal with Corvidae, full published case study here.

The Impact

  • Customer Acquisition Cost reduced by 46% for Google Ads activity by replacing cookie-based journeys with Corvidae conversion journeys
  • An increase in ROAS of 89%
  • CPC was reduced by 34%
  • Played a key role in new customer acquisition which has strengthened QVC Germany’s customer base

Press Enquiries

Media contact

Chris Liversidge, CEO and Founder, Corvidae AI

[email protected]

Interview requests, briefings, technical deep-dives and analyst enquiries all to the same address.

Who To Talk To

Spokespeople

Chris Liversidge - CEO and Founder. Founded QueryClick in 2008 and built Corvidae from it. Named inventor on Corvidae's granted patents, with two decades in performance marketing. Speaks on market structure and platform transparency, the economics of measurement, the case against third-party data dependency, and the company story.

Justin Matters - Data Science Lead. Leads the team responsible for the session reconstruction models. Speaks on how predictive session reconstruction works, why deterministic matching fails, and technical comparisons with competing methods.

Michael Hirons - Customer Success Lead. Leads client onboarding and ongoing performance. Speaks on implementation, integration with client stacks, what adoption looks like in practice, and time to measurable result.

For Direct Quotation

Quote bank

Pre-approved and free to use without further permission. All attributed to Chris Liversidge, CEO and Founder, Corvidae AI.

  • Every measurement tool on the market depends on data somebody else controls.
  • Your analytics can tell you what happened. It cannot change what happens next.
  • Google's algorithm is only as good as the data you feed it, and it cannot see the top of your funnel.
  • GA4 is free, which is exactly why nobody audits what it costs them.
  • We do not track people. We reconstruct sessions.
  • Everyone else replaced the cookie with a different identifier. We stopped using identifiers altogether.
  • Cookieless is not a compliance feature. It is the only architecture that still works.
  • More revenue from the ad budget you already have.
  • You do not have to believe our attribution. Look at the increment.
  • The industry spent a decade buying better reports. Nobody bought better bidding.

Questions Journalists Ask

An enterprise-grade analytics and attribution platform which also automates improved Ad performance.

Corvidae increases the revenue generated by existing Google and Meta ad budgets, and gives transformationally accurate attribution across all marketing activity.

Patented AI reconstructs customer journeys from raw first-party data, with no cookies or third party identifiers for unified attribution view, then feeds that data back into ad platforms' bidding algorithms.

Parent company QueryClick Limited, was founded in 2008.

Edinburgh, Scotland

QueryClick Limited

£5m, led by Archangels with the Scottish Investment Bank

Patents are Granted in the United States, Europe, the United Kingdom, Canada and Australia

Clients include: RS Group PLC, QVC, Menkind, Macmillan, Groupe SEB

£1,200 per month

Last-click is heavily biased towards direct visits.

Last non-direct click - as used by GA4 and Adobe - ignores the influence of all the preceding channels. First click rewards only the first touchpoint.

Only using rules based models means marketers can never understand the true value of their marketing mix. Without Multitouch Attribution (MTA) models it is impossible for marketers to understand the impact each channel - or ad campaign - has on total revenue and the customer journey.

Corvidae attributes revenue to the most granular level possible - the individual visit or impression, and then allows aggregation up to insightful segments.

Shapley and Markov models cannot offer a value to such a granular level, as they are statistical models.

Only by first stitching and applying an MTA model like Corvidae's LSTM model can you then apply further statistical modelling to avoid over-fitting and the failure of the underlying mathematics in MMM, Shapley or Markov approaches.

Corvidae AI FAQs

Latest Insights
AdCP is an attempt to allow natural language ad buying

Corvidae AI & The Agentic Future

Chris Liversidge: CEO & Founder @ Corvidae AI By Chris Liversidge · 20th November, 2025 · AI
Media buying is undergoing a profound transition today, as Agentic Operators are launching and presenting as viable challengers to incumbent Adtech, potentially upending the decade-long relationship between advertiser and audience platforms.
Our Accreditations & Partnerships

Contact Us

Newsletter Signup