Measurement guide
How to measure the ROI of an employee advocacy program
Four KPIs with the arithmetic behind each, the baseline you need before anything starts, and the two claims that get a program report taken apart in the first budget meeting.
What a program report can honestly claim
Advocacy reports fail in one of two directions. Either they claim revenue the program cannot trace, or they retreat into impression totals that nobody in the room trusts. Both get taken apart in the first budget meeting.
There is a defensible middle. You measure participation, development and reach against a baseline you set yourself, you name what you are not claiming, and you report the same four numbers every quarter so the trend does the arguing.
The baseline, and why it has to exist before the kickoff
A program without a baseline can only be reported in absolutes, and absolutes invite the wrong question. 340,000 impressions last quarter sounds like a lot until somebody asks what it was before. Record five things per participant on the day they join, before any training and before the first post.
Do it per person, not as a program total. Program totals hide the only movement that matters, which is people crossing from inactive to active.
If the program is already running, today becomes the baseline and you say so in the report. A baseline set in month seven is worth more than a reconstructed one, because reconstructed numbers get questioned and the questioning eats the meeting.
Four KPIs, and how to calculate each
1. Participation rate
Active participants divided by enrolled participants, per quarter. Define active before you measure it, in writing, and then do not change the definition. A workable threshold is three posts in the quarter. The definition matters more than the threshold, because a definition that moves makes every trend meaningless. This is the KPI that decides whether the program exists. A program with 40 enrolled and 6 active is a program with 6 people in it.
2. Development against baseline
Per participant, current value divided by baseline value, for followers and for engagement rate. Report the median across participants, not the mean. The mean is wrong here and it is wrong in a specific way. One person with a post that travels moves the program mean by a third, and next quarter it drops back and looks like a decline. The median tells you whether the middle of your program moved.
3. Engagement quality
Median engagement rate per participant over the period, measured as interactions divided by impressions where you have impressions, and interactions divided by followers where you do not. Say which of the two you used. They produce different numbers and mixing them across quarters is the most common way these reports quietly break.
4. Consistency
Share of participants who published in every month of the quarter. This one predicts the others. Consistency drops before reach drops, usually by about a month, and it is the KPI you can actually act on, because somebody going quiet is a conversation and a reach decline is a post-mortem.
The quarterly review, in three parts
Leadership asks three questions about a program like this, in this order, and a report structured any other way gets interrupted.
Where did we start
The baseline, on one line per KPI. Same four numbers every quarter, same definitions, no additions. A report that grows a new metric each quarter reads as looking for a good number.
What moved
The four KPIs against baseline, with the median and the spread. The spread matters: 6 of 40 people carrying the entire program is a different situation than 30 of 40 moving a little, and both can produce the same average.
What we do next quarter
Two or three decisions, each attached to a number. More support for the six who are carrying it. A conversation with the eleven whose consistency dropped. Or the honest one, which is winding down a cohort that has not moved in two quarters.
Two claims that will cost you the room
Revenue the program cannot trace
Unless a deal came in through a tracked link and closed inside the quarter, attributing revenue to the program is a claim the CFO will test on the spot. Report pipeline influence only where the CRM can show it, and say so in the report.
Reach as the headline number
A quarter with one post that traveled looks better than a quarter of steady work, and the quarter after it looks like a slump. Lead with participation and median engagement, and keep reach underneath as context.
Doing this without a spreadsheet
All four KPIs are calculable by hand. What makes it painful is that they need the same definitions applied to every participant over the same period, quarter after quarter, and a spreadsheet maintained by one person does that until that person is on holiday. whoranks tracks the four inputs per participant on one scale and keeps the baseline from the day you add them, which is the part that is hard to reconstruct later.