Adometry was an Austin, Texas company that sold cross-channel marketing attribution: software that took the record of every advertisement a person saw before buying something, then divided the credit for that purchase across those exposures. Google acquired it in May 2014, relaunched the technology as Google Attribution 360 in 2016, and eventually folded the surviving parts into products that no longer carry the name. The term still appears in vendor documentation and agency pages, denoting both a defunct company and a technical lineage running into the attribution reporting used now.

The problem it addressed was last-click accounting, under which the final touch before a purchase collects all the credit and every display impression, email and television spot preceding it collects none. Adometry belonged to a category of vendors, alongside Convertro, Visual IQ, ClearSaleing and C3 Metrics, selling fractional attribution: statistical models assigning partial credit to each touchpoint on a path.

From click fraud to attribution

The company did not begin in attribution. It began as Click Forensics, a click fraud auditor whose Click Fraud Indexwas for several years one of the few public numbers on invalid traffic in paid search. According to Click Forensics, the industry-wide rate hit 17.1 percent in the fourth quarter of 2008, up from 16.0 percent in the third and the highest reading since the index began in 2006. On content networks including Google AdSense it ran at 28.2 percent. The figures drew criticism: the company publishing them also sold detection services, and never fully disclosed its methodology.

Founding dates conflict. Co-founder Tom Cuthbert dates Click Forensics to 2006, TechCrunch and ClickZ to 2007, databases including Tracxn and CB Insights to 2005. Silicon Hills News reports it started in San Antonio on a $500,000 Austin Ventures seed round before relocating.

On 1 March 2011 Click Forensics bought a smaller company, Adometry, Inc. of Redmond, Washington, which sold display ad verification. Terms were undisclosed. The buyer took the target's name and launched an ad analytics suite the same morning. Paul Pellman, then chief executive, told AdExchanger the move expanded the product line rather than pivoting it.

Total venture funding is also disputed: roughly $44 million according to TechCrunch, $29.1 million according to Silicon Hills News citing the company, $30.6 million according to Tracxn. An $8 million Series D led by Shasta Ventures closed in December 2012.

How the attribution engine worked

The mechanics are documented in unusual detail because Adometry patented them. Application US20160034948A1, filed by Shi Zhong and Robert Lee Marsa, claims priority to a provisional dated 28 February 2013 and was assigned to Google on 30 October 2015. Its status is abandoned.

Collection ran on two tag types. Ad tags, described in the filing as 1x1 pixels embedded in publisher page code, recorded impressions including whether the creative sat above or below the fold and how long it stayed in view. Page tags on the landing page recorded arrivals and referrers. Each event row carried 23 fields, among them server timestamp in UTC, cookie ID, source IP, interaction type where 0 denotes an impression and 1 a click, campaign, placement, publisher and creative IDs, an iframe visibility flag, position on page, time on ad, time on page, demographic segments and revenue.

From those rows the platform assembled, for each converting user, the set of events preceding the conversion, then generated leave-one-out subsets. Credit came from conditional probabilities estimated by counting users. The filing evaluates four models and rejects three: naive Bayes and conversion index formulations miss interactions between events, while a conditional importance model breaks monotonicity by assigning a combined event less credit than its parts. The preferred marginal importance model normalises the probability of seeing an event given the others and a conversion against the probability of seeing it given the others alone, discounting touchpoints that would have appeared regardless.

Granularity created a sparsity problem. Defining an event as two impressions from one campaign with one creative on one site three days earlier produces cells too thin to estimate reliably, even from billions of monthly impressions. The system therefore ran the analysis at levels arranged in a hierarchy, from campaign alone up to campaign plus site plus frequency plus recency, combining results through hierarchical Bayesian shrinkage with a sigmoid confidence function weighted by user counts.

Channels with no user-level data were handled separately. Direct mail, television, radio and some social inventory yield no observation of people who saw an advertisement and did not convert. For those, the platform aggregated online data to match, sorted channels into funnel stages by conversion rate, splitting branded from non-branded search, then ran multi-stage least squares regression with instrumental variables to derive channel weights, which multiplied the user-level importance scores.

Where it sat in the transaction flow

Adometry sat in the measurement layer, not the buying layer. In December 2012 it released Attribute Catalyst Framework, which pushed fractional credit scores back into buying platforms as data feeds, initially through Advertising.com, DoubleClick Bid Manager, MediaMath and Videology. A 2013 partnership put the models behind Acxiom's Audience Operating System.

Google, and what happened next

Google announced the acquisition on 6 May 2014. Financial terms were never disclosed. Hours later AOL announced its intent to buy Convertro. Forrester analyst Tina Moffett noted the tension in both deals: the acquirers sell media and now also grade it.

That tension was already visible in vendor evaluations. According to AdExchanger, Forrester's Wave on attribution had rated Adometry and its independent peers highly while criticising Google's own tool as suited mainly to marketers already committed to Google Analytics, display and search. James Green, then chief executive of Magnetic, told MediaPost that Google's attribution offering "was never completely unbiased."

Rebuilding took nearly two years. On 15 March 2016 Google launched the Analytics 360 Suite, six products sold separately, in which Google Analytics Premium became Analytics 360 and Adometry became Google Attribution 360, described by Google as rebuilt from the ground up. It shipped three capabilities: digital attribution across user-level touchpoints, marketing mix modelling covering television, radio and print alongside factors such as seasonality, and TV attribution matching minute-by-minute airings data against search and site behaviour.

Babak Pahlavan, then senior director of product management, told TechCrunch the two years since the acquisition had gone into rebuilding the tool on Google's stack and simplifying it. Simplification eventually eliminated the product. On 23 May 2017, at Google Marketing Next, Google announced a free beta called Google Attribution, described by Search Engine Land as a stripped-down Attribution 360. Bill Kee, group product manager for attribution, told the keynote: "This year, we're solving the attribution problem."

No public sunset notice for Attribution 360 surfaced in research for this article. Bounteous, writing in June 2022, described the standalone product as scrapped, leaving traces in corners of the Google Marketing Platform stack. The modelling approach survived, embedded rather than sold. Data-driven attribution became the default in Google Ads and Google Analytics 4, and in April 2023 Google removed the first-click, linear, time decay and position-based models, stating that fewer than 3 percent of conversions used them.

Why the lineage matters

The acquisition marked the point at which independent multi-touch attribution began consolidating into platforms that also sell inventory. Two years later Google restricted user-level data export through Ads Data Hub, which AdExchanger characterised as narrowing the options for independent modelling. Google assembled a roster of external checkers instead, and by 2018 the Google Measurement Partners programme covered viewability, brand lift, app attribution and marketing mix modelling across more than 20 firms.

The technique Adometry sold has since been partly displaced by the aggregate method it used only as a fallback. Marketing mix modelling returned to favour as user-level identity degraded, and Meridian opened globally in January 2025.

Limitations and disputes

Three criticisms attached to the technology and never detached. The first is identity: path reconstruction depended on cookies and broke across devices and browsers. The second is the counterfactual gap conceded in the patent itself, which states that non-converting users on offline channels cannot be observed. The third is conflict of interest, raised at both 2014 acquisitions and unresolved since. Buying measurement from a media seller asks one party to mark its own work, a critique that now attaches to open-source platform tooling as much as it did to Attribution 360.

Disambiguation

Adometry, Inc. of Redmond was a display verification vendor founded around 2006 and acquired in March 2011 by Click Forensics, which adopted its name. References before that date usually mean the Redmond company, references after it the Austin one.

Attribution 360 was the paid enterprise product built from Adometry. Google Attribution was a separate free beta announced in 2017. Neither survives under those names.

Analytics 360 was the renamed Google Analytics Premium, a web analytics product. Attribution 360 sat alongside it in the same suite but did a different job, and the names are frequently conflated.

Convertro and Visual IQ were direct competitors. Convertro went to AOL the same day Adometry went to Google, and Visual IQ was later bought by Nielsen. The standalone category effectively ceased to exist.

Recent developments

At Google Marketing Live 2026 on 20 May, Meridian was integrated directly into Google Analytics 360, placing marketing mix modelling inside the same enterprise interface that once sold Attribution 360 as a separate line item. The same announcement introduced Qualified Future Conversions, a Gemini-powered metric linking current spend to predicted future sales.

Meridian gained a code-free Scenario Planner in February 2026, and Meridian GeoX and Meridian Studio were disclosed on 5 May 2026. The daily fractional credit scores Adometry once computed from 23-field event logs have given way to Bayesian inference over aggregated spend.

Timeline

  • 2005 to 2007: Click Forensics founded in San Antonio, Texas, with seed funding from Austin Ventures. Sources disagree on the year.
  • 2006: Click Fraud Index begins publishing quarterly click fraud estimates.
  • Q4 2008: Click Forensics reports a 17.1 percent industry click fraud rate, its highest reading.
  • 1 March 2011: Click Forensics acquires Adometry, Inc. of Redmond, adopts its name and launches an ad analytics suite.
  • January 2012: Ad Analytics Suite 2.0 released with Smart Overlap reporting and frequency analysis.
  • December 2012: Attribute Catalyst Framework pushes attribution credit into DSPs including MediaMath and Videology.
  • 28 December 2012: $8 million Series D led by Shasta Ventures closes.
  • 28 February 2013: Provisional patent filed covering fractional attribution using user-level and aggregate data.
  • September 2013: Attribution models integrated into Acxiom's Audience Operating System.
  • 6 May 2014: Google announces the acquisition of Adometry. Terms undisclosed. AOL announces its intent to acquire Convertro the same day.
  • 30 October 2015: Adometry patent rights assigned to Google Inc.
  • 15 March 2016: Google Analytics 360 Suite launches; Adometry becomes Google Attribution 360.
  • 23 May 2017: Google announces a free Attribution beta at Google Marketing Next.
  • July 2018: Google Measurement Partners programme consolidates third-party measurement partners.
  • June 2022: Trade analysis describes the standalone Attribution 360 product as scrapped.
  • April 2023: Google announces removal of four rules-based attribution models from Ads and Analytics.
  • March 2024: Meridian unveiled as an open-source marketing mix model.
  • 29 January 2025: Meridian opens globally.
  • 19 February 2026: Meridian Scenario Planner launches.
  • 20 May 2026: Meridian integrated into Google Analytics 360 at Google Marketing Live.

Summary

Who: Adometry was founded in Texas by a team including Tom Cuthbert and led at the time of its sale by chief executive Paul Pellman, and was backed by Austin Ventures, Sierra Ventures and Shasta Ventures. Google acquired it. Babak Pahlavan and Bill Kee led the Google product teams that absorbed and then replaced the technology.

What: A SaaS cross-channel attribution platform that reconstructed user paths from tag data, assigned fractional conversion credit using probabilistic models combined by hierarchical Bayesian shrinkage, and handled channels without user-level data through aggregate regression with instrumental variables.

When: Founded between 2005 and 2007 as Click Forensics, renamed Adometry in March 2011, acquired by Google in May 2014, relaunched as Attribution 360 in March 2016, and discontinued as a standalone product without a public sunset notice, with trade coverage confirming its removal by 2022.

Where: Headquartered in Austin, Texas, with satellite offices in Redmond, Washington, and London. Its technology now sits inside Google Marketing Platform and Google Analytics rather than in a separate application.

Why: The company existed to break the last-click convention that concentrated credit in the final touchpoint before a purchase. Its acquisition marked the absorption of independent multi-touch attribution into platforms that also sell media, an unresolved conflict that shapes how measurement is bought and disputed today.