508. How much can we trust LinkedIn's numbers?

The episode opens with a postbag question from LinkedIn trainer Ashley Leeds: if people block a connection request or message, does LinkedIn penalize the sender? John Espirian says there are no firm statistics confirming this, but he thinks it is reasonable to assume that sending “loads and loads” of invitations that are ignored could produce a negative signal. If many people reject invitations, LinkedIn may decide the sender resembles a spammer and be less likely to promote their content across the platform. John explains that recipients can ignore an invitation, leave it untouched, or sometimes select “I don’t know this person.” He suspects that the latter action, especially when repeated, could send a stronger negative signal. His advice is to target people who are likely to accept, warm up prospective connections by commenting on their posts, and include a personal invitation note rather than a sales-like introduction. Free accounts can personalize only three invitations per month, while Premium accounts have no such limit. Although the effect is unconfirmed, John concludes that a low acceptance rate probably does not bode well for a sender’s LinkedIn future.

John then turns to reports from listeners who have seen unusually low LinkedIn impression counts. In some cases, the impression number has been lower than the number of reactions and comments. He points out that this cannot make sense: an engagement can occur only after LinkedIn has shown the post to someone, so engagements cannot logically exceed the number of impressions. He has also noticed lower-than-usual impression figures himself, although the counts sometimes recover if users wait a couple of days. That delay makes him less confident in the numbers displayed in the dashboard. Even so, John argues that impressions are not necessarily the most useful measure. Comments are more valuable because they show that a real, identifiable person spent time responding, while direct-message conversations stimulated by a post may be even more useful for someone trying to generate business. LinkedIn does not show those conversations in its analytics dashboard. John says impression counts are mainly useful to people demonstrating reach to brands or advertisers. For an ordinary user seeking conversation and new business, he recommends prioritizing engagement and the private conversations content creates rather than treating the raw reach number as the main measure of success.

Microsoft’s quarterly financial results include LinkedIn figures because Microsoft owns the company. The headline number is 1.4 billion LinkedIn members, but John says listeners should take it “with a pinch of salt.” He believes this may count accounts created rather than unique people, and says the figure is likely inflated by fake, duplicate, and dormant accounts. He uses the United States figure as an example of why the numbers look questionable. LinkedIn reports 265 million US members, even though the total US population is about 350 million and roughly 18% of the population is under 16 and therefore unlikely to have a LinkedIn account. John therefore finds it “very sketchy” to treat 265 million as the number of genuine US users. The reported top five countries by membership are the United States with 265 million, India with 193 million, Brazil with 103 million, the UK with 48 million, and Indonesia with 41 million. John says he is tracking these figures over time on a membership-tracker page, while warning that the published totals cannot be independently verified from the information available to him.

The quarterly figures include several product and usage claims. LinkedIn says more than 10 million members use its AI-powered job search daily, while 25,000 customers use its agentic hiring solutions. Time spent viewing feed content is reportedly up 15% year over year, and LinkedIn Talent Solutions has grown 10%. John notes that greater time spent on the platform should eventually support more revenue, which explains why Microsoft and LinkedIn emphasize it. LinkedIn CEO Dan Shapiro is also quoted as reporting a 25% year-over-year increase in paid video ads, a 63% increase in ads featuring creators, and a 121% increase in Premium Company Pages. John is surprised by the video-ad growth but is not impressed by the creator-ad figure, since he considers it natural for advertisements to use creators. He also wants to see absolute numbers behind the Premium Company Pages percentage, because a large percentage increase from a small starting point may not mean much. John says his own trial of Premium Company Pages has not helped and considers the product expensive. He objects that verifying a company page, and removing fake employees from it, effectively requires paying for Premium Company Pages and verification—making a basic administrative feature feel gated.

LinkedIn’s press-room figures say that, every minute, 110 new people add a profile, 9,000 job applications are submitted, 2 million feed updates are viewed, 19,000 connections are made, and 119 hours of learning content are consumed. John identifies LinkedIn’s approximately $20 billion in annual revenue as the most consequential number for Microsoft. He compares that with Microsoft’s roughly $26 billion purchase of LinkedIn about ten years earlier and calls LinkedIn a “real cash cow.” To estimate how many people are genuinely active, John uses data LinkedIn disclosed about users in the European Union. His rough calculation suggests that around 30% of the accounts may represent active users. Applying that proportion to 1.4 billion accounts produces an estimated 420 million active users worldwide—but he stresses that this is an extrapolation, not an official figure. LinkedIn defines an active user as someone who logged in once during a month, so the number who post, comment, or send messages regularly is much smaller. John’s practical conclusion is encouraging: people who consistently post, comment, and message are part of a relatively small minority, not merely a tiny voice lost among 1.4 billion active participants. He urges listeners not to give up, while acknowledging that he cannot fully trust or verify LinkedIn’s published numbers.