Our process was developed, and is conducted, by litigation consultants who understand both the nuanced issues involved in complex commercial legal disputes and the extraordinary level of rigor demanded of evidence relied upon by expert witnesses. Our experienced analysts are front and center at every step of the process to thoughtfully transform raw data into defensible evidence for experts to rely on – evidence that is considered probative by judges and compelling by juries. From the outset of each project, we work collaboratively with lawyers and/or experts to determine the scope, key questions and realistic objectives for the analysis.
Our process was developed by litigation consultants
OUR APPROACH IS BASED ON THREE CRITERIA:
THE VOLUBLE PROCESS
Experienced Litigation Experts Guide Every Step of Our Process
(click on image to view)
OVERVIEW OF OUR METHODOLOGY
Extensive Background Research
In advance of any data collection, we conduct extensive background research on the relevant products, companies, and competitors. This background research informs how we develop queries that create the dataset for the case.
Carefully Constructed Queries
Our queries are carefully constructed to capture as much of the relevant social media conversation as possible while filtering out irrelevant posts.
We have the ability to search major social media sitesThe conversation occurring across these major social media sites is representative of an extraordinarily large portion of the population. Facebook alone has more than 1.8 billion active users, and the 10 largest social media sites collectively have more than 5 billion users. All told, 85% of U.S. consumers currently use at least one social platform. such as Facebook, Twitter, Tumblr, and YouTube, along with forums, blogs and product reviews. We customize our searches based on the needs of the case and can scrape specialized sites relevant to the case and also use Facebook’s, Twitter’s and other sites’ APIs (Application Program Interfaces) to collect additional relevant information.
To ensure we have captured the optimal dataset for our analysis, we continually refine our queries (adjusting the filters and Boolean strings) as we review posts and understand what consumers are saying, as well as where and how they are saying it.
Additionally, using custom algorithms and metrics that we refine for each project, we clean the total universe of potentially relevant posts, filtering out additional irrelevant and unauthentic posts.
Integrity of Data
Voluble has licensed software to protect the integrity of its data and ensure it is documented in accordance with evidentiary rules (F.R.E. 901(b)(9) and other standards of authentication). Any online evidence cited in our reports is retained by a third-party firm in order to ensure a clean digital chain of custody and the preservation of authentic records. Any associated metadata is retained in the same manner. Affidavits attesting to the integrity of the data and the digital chain of custody are available upon request.
Using the scrubbed dataset we have constructed, we are able to conduct a wide range of analyses depending on the specific questions at issue. For example, we can perform quantitative analyses to:
- Determine how often consumers talk about an at-issue trademark or trade dress element
- Compare the volume of online conversations across multiple brands and/or at various dates
- Assess the extent to which consumers mention the plaintiff’s and defendant’s brands together in a single post.
We can perform qualitative analyses of the relevant social media posts, applying various filters to identify posts that would make compelling anecdotal evidenceThere is a huge number of fake accounts and bots, both active and dormant, on various social platforms. Bots can artificially inflate the following and support of certain products, companies, or people, while smearing competitors with fake news reports. Voluble has developed tools that enable us to identify bot generated posts. These tools rely on subtle differences between human and bot behavior, such as the higher volume of activity, the excessive use of certain special characters, the unoriginal content, and the use of external links that characterize posts by bots. to support our quantitative analyses or the work of others, such as survey or marketing experts.
Depending on the volume of posts and the available budget for the project, we can perform these analyses on the entire dataset of relevant posts or on a statistically appropriate and defensible sample of posts.