Right-Sizing Tiers

Seat reduction is the obvious saving. Tier reduction is the quieter one and it is frequently larger per user.

Why organisations over-tier

The premium tier was needed for one feature at purchase, and everyone was put on it.

A single requirement drove the whole account — an integration, a compliance feature, an administrative control.

The upgrade was easy and nobody revisited it.

Sales positioned the tier, not the requirement.

Tiers changed after purchase and the feature moved down a level.

The analysis

List what your tier includes that the one below does not. From the vendor's current pricing page, not from your memory of it at purchase.

For each distinguishing feature, establish whether it is used, and by how many people. Usage data from the vendor's admin console where available; otherwise ask the owner.

Frequently the answer is one feature, used by a few people, or a feature nobody has used at all.

The options

Downgrade entirely, if the distinguishing features are unused.

Split tiers. Premium for the users who need the feature, standard for everyone else. Many vendors support this and few volunteer it.

Buy the feature separately, where it is available as an add-on.

Replace the requirement. If one integration drives a premium tier for two hundred users, the integration may be achievable another way.

Negotiate the feature into the lower tier, which is occasionally possible at renewal for a committed customer.

Where it is harder than it looks

Administrative and security features are frequently what distinguishes tiers, and they apply to the account rather than to users. Single sign-on behind a premium tier is the standard example, and downgrading to save money at the cost of federation is a bad trade.

Compliance features — audit logs, retention controls, data residency — similarly apply account-wide.

Support levels are bundled into tiers and matter more than they appear when something breaks.

Weigh these deliberately rather than optimising on price alone. A tier that includes SSO and audit logging is frequently worth its premium on governance grounds even if the productivity features go unused.

The mixed-tier conversation

Ask directly: can we run standard for most users and premium for a named subset.

Some vendors support it natively. Some allow it contractually. Some refuse.

The refusal is worth testing at renewal, with usage data attached.

Where it is possible, the saving is proportional to how few people actually need the premium features, which is usually very few.

Recording it

Per application: current tier, tier below, distinguishing features, and which of those are used.

Reviewed annually, because vendors restructure tiers and a feature that justified a premium may have moved.

This is a short table and it produces a finding on a meaningful proportion of applications the first time it is filled in.

The features that justify a premium tier

Some tier differences are worth paying for on governance grounds regardless of productivity value.

Single sign-on, which many vendors place behind a higher tier. Downgrading to save money at the cost of federation is almost always the wrong trade.

Audit logging, which you need for incident response and compliance evidence.

Administrative controls — role separation, session policies, device restrictions.

Data residency options.

Retention and legal hold controls.

Deletion verification and export capability, which is the exit provision.

Assess these separately from the productivity features. A tier that costs more and gives you federation, audit logs and export is frequently worth it even when nobody uses the collaboration features it was sold on.

The annual tier review

Vendors restructure their tiers, and a premium justified two years ago may no longer be.

Pull the vendor's current pricing page for each material application.

Compare against what you are on and what you were told at purchase.

Check whether the feature that justified your tier has moved down a level, which happens regularly as products mature.

Check whether a new tier has appeared between the two you were choosing between.

Check whether your usage has moved you into a different band.

Twenty applications, an afternoon, once a year, and it reliably finds at least one tier that can drop.