Intranet rollout: a step-by-step plan for the first 90 days

Most intranet projects fail for reasons that have nothing to do with technology. The system works, content gets loaded, and by month three the daily active user count is falling. The plan below splits the first 90 days into three phases to prevent exactly that.

Days 1–30: narrow scope, real content

The first month serves a single purpose: don't launch an empty intranet. If an employee's first visit shows nothing useful, there won't be a second one.

• Week 1 — decide the scope. Start with three modules only: announcements, people directory and documents. Deliberately leave the rest switched off.
• Week 2 — organisational data. Department tree, managers, employee list. If you have a corporate directory, set up matching now; maintaining the same data twice becomes tiring from the first month.
• Week 3 — content. Upload the twenty most-requested documents. "Let's migrate everything" eats months; the most-asked-for documents produce most of the value.
• Week 4 — pilot group. Pick 20–40 people from different departments, and make sure a field team is included.

Days 31–60: pilot and correction

A pilot is not a demo; it is a period in which real work is done. This month, announcements must genuinely be published on the intranet and not repeated in parallel by email. Parallel publishing makes the pilot result impossible to measure.

Three things to measure:

1. Weekly active user rate. What share of the pilot group signs in at least once a week?
2. Announcement read rate. How many people opened the critical announcement, and how many acknowledged it?
3. Feedback. What couldn't they find? The answer to that question usually rewrites the menu.

Always end the pilot with a correction round. The most frequent findings are menu naming, homepage layout and mobile usability.

Days 61–90: launch and habit

General availability should be a campaign, not an email. Three practices that work:

• One mandatory task. During launch week, give everyone one real job that can only be done on the intranet: completing a survey, acknowledging a policy.
• Managers first. If department heads send their team's first announcement through the intranet, the habit spreads downward.
• Close the old channel. As long as the chat group stays open for announcements, the intranet stays in second place.

At the end of the third month, open the second wave: requests, recognition, surveys. The reason for that order is that every module added before the first wave becomes habit competes for attention. You can review the full set on our modules page.

Content migration: what should move?

Wanting to move everything from the old portal is the trap that delays rollouts most. What works is sorting into three boxes.

Move: documents actually used in the last year, current, with a known owner. In most companies that is under ten per cent of the archive.

Rewrite: documents whose content is valid but whose format is dated. Rewriting them into the new structure is faster than migrating them as-is.

Archive: everything no longer used. It isn't deleted, but it doesn't enter the new intranet either.

The quickest way to make this split is to look at access logs in the old system. Migrating a document nobody ever opened only pollutes search.

Four common mistakes

Switching on every module on day one. The homepage gets crowded and nothing gets owned.

Not assigning content owners. Every section needs a named owner; otherwise content goes stale within six months. We cover this in the content ownership article.

Leaving mobile until later. For field teams mobile is not a secondary channel — it is the only one.

Not measuring. Without numbers, "did it work?" becomes an argument about opinions six months later.

What should day 90 look like?

In a healthy rollout, the weekly active user rate is above 60 per cent by the end of month three, critical announcements approach an 80 per cent read rate, and repetitive questions to HR visibly decline. If those three numbers aren't there, the problem is usually not the product but a scope that was drawn too wide.

To settle the budget beforehand see our cost article; for vendor comparison, the selection guide.