The Challenge
Communications teams managing issues across channels frequently make a category error: they respond to an average sentiment score that masks dramatically different platform realities. A single public statement calibrated to a blended negative majority might be perfectly appropriate for one platform and actively counterproductive on another.
The Greenland acquisition story created a clean test case: a single policy announcement with high-salience geographic and sovereignty implications, generating engagement across platforms with meaningfully different user demographics. The question was whether the platforms differed only in degree — or whether the strategic implications were categorically different.
The Approach
Rolli IQ analyzed Meta platforms (Instagram, Facebook, Threads) and Twitter/X separately, tracking sentiment distribution for each rather than aggregating across platforms. Both platform sets showed strong negative majorities — approximately 66% negative on each — so a surface-level reading would have missed the critical difference.
The divergence was in the positive ceiling: Meta showed only 1% positive engagement; Twitter/X showed approximately 20% positive. Same negative majority — but a 19-point gap in the audience that could be activated for amplification. Platform-specific strategy implied two entirely different response postures.
The investigation confirmed the divergence was not statistical noise from different user base sizes. The composition of positive engagement on Twitter/X — the accounts involved, their network connections, and their organic reach — represented a viable amplification surface. The equivalent audience simply did not exist on Meta for this topic.
The Findings
- 1%positive sentiment on Meta platforms (Instagram, Facebook, Threads)
- 20%positive sentiment on Twitter/X — same negative majority, different ceiling
- 19 ptsplatform sentiment gap confirmed as actionable strategy signal — not noise
- ~66%negative sentiment on both platforms — blended average would have masked the gap
“The negative majorities were identical. But Twitter/X had a 20% positive ceiling; Meta had 1%. Those aren't the same strategy — and a blended average would have hidden both.”