Thomson Reuters has upgraded its legal AI tool CoCounsel with new capabilities, but the company faces a familiar marketing challenge: explaining technical improvements in ways that resonate with lawyers and law firms.
The "new and improved detergent" problem refers to the difficulty brands encounter when advancing products. Manufacturers know their formula works better, but consumers struggle to understand why the change matters. Thomson Reuters confronts this same dilemma with CoCounsel, its generative AI platform designed to assist lawyers with research, document review, and legal analysis.
CoCounsel operates within Thomson Reuters' Westlaw ecosystem, offering lawyers AI-powered assistance on tasks that typically consume billable hours. The tool uses large language models trained on legal content to generate research memos, summarize depositions, and identify contract terms. These functions address genuine workflow problems for practitioners managing heavy caseloads or extensive document sets.
Thomson Reuters' recent enhancements presumably improve CoCounsel's accuracy, speed, or breadth of legal knowledge. Yet articulating those gains presents a commercial hurdle. Legal professionals need concrete examples of how upgrades reduce their workload or improve results. Vague claims about "enhanced performance" or "better understanding" fail to justify switching costs or adoption efforts within firms.
The marketing obstacle reflects broader challenges in legal technology adoption. Lawyers remain skeptical of AI tools until they see measurable benefits. Features matter less than outcomes. A firm needs to understand whether upgraded CoCounsel will reduce the time spent on legal research by 20 percent or 40 percent, whether it catches contract provisions human reviewers miss, and whether it justifies the licensing fee.
Thomson Reuters must translate technical refinements into business value. That requires specific case studies showing time savings, clearer benchmarks comparing CoCounsel's performance against manual work, and transparent documentation of limitations. Without that specificity, even significant AI improvements risk