Naming a client beats quoting a percentage, by how much

Cold-email templates typically pick one kind of proof to include: a client name, a percentage improvement, a case study number, a timeframe claim, or general social proof language. This study compared six such segments drawn from 287,790 cold emails sent between 2026-01-26 and 2026-08-14, and two of them are worth comparing directly because they represent the two most common instincts, name-drop a real customer versus cite a statistic.

Emails that named a specific client, 10,958 of them in this sample, replied at 1%, with 99% getting no reply. Emails that instead quoted a percentage statistic, 79,293 of them, replied at 0.8%, with 99.2% getting no reply. The named-client segment out-replied the percentage-stat segment, and the percentage-stat segment was the lowest-performing of all six segments in the study, lower even than emails with no proof element at all, which replied at 1.3%.

It is worth being precise about the size of this gap and what it does and does not mean. Both segments are measured against the same overall study baseline, where 1.2% of all 287,790 emails replied and 98.8% did not, so both tactics are operating in a low-reply environment generally. The gap between named-client and percentage-stat performance is a difference measured in fractions of a percentage point on a base rate that is itself small, roughly one in a hundred emails. That is a real, directionally consistent difference across a combined sample of 10,958 plus 79,293 emails, but it is not a difference between a channel that works and one that does not. Neither tactic produces a high reply rate in absolute terms.

Why might naming a client outperform quoting a percentage? A named client is a specific, checkable claim, a recipient can in principle look up whether that company exists and works with the sender, and specificity of that kind often reads as more credible than a number, which can be sourced from anywhere or nowhere. A percentage stat, especially an unsourced one, is a claim a recipient cannot verify at a glance, and cold-email recipients who deal with high volumes of outreach may have learned to discount unverifiable numeric claims as a category, a subject line boasting a specific-sounding lift being a familiar and often-ignored pattern.

A competing explanation: senders willing to name a real client by name may on average be further along in building a case study or relationship, meaning the named-client segment could be capturing companies with generally stronger proof and messaging, not just a different proof format. This study segments by content type, not by sender maturity or company stage, so it cannot rule that out. A cleaner test would hold the rest of the email fixed and randomize only whether the proof element is a client name or a percentage, within the same sender and list.

For a team deciding what to put in an outreach template today, the practical takeaway is narrower than "never use numbers." It is that, in this dataset, a specific named client outperformed a percentage claim, and a percentage claim underperformed even having no proof element at all. That ordering is the finding. Whether it holds for a specific list, industry, or sender is a question this study's group-level spread, reply rates ranging from 0% to 4.5% across 33 groups, suggests is worth testing locally before assuming.