Shared viewing suits presentation, not parallel work

What it is

Shared viewing — follow mode, screen sharing, any viewport coupling — has a crisp scope: it serves one-way knowledge transfer (one person walking others through material: walkthrough, demo, review) and does not serve parallel joint production (everyone working simultaneously on different parts of the material). Using a presentation tool as a workbench (co-editing a proposal with everyone sharing one view the whole time), or the reverse — presenting with free viewports where everyone pages separately and verbal alignment burns the session — are both mismatches between mode and task. The test: if a large share of task time demands "everyone looking at the same thing right now," use shared viewing; if it is "each on their own part, aligning occasionally," use free viewports plus awareness cues.

Why it happens

Coupling and parallelism are structurally exclusive. Shared viewing compresses many people's viewports into one, so at any moment the whole session has a single focus of attention — which is exactly what makes presentation efficient: information, reference, and pacing all aligned, with zero "where are we now" overhead. The cost is that navigation authority is collected into one person's hands; everyone else cannot look at the part they personally care about, and individual exploration, cross-checking, and private thinking are all blocked. Parallel work is the opposite: several people advance separate regions at once, efficiency comes from distribution, and it requires independent viewports and individually built reference frames. The two task classes have exactly opposite viewport needs, so no setting optimizes both — shared-viewing time and parallel output trade off directly. Healthy collaboration sessions therefore switch modes by phase: shared viewing to align understanding (walkthrough, decision), decoupled execution, re-alignment when needed.

Where it stops holding

The boundary is not binary; intermediate forms exist but at a cost. Loose following (align only when the leader pauses) partially accommodates "listen while reading ahead," but both experiences are discounted; split sharing (private views plus one shared region) fits tightly coupled tasks like pair inspection, at rising implementation complexity. Shared viewing also has a far lower headcount ceiling than parallel collaboration: broadcast-style walkthroughs scale to dozens, but as soon as bidirectional interaction is needed, co-viewing beyond five or six people decays into a speaking queue. And the more a task depends on individual rhythm (deep reading, verification, drafting), the more harmful forced synchronization becomes; the more it depends on collective rhythm (demo, review walkthrough), the more valuable it is.

Applying it

• Structure sessions by phase: shared viewing or following for walkthrough and decision phases; free viewports for execution — and make the switch an explicit, cheap mode control.
• In shared canvases and documents, provide a transient "look at this together now" convocation (pull everyone to one spot) instead of a permanently locked common view.
• Give the presenter signals (raised hands, detachments) as cues to switch back to discussion or slow down.
• Verification: track mode switches per session and their timing; sessions stuck long in shared viewing while showing heavy detach-and-return cycling are parallel work misfiled into presentation mode.

Related

• Same group: V3.04.1 Follow mode replicates one person's viewport changes to everyone else · V3.04.2 Fast movement by the followed party disorients followers · V3.04.3 Following must be escapable at any time, returning to one's own place · V3.04.4 Who follows whom must be visible to both sides
• Nearby: V5.02 Synchronous versus Asynchronous · V5.06 Screen Sharing and Pointing
• Search terms: co-browsing · shared viewport · presentation versus parallel work · coupled navigation