Employee advocacy program rollout: the 90-day playbook

A 90-day employee advocacy rollout that survives week six — voice capture first, 120-profile watchlists second, gamification on engagement earned last.

Employee advocacy program · Advocacy rollout · LinkedIn team rollout

Key takeaways

Sequencing is the whole game — voice first, watchlists second, gamification last. Reversing the order produces a launch spike and a week-six collapse.
Days 1–30 are voice capture — connect via OAuth, start the comment cadence small, and let each person's drafting model settle through roughly 20 to 40 edited comments.
Days 31–60 build the targeting — a 120-profile watchlist per rep (roughly 60 buyers, 30 amplifiers, 30 deal-stage targets) and the owner-approval workflow that keeps every action consented.
Days 61–90 add the scoreboard — leaderboards ranked on engagement earned from watchlist buyers, never on volume pushed, because whatever the leaderboard counts is what people will game.
Instrument attribution before the participation push, so the day-90 leadership review shows pipeline touched rather than share counts.
Roll out to a committed pilot cohort before the wider team — a working motion recruits better than a mandate.

Why rollouts die at week six

The standard advocacy rollout is a launch email, a content library, a leaderboard, and a prize — and it produces a participation spike that decays to three volunteers within six weeks. The failure is not enthusiasm; it is sequencing. Gamification arrived before anyone trusted the content coming out under their name, targeting never arrived at all, and the metrics counted pushes instead of responses — so the rational employee did the minimum, then stopped. (The deeper model-level reasons live on the employee advocacy platform hub, which compares broadcast advocacy against the engagement-led model this playbook rolls out.)

The fix is to build the program in the order the motion actually depends on: voice before volume, targeting before scoreboard, attribution before applause. Ninety days, three phases, each with a definition of done.

Days 1–30: voice capture

The first month has one goal — every participant trusts what drafts under their name — because no targeting or gamification can rescue a program whose members are embarrassed by their own output.

The pilot cohort connects through LinkedIn's official OAuth flow — no extensions, no shared credentials — and starts a deliberately small daily cadence: a handful of comment drafts a day, each one edited and approved by the person before it goes anywhere. Those edits are the training signal. The voice model learns from each person's comment history and corrections, and it settles into their real register after roughly 20 to 40 edited comments — for a rep doing a few a day, that is two to four weeks, which is why this phase gets a full month and why skipping it shows.

What to deliberately not do in this phase: no leaderboard, no participation mandates, no company-content push. The only management attention is a weekly fifteen-minute check-in asking one question — do the drafts sound like you yet?

Definition of done: every pilot member approves most drafts with light edits rather than rewrites, and nobody has published anything they would not have written themselves.

Days 31–60: watchlists and the approval workflow

The second month points the now-trusted voice at the right people — a 120-profile watchlist per rep — and hardens the consent workflow before the volume grows.

Each rep builds a watchlist of 120 profiles, weighted roughly 60 buyers, 30 amplifiers, 30 deal-stage targets, drawn from their actual book — the composition logic and sourcing tactics are in how to build a LinkedIn watchlist. SDRs weight toward net-new buyers; AEs weight toward open-opportunity stakeholders. The watchlist is what turns daily engagement from a content hobby into a sales motion, and the sales-team version of the playbook covers the rep-level cadence in detail.

In parallel, two pieces of plumbing go in:

The owner-approval workflow, formalized. Every action still waits for the profile owner's explicit yes — that never changes as volume grows. Marketing's role is curating targets and surfacing themes, never publishing as a person. Write this down as program policy now, while the cohort is small, so it is culture rather than a rule bolted on later.
The attribution lane, before the participation push. Watchlist engagement starts syncing to the CRM via the LinkedIn-to-HubSpot sync in this phase — not day 89 — so that by the time leadership asks what the program produced, the answer is a CRM report. What to count and what to ignore is its own discipline, laid out in employee advocacy analytics.

Definition of done: every rep runs a full 120-profile watchlist, daily engagement lands on watchlist buyers rather than the open feed, and engagement is visibly accruing on CRM contact timelines.

Days 61–90: gamification done right

Only now does the scoreboard arrive — and it ranks engagement earned, never volume pushed, because whatever the leaderboard counts is what people will game.

The broadcast-era leaderboard counted shares, so it manufactured sharing — hollow, identical, buyer-repellent. The engagement-led leaderboard ranks what buyers gave back: reactions and replies earned from watchlist profiles, weighted toward buyers over peers. Under that metric, the only way to climb is to consistently say things real buyers respond to — the leaderboard becomes a proxy for market resonance instead of a proxy for obedience. Keep it light: a weekly rhythm, visible to the team, celebrated briefly. The moment a leaderboard becomes a performance-review input, people optimize it adversarially.

The same month, widen the cohort. The pilot's results — real buyer replies, engagement on deal-stage accounts, the first attributed touches — are the recruiting material. A working motion sells itself internally in a way no launch email can; this is why the wide rollout comes last, not first.

Close the 90 days with the leadership review the program was instrumented for: engagement received from buyers, pipeline attributed in the CRM, and the participation floor. Calibrate the engagement numbers against real platform distributions — The State of LinkedIn gives the honest baselines — so nobody judges a healthy program against a fantasy average.

Definition of done: the leaderboard has run for four weeks without producing forced-comment behavior, the second cohort is through voice capture, and the day-90 review led with attributed pipeline.

The failure modes, named in advance

Four ways this rollout goes wrong, so you can catch each one in the phase where it starts:

Skipping voice capture because week one feels slow. The drafts read generic, reps quietly disengage, and by the time the leaderboard launches there is nothing authentic to amplify. The 20-to-40-edited-comments settling period is the program's foundation, not its warm-up act.
Watchlists built by marketing instead of the rep. A list the rep did not shape is a list the rep will not work. Marketing curates and suggests; the rep owns the final 120.
Leaderboards on volume. Covered above — measure pushes and you will drown in pushes. If you take one rule from this page, rank on engagement earned.
A territory whose buyers are not on LinkedIn. No rollout fixes this. Check buyer activity during the pilot-selection step, and route reps whose markets live elsewhere to other motions without stigma.

If you would rather pressure-test the sequencing against your own team before committing a quarter to it, a 30-minute walkthrough will map the three phases onto your roster — which reps pilot, what their watchlists contain, what the day-90 review can honestly promise your leadership — and flag in the meeting if your buyer activity makes phase two a stretch.

FAQ

Common questions

How long does an employee advocacy program take to roll out?

Ninety days to a defensible leadership review, on this sequencing — a month of voice capture, a month of watchlist build and approval workflow, a month of gamification and reporting. The commonly promised two-week rollout is possible only in the broadcast model, where launch means employees got a login and a content library — which is also why those programs have nothing left to show by week six.

Who should be in the first advocacy cohort?

A pilot of genuinely willing people whose buyers are active on LinkedIn — typically a founder or exec sponsor, two or three reps who already engage a little, and one marketing operator who owns the program. Six to ten people is plenty. Skip anyone joining under duress; a skeptic converted by a working pilot recruits the rest better than any mandate, and an early cohort of conscripts kills the program's internal reputation before it has results.

Why do employee advocacy leaderboards backfire?

Because whatever the leaderboard counts is what people will produce. Rank on posts pushed or shares clicked and you get volume — forced comments, hollow reshares, activity theater that buyers can smell. Rank on engagement earned from watchlist buyers and the only way to climb is to say things real buyers respond to. Gamification is a multiplier on whatever behavior you measure; that is exactly why it goes last, after the behavior being multiplied is the right one.

What does the owner-approval workflow mean in practice?

Nothing appears on anyone's profile without that person approving it. Each teammate connects through LinkedIn's official OAuth flow — no browser extension, no shared passwords, no session cookies — and every drafted comment or post waits for the owner's edit and explicit yes. Marketing curates targets and themes; it never publishes as a person. This is both an ethics position and an adoption one — people engage honestly through a motion they visibly control.

What should the day-90 leadership review contain?

Two headline numbers and one distribution. Engagement received from watchlist buyers — filtered to buyers, not peers — and pipeline attributed, meaning deals in the CRM carrying LinkedIn touches from the program. Plus the participation floor — how many reps ran the daily motion in the final month. If the attribution plumbing went in during days 31–60 as this playbook sequences it, the pipeline number is a CRM report, not an estimate assembled the night before.