Pages
Teams & Users — Strip Axis

Source: qdash /teams — c67 simplicity ↔ expressiveness study, 2026-08-15

The lensed chip strip sits at 0 on this axis. Four steps toward simplicity — each sheds a channel until only a sentence remains. Four steps toward expressiveness — each adds a channel until the lens switcher itself dissolves, because everything shows at once. The axis has a punchline: the switcher exists exactly at the midpoint — go simpler and you lose the lenses one by one; go richer and you stop needing them.

Lens (where a treatment can listen):
−1

Quiet chips

the bar and sub-line move to hover
keeps: lens, sort, amber landmark · loses: proportion at a glance (hover recovers the stat sheet) · earns: half the height — fits above the roster without claiming a section
−2

The ticker

the strip as one line of prose
· · · · ·
keeps: lens, sort, every value · loses: chip geometry, click-target size, the bar · earns: one text row — survives anywhere (a header, a Slack digest, a mobile width)
−3

One bar

five chips collapse into a single proportional object
Payments 24%Platform 25%Data 20%Frontend 8%Unassigned 23%
keeps: proportion (its whole point), amber share · loses: every lens except Spend — the first casualty of abstraction is flexibility · earns: the division of the whole in literally one glance
−4

The headline

the strip answers only "what's notable"
Platform leads spend at $339 (25%) — and $306 (23%) has no owner ·
keeps: the single most decision-relevant fact per lens, the amber stake · loses: the teams themselves — no comparison, no roster of values · earns: verdict-line grammar; this is what the Teams section becomes in the all-clear state, and arguably on the c68 landing
0

Baseline — the lensed chip strip

where we are now (strip-metrics page)
+1

Stat chips

all five lenses at once, in miniature — the switcher starts dissolving
Payments$318
per head $106active 2/3agents 2.7clear
Platform$339
per head $85active 3/4agents 2.51 Elev
Data$269
per head $67active 2/4agents 2.31 Crit
Frontend$108
per head $36active 1/3agents 2.3clear
Unassigned$306
per head $77active 1/4agents 1.3clear
keeps: chip form, amber landmark · loses: the lens (nothing to switch — it ignores the tabs above), scanability of any single metric across teams · earns: the follow-up question answered without a click
+2

Fingerprints

four normalized columns per team — compare shapes, not numbers
Payments
clear
Platform
1 Elev
Data
1 Crit
Frontend
clear
Unassigned
clear
columns, each scaled to fleet max:
spend · $/head · active · agents
risk stays words below — never a height
keeps: every ratio comparable at once · loses: absolute values (heights are normalized — the numbers live in tooltips), instant legibility for newcomers · earns: team "shapes" — Payments reads tall-spend/tall-head, Support-Ops would read short-spend/full-active
+3

People cards

the team as its people — one dot per person, state-colored
Payments$318
pna
2 running · clear
Platform$339
msrj
3 running · 1 Elevated
Data$269
sdtf
2 running · 1 Critical
Frontend$108
cjl
1 running · clear
Unassigned$306
cfc(
4 unmapped · 1 running
Critical openElevated openrunning nowidleunmapped
keeps: the selected lens as the top-right value, risk/activity per person · loses: compactness (dots need room; at 31-person ML-Infra the dots become a density strip) · earns: composition — the strip stops aggregating and shows who; felix IS the red dot, and clicking him is the roster filter
+4

The matrix

the strip unrolled into a scoreboard — every cell keeps its own encoding
TeamSpend$/personActiveRiskAgents
Payments$318$1062/3clear2.7
Platform$339$853/41 Elev2.5
Data$269$672/41 Crit2.3
Frontend$108$361/3clear2.3
Unassigned$306$771/4clear1.3
every column keeps its own bar semantic — share, scaled, meter, none · click a column header to sort (ship: DataTable)
keeps: everything — all metrics, all encodings, sortable · loses: the strip form entirely (this is a table claiming a section), the glanceability that made chips work · earns: the power-user view; no lens, no clicks, one scoreboard
Verdict

What the axis taught

Recommendation: baseline holds the section; −4 takes the all-clear

The lensed chip strip stays the working state's Teams section. The headline (−4) is what it collapses to when everything is clear — the verdict-line grammar the page already speaks, and the natural Teams voice on the c68 landing. One component, two densities, state-driven like everything else on this page.

People cards (+3) are the challenger worth testing

The only treatment that changes what the strip IS — composition instead of aggregate. felix IS the red dot; clicking him is the roster filter. Test against baseline at checkpoint; the open question is density (a 31-person team turns dots into a texture — needs a compressed form above ~12).

The matrix (+4) is the sub-route, not the strip

Everything, sortable, no lens — but it's a table claiming a section, and the roster below is already a table. If the checkpoint wants it, it's the Cost Center team table's upgrade (or /teams/mapping's sibling), not this strip. Ship note: it's DataTable with five slotted columns — nearly free.

Abstraction's tax is lens flexibility

One bar (−3) can only speak share-of-total; the headline (−4) speaks one fact. That's the axis lesson in one line: every step simpler removes a question the strip can answer, every step richer removes the need to ask. The switcher lives exactly where questions and answers balance.

Quiet chips (−1) are the mobile/narrow answer

Same lens, same sort, half the height — the strip's responsive collapse, not a replacement. The ticker (−2) is the export form: a Slack digest line, a header summary. Neither competes with baseline on the page; both are the same data dressed for other rooms.

Stat chips (+1) answer the tooltip's job, worse

All five numbers visible kills cross-team scanning — you read chips one at a time. The baseline's hover-tooltip carries the same follow-up without the noise. Fingerprints (+2) are the interesting failure: shape-comparison is real, but normalized heights without numbers demand expertise the strip's audience doesn't owe us.

Qpoint Brand Style Guide