When Google sunset Universal Analytics, the data went with it. A recent Rule 37(e) decision shows what happens when an IT employee tried to preserve litigation data without direction from counsel — and why the consequences went far beyond a sanctions motion.
By Kelly Twigger
Podcast | Transcript
This week’s Case of the Week examines a spoliation decision that every litigator needs on their radar — not just because of the Rule 37(e) analysis, but because of the source of ESI at the center of it: website traffic data stored in Google Analytics, a platform that millions of organizations use and that Google permanently shut down in 2024.
The decision, issued by United States Magistrate Judge Victoria Reznik of the Southern District of New York, addresses what happens when a party knows a data platform is being deprecated, attempts to preserve the data, but fails to preserve it at the level of granularity that matters — all because no attorney was involved in the preservation process.
The Case
PharmacyChecker.com is an organization that provides consumers with comparative drug pricing information from domestic and international pharmacies. It filed suit in August 2019 against the National Association of Boards of Pharmacy and related entities, alleging Sherman Act violations. The core allegation was that NABP conspired to push PharmacyChecker out of the marketplace by placing it on a “Not Recommended Sites” list and then reaching agreements with search engines like Google and Bing to incorporate that list into their search rankings. The result, according to PharmacyChecker, was a measurable drop in organic web traffic from those specific search engines.
Web traffic data — and specifically which search engines that traffic came from — was therefore central to PharmacyChecker’s theory of the case and its damages.
Google Analytics as a Source of ESI
For litigators who don’t work in the marketing or digital analytics space, Google Analytics is a free platform that millions of websites use to track visitor data: who is visiting the site, where those visitors are coming from, what pages they view, and how long they stay. Critically, the platform captures granular data — it can break down total traffic by the specific search engine source, showing how many visitors came from Google, how many from Bing, how many from direct links, and so on.
That source-by-source breakdown is exactly what was at issue here. During earlier phases of discovery, PharmacyChecker produced at least one report from Google’s Universal Analytics platform showing monthly organic traffic broken down by specific search engine, covering January 2015 through August 2021. Later productions included refreshed traffic data, but only at a high level — overall traffic volume without the search-engine-specific breakdown.
The Platform Sunset
In March 2022, PharmacyChecker learned that Google was replacing Universal Analytics with a new platform called Google Analytics 4. Google announced that any data on the old platform would be unavailable after July 1, 2024. The platform was being sunset, and the historical data was going with it.
One of PharmacyChecker’s IT employees — someone who was aware of the litigation and had helped generate earlier traffic data reports — began downloading data from Universal Analytics before the platform went offline. But the employee did this without direction from counsel. He downloaded what he believed was relevant based on his own understanding, aiming to capture a “comprehensive view of the overall traffic.” He preserved the aggregate numbers but did not download the granular, search-engine-specific data for every time period. For the 19-month window from September 2021 through May 2023, only overall traffic data was preserved. The source-by-source breakdown was never downloaded, and once Universal Analytics went offline, it could never be reproduced.
The Dispute
In April 2024, NABP served a request for production seeking all documents and ESI related to website traffic, lost traffic, and damages. PharmacyChecker refreshed its earlier reports but again produced only high-level data. A dispute arose, and after motion practice, the Court ordered PharmacyChecker to produce a report with traffic data broken down by specific search engine source. PharmacyChecker produced it — but with a 19-month gap. At the time of the Court’s order, PharmacyChecker’s counsel mistakenly believed the granular data for that period had been retained.
NABP moved for spoliation sanctions under Rule 37(e).
The Court’s Analysis
Magistrate Judge Reznik’s analysis proceeded under Rule 37(e)(1) — not (e)(2). That distinction is important. Subsection (e)(2) authorizes more severe sanctions like adverse inference instructions, dismissal, or default judgment, but requires a finding that the party acted with intent to deprive. Under (e)(1), the court can impose measures no greater than necessary to cure the prejudice caused by the loss of ESI. The entire decision — including the sanctions imposed — is governed by that proportionality standard.
The threshold elements were largely undisputed: the lost data was ESI, PharmacyChecker had a duty to preserve it, and it could not be restored. The contested issues were whether PharmacyChecker took reasonable steps to preserve the granular data and whether NABP was prejudiced by its loss.
Reasonable Steps. The Court framed the reasonable steps inquiry as a negligence analysis and found that PharmacyChecker failed. The reasoning rested on several facts: PharmacyChecker’s own complaint alleged lost traffic from Google and Bing specifically; it had already produced granular, search-engine-specific data during phase 1 discovery; and the data was central to its claims. Given those facts, it was reasonable to expect that search-engine-specific data would remain relevant throughout the litigation.
But the critical finding was about what happened — and what did not happen — after PharmacyChecker learned that Universal Analytics was going offline. The IT employee downloaded data without direction from counsel. No attorney assessed what data needed to be preserved in the context of the litigation. No one contacted opposing counsel to discuss the ramifications of losing access to the platform. And no one followed up to verify that the right data had been preserved. The Court found that these failures, viewed together, constituted negligence.
The Court also noted that the parties’ discussions about “refreshing” the traffic data lacked precision. PharmacyChecker’s counsel understood the term to cover only the high-level reports ordered by Magistrate Judge Davison earlier in the case. NABP’s counsel understood it to include every report previously produced. Magistrate Judge Reznik observed that much of the protracted motion practice could have been avoided with clearer communication before the database went offline.
Prejudice. The Court’s prejudice analysis is where this decision stands out. Magistrate Judge Reznik did not simply find prejudice or no prejudice — she calibrated the degree of it with a level of nuance that is not often seen in spoliation decisions.
The Court found that NABP was prejudiced because the granular data was relevant to PharmacyChecker’s claims and would have aided NABP’s defense. But it found the prejudice was mitigated by several factors: NABP could still cross-examine PharmacyChecker’s experts about the reliability of their conclusions; NABP received granular data spanning January 2015 to April 2025 with only the 19-month gap missing; and the evidence of confounding factors that might explain changes in traffic during the missing period largely existed outside the search traffic data itself.
The Court even observed that in some ways, the missing data gives NABP a stronger argument against PharmacyChecker’s damages than if the data had been produced, because NABP can now attack the certainty and reliability of any damages claim during that window.
Sanctions. Having found prejudice under Rule 37(e)(1), the Court turned to what sanctions were appropriate. It declined to preclude PharmacyChecker from presenting evidence on liability and damages during the 19-month period, finding that remedy too severe given eight and a half years of produced granular data showing Google consistently accounted for most organic search traffic. Instead, the Court ordered that PharmacyChecker must make clear that any Google and Bing traffic figures and related damages for the missing period are approximations or estimates, not precise measurements.
The Court granted NABP’s request to present evidence of the data loss at summary judgment and trial, and permitted NABP to seek a jury instruction on the missing data. It awarded attorneys’ fees and costs for the sanctions motion and the pre-motion letter, but denied fees for the earlier motion to compel, which would have been filed regardless of the spoliation.
Takeaways for Litigators
Counsel must supervise ESI preservation — not delegate and walk away. This is the central lesson. PharmacyChecker had the data. The traffic numbers documented that web traffic fell after it was blacklisted. The granular breakdown by search engine was already in the record. All counsel had to do was ensure it was preserved when the platform went away. Instead, an IT employee made preservation decisions without any attorney involvement, and the most critical data — the source-level detail — was lost.
The consequences are significant. PharmacyChecker now has to navigate presenting its case to a jury while disclosing that it lost evidence, caveating its damages as estimates, and putting a witness on the stand who is credible enough that the jury doesn’t punish them for the loss but not so apologetic that it undercuts confidence in the rest of their case. NABP, with the right lawyer, now has the opportunity to drive a truck through what would otherwise be an ironclad damages case. The failure impacts settlement leverage, increases litigation costs, and creates a narrative problem that didn’t need to exist. There is a direct line from counsel’s absence in the preservation process to a harder case to prove.
Identify third-party platform ESI early and preserve it deliberately. Google Analytics data, website traffic data — these are sources of ESI that may not come up in every case, but when a party’s claims or damages depend on web traffic, counsel needs to identify them early. And it’s not just Google Analytics. Any third-party platform that stores relevant data — marketing platforms, CRM systems, cloud-based SaaS tools — can be deprecated, sunset, or changed at any time. Counsel must ask: where does this data live, who controls it, and what happens if access goes away?
Be precise in meet and confer communications. Magistrate Judge Reznik explicitly flagged that the protracted motion practice could have been avoided with clearer communication. When discussing supplemental or refreshed productions with opposing counsel, identify exactly which reports, which fields, and which level of detail will be included. Put it in writing. When the data source is at risk of disappearing, over-communicate.
Confront preservation failures transparently. My reading of this opinion is that PharmacyChecker’s counsel realized the error and, rather than confronting it, argued that what they produced satisfied prior orders. They dug in. But the Court’s footnote suggests a different path was available: disclose the gap early, work with opposing counsel to agree on how to handle the missing data, and potentially stipulate to estimates or interpolation. That would have required a level of candor and maturity that we don’t often see in litigation. But look at the outcome — the sanctions imposed are measured and fair. PharmacyChecker can still make its case. Had they gotten ahead of the issue, they likely could have reached a similar result without the motion practice, the fees, or the credibility damage. How you respond to a preservation failure matters as much as the failure itself.
Use the prejudice analysis as a roadmap. Magistrate Judge Reznik’s calibrated approach to prejudice — finding it real but mitigated — is a model for how these arguments should be made. Whether you are moving for sanctions or defending against them, the prejudice argument under (e)(1) is where the outcome gets shaped. Show the court specifically how the absence of data harms your ability to litigate, or show the court what the other side can still do with the evidence that exists. Not every judge will conduct this analysis on their own. It is counsel’s job to guide the court to a calibrated finding, because the degree of prejudice controls the severity of the sanctions.
Conclusion
Every organization with a website is using some form of analytics to track traffic. Every organization using cloud-based tools, marketing platforms, CRM systems, or third-party SaaS products is storing potentially relevant ESI on infrastructure that someone else controls — and can change or shut down at any time. As litigators, we have to identify these sources early, involve ourselves in the preservation process, and follow up to make sure it was done right. If this case teaches us anything, it’s that the cost of getting it wrong isn’t just a sanctions motion. It’s the slow erosion of your ability to prove your case — or defend against one — when it matters most.
This decision is available on the Minerva26 platform with full issue tagging. If you’re a litigator navigating discovery strategy and want to stay ahead of decisions like this one, visit Minerva26.com to learn more or schedule a demo. Every decision covered on Case of the Week is searchable by issue, jurisdiction, and judge. If it’s about the discovery of ESI, it’s covered in Minerva26, your discovery strategy platform.
- 📩 Get the newsletter
- 🎧 Listen to Meet and Confer
- 🔁 Share this post on LinkedIn
