LoginSchedule demoLoginSchedule demoFirst-party data across 188 enterprise programs: clicks per share halved while engagements held flat. A click-based scoreboard is grading the platform, not your program.
Somebody is about to walk into a quarterly review and explain why their advocacy program is failing. They will have a chart. The chart will show clicks per share falling steadily for two years, and it will be accurate, and the conclusion they draw from it will be wrong.
I know because I have the same chart, across roughly 190 enterprise programs, and it says the same thing.
Clicks per employee share fell from about 4.2 in early 2024 to about 1.6 by mid-2026. The median program went from 2.9 to 1.2. That is a decline of roughly 60 percent, and it is not concentrated in weak programs, or new programs, or one industry. It is a smooth slope across the whole portfolio.
Here is the part that changes what it means. Over the exact same window, engagements per share did not move. It sat between roughly 3.7 and 4.3 the entire time and it is 3.7 today.
One number halved. The other did not budge. Those are not the fingerprints of programs getting worse. They are the fingerprints of a platform changing what it distributes.
Short and plain, because you should be able to check my work.
This is EveryoneSocial’s own portfolio, not a survey and not a vendor study. Around 190 enterprise programs on real paying plans, measured on a trailing twelve month basis, restricted to accounts with a genuine share base of at least 25 shares in the window so that a program with nine shares cannot swing the average.
If you have read our earlier benchmark table, the account counts will not match, and the difference is the screen rather than the portfolio. That table qualified accounts on registered users. This analysis qualifies them on at least 25 shares in the trailing window, which is the tighter test, because a workspace can be fully provisioned and still be barely sharing. That screen is why the number here is 188 accounts rather than the larger count you saw before.
Engagements per share is reactions and comments divided by shares. Clicks per share is outbound link clicks divided by shares. Neither is divided by impressions, which matters, because plenty of teams hear “click through rate” and assume the advertising definition. These are per-share averages.
No customer is identified anywhere in this, and no individual program’s numbers appear. It is aggregate only.
Three things happened, and none of them are about your content calendar.
LinkedIn suppresses posts that send people off LinkedIn. A post carrying an outbound link is distributed less widely than one that keeps the reader on the platform, and that gap was wider still in earlier years. Fewer impressions on the post means fewer clicks from the post. That is most of the story right there.
Company page organic reach has fallen sharply. Brands now reach a fraction of their own follower base without paying for it. That makes employee sharing more valuable, not less, but it lowers the ceiling on link-driven traffic across every channel a brand owns.
The ranking model changed. LinkedIn now rewards on-platform dwell time and real conversation, which means documents, carousels, video and text posts win, and those formats generate engagement rather than link-outs. The full mechanics of that ranking change are covered separately and I am not going to rebuild them here. The relevant consequence is narrow: the platform rebuilt itself to reward the thing that shows up in your engagement number and to penalize the thing that shows up in your clicks number.
Which means clicks per share is now, functionally, a measure of how much LinkedIn dislikes link posts. Grading your program on it is grading your program on a number the platform has publicly committed to shrinking.
I want to be specific about this, because it is the reason I trust the trend and distrust any single month.
When we first pulled this data, engagements per share appeared to fall off a cliff in one month of 2025 and then slowly climb back. It looked like a genuine engagement collapse across the portfolio. It was not.
The engagement numerator never moved. It sat at right around 5 million the whole time. What moved was the denominator: total shares in the trailing window nearly doubled in a single month because of how shares landed in the export, then normalized. Divide a stable number by a temporarily inflated one and you manufacture a crisis that never happened.
Then, recomputing this last week against a fresh export, the same thing happened again in the newest month. Engagements flat at 5.4 million, clicks flat at 2.3 million, and the share denominator up 25 percent in one month, with the median program’s share count jumping 1.4x. Both per-share ratios “dropped.” Neither actually did.
We caught it the second time because we had been burned the first time. That is not a story about our data quality being great. It is a story about why a single month is never evidence. A per-share metric has two moving parts and you cannot tell which one moved by looking at the ratio.
If you take one operational habit from this post, take that one. Before you report a per-share number as up or down, look at the numerator and the denominator separately. Most of the alarming quarter-over-quarter swings in advocacy reporting are denominators.
The fix is not a better click metric. It is a different scoreboard.
Set engagement-based expectations and put them in writing at launch. Across the portfolio, a healthy floor is 3 to 4 engagements per share and 1 to 2 clicks per share. Top quartile programs still clear 6 or more engagements and 2 or more clicks. Those are floors for expectation-setting, not promises, and per-share performance varies enormously by program maturity and content mix.
Report the two numbers separately, always. Blending them into one “per share performance” figure hides the entire story. One is steady and one is being deliberately compressed by the platform. Averaging a stable metric with a suppressed one produces a number that means nothing and trends downward forever.
Stop treating link post click-through as the health metric. It is a legitimate diagnostic for a specific campaign with a specific landing page. It is not a measure of whether your program is working, and the moment it becomes the headline, your advocates start optimizing for the format the algorithm punishes.
Coach for native value. The post should be worth reading without the click. Keep the link secondary. This feels like giving up traffic and it is the opposite: protecting reach is what eventually produces more total clicks, because a link on a post nobody sees converts at zero.
Reset your executive’s expectations with the trend, not the snapshot. An executive who was told in 2023 to expect 4 clicks per share is going to read 1.6 as failure, and no amount of context in the appendix will fix that. Show them the two-year slope. Show them that engagement held. Show them it happened to everyone.
Then move the real argument downstream. Clicks per share was always a proxy, and a weak one. What the business wants to know is whether advocacy produced traffic that converted, pipeline that closed, and candidates that applied, and that lives in attribution you configure once, not in a platform metric you have no control over. The full case for that scoreboard is in How to Measure Employee Advocacy ROI (Without Counting Shares).
Some programs genuinely are underperforming, and this post is not cover for them. If your shares are falling, your active sharer count is shrinking, or your engagements per share is well under 3, you have a real problem and it is not LinkedIn’s fault.
But if your share volume is steady, your engagement per share is in the 3 to 4 band, and the only thing sliding is clicks, your program is performing exactly in line with a portfolio of 190 enterprise programs and the thing that needs fixing is your reporting.
That is a much cheaper problem to solve than the one you thought you had. It is also the one nobody gets budget to fix, because “our metric is wrong” is a harder ask than “our program is broken.”
Make the ask anyway. The alternative is defending a number that is designed to go down.
Want to know where your program actually sits? Both numbers, read separately, against a portfolio of roughly 190 enterprise programs. Book a demo and we will run yours against the benchmark.
See the platform that comms, marketing, and compliance teams use to run their advocacy program.
