Dumbbell Chart | Zyeon UI

status="ready" · 9 markets sorted by improvement. Japan launched mid-year (0% → 12.6%), so its relative change is undefined rather than ∞; Italy read 29.8% twice and gets the ringed "unchanged" marker instead of a lone dot.

Checkout conversion by market

Share of carts that reached payment · Q1 vs Q4

Q1Q4improvementdecline

Rows ordered by improvement, best first. Higher is better, so an improvement is drawn as a rise and carries the up arrow. Bar length is the change on one scale shared by every row; the hollow dot is Q1, the filled dot Q4. Arrow keys move between rows.

Checkout conversion by market: every category at Q1 and Q4, with its change. Rows are ordered by improvement, best first. Rising values are improvements.
Rank	Category	Q1	Q4	Change	Relative	Direction
1	Germany	42.1%	61.8%	+19.7%	+46.8%	up, an improvement
2	Japan	0%	12.6%	+12.6%	undefined (started at zero)	up, an improvement
3	Netherlands	44.5%	52.3%	+7.8%	+17.5%	up, an improvement
4	Spain	38.4%	44%	+5.6%	+14.6%	up, an improvement
5	Sweden	47.7%	49.1%	+1.4%	+2.9%	up, an improvement
6	Italy	29.8%	29.8%	no change	0%	unchanged
7	France	51.2%	47.6%	-3.6%	-7%	down, a decline
8	United Kingdom	58.9%	55.2%	-3.7%	-6.3%	down, a decline
9	Poland	33%	25.4%	-7.6%	-23%	down, a decline

status="empty" · the contract parsed, there is simply nothing to compare yet.

Checkout conversion by market

Share of carts that reached payment · Q1 vs Q4

Nothing to compare yet

A row appears once a category has a value at both Q1 and Q4.

status="loading" · six pulsing bars, aria-hidden, with an sr-only role="status" line. No timers.

Checkout conversion by market

Share of carts that reached payment · Q1 vs Q4

higherIsBetter={false} · p95 latency: a fall is the good news, so the falls are coloured as improvements and sort to the top while the one regression sinks. Four-digit values widen the delta lane.

p95 latency by endpoint

Before and after the edge cache rollout

BeforeAfterimprovementdecline

Rows ordered by improvement, best first. Lower is better, so an improvement is drawn as a fall and carries the down arrow. Bar length is the change on one scale shared by every row; the hollow dot is Before, the filled dot After. Arrow keys move between rows.

p95 latency by endpoint: every category at Before and After, with its change. Rows are ordered by improvement, best first. Falling values are improvements.
Rank	Category	Before	After	Change	Relative	Direction
1	/v2/projects/{id}/settings	790 ms	374 ms	-416 ms	-52.7%	down, an improvement
2	/v2/webhooks/deliveries	585 ms	212 ms	-373 ms	-63.8%	down, an improvement
3	/v2/organizations/{id}/members	1,048 ms	754 ms	-294 ms	-28.1%	down, an improvement
4	/v2/invoices/{id}/pdf	1,079 ms	818 ms	-261 ms	-24.2%	down, an improvement
5	/v2/search	479 ms	292 ms	-187 ms	-39%	down, an improvement
6	/v2/exports	821 ms	681 ms	-140 ms	-17.1%	down, an improvement
7	/v2/events:batch	902 ms	1,238 ms	+336 ms	+37.3%	up, a decline

Negatives that mean something · two lines lost money and Cloud crossed zero between the readings, so a dashed rule marks 0 and the bars are read against it.

Operating margin by product line

FY23 vs FY24 · negative means the line lost money

FY23FY24improvementdecline

Rows ordered by improvement, best first. Higher is better, so an improvement is drawn as a rise and carries the up arrow. Bar length is the change on one scale shared by every row; the hollow dot is FY23, the filled dot FY24. Arrow keys move between rows.

Operating margin by product line: every category at FY23 and FY24, with its change. Rows are ordered by improvement, best first. Rising values are improvements.
Rank	Category	FY23	FY24	Change	Relative	Direction
1	Cloud	-3.2%	6.8%	+10%	+312.5%	up, an improvement
2	Hardware	-8.4%	-2.1%	+6.3%	+75%	up, an improvement
3	Support	21.4%	24%	+2.6%	+12.1%	up, an improvement
4	Services	12%	9.5%	-2.5%	-20.8%	down, a decline
5	Legacy licences	4%	-6.5%	-10.5%	-262.5%	down, a decline

Degenerate · every value is 25 seats. The span is 0, so a linear scale would divide by zero; every marker is centred instead and the summary says the scale carries no information.

Seats per workspace

Every workspace is on the same fixed 25-seat plan

RenewalTodayimprovementdecline

Rows ordered by improvement, best first. Higher is better, so an improvement is drawn as a rise and carries the up arrow. Bar length is the change on one scale shared by every row; the hollow dot is Renewal, the filled dot Today. Arrow keys move between rows.

Seats per workspace: every category at Renewal and Today, with its change. Rows are ordered by improvement, best first. Rising values are improvements.
Rank	Category	Renewal	Today	Change	Relative	Direction
1	Acme	25 seats	25 seats	no change	0%	unchanged
2	Globex	25 seats	25 seats	no change	0%	unchanged
3	Initech	25 seats	25 seats	no change	0%	unchanged
4	Umbrella	25 seats	25 seats	no change	0%	unchanged

Degenerate · a single row: the band, the "1 of 1" readout and the summary all hold at a sample of one, and the arrow-key hint drops out because there is nowhere to move.

Trial-to-paid conversion

The only cohort that has finished its trial window

Jan cohortFeb cohortimprovementdecline

Rows ordered by improvement, best first. Higher is better, so an improvement is drawn as a rise and carries the up arrow. Bar length is the change on one scale shared by every row; the hollow dot is Jan cohort, the filled dot Feb cohort.

Trial-to-paid conversion: every category at Jan cohort and Feb cohort, with its change. Rows are ordered by improvement, best first. Rising values are improvements.
Rank	Category	Jan cohort	Feb cohort	Change	Relative	Direction
1	Self-serve	18.4%	23.9%	+5.5%	+29.9%	up, an improvement

sort="magnitude" · 24 cities at rowHeight={24}, valueLabels="never", labelWidth={168}. All 24 rows are rendered — the card grows, it never scrolls a row out of sight.

Weekly active accounts by city

24 cities · week 12 vs week 24

Week 12Week 24improvementdecline

Rows ordered by how far they moved, largest first. Higher is better, so an improvement is drawn as a rise and carries the up arrow. Bar length is the change on one scale shared by every row; the hollow dot is Week 12, the filled dot Week 24. Arrow keys move between rows. Point at or focus a row to read its two values.

Weekly active accounts by city: every category at Week 12 and Week 24, with its change. Rows are ordered by how far they moved, largest first. Rising values are improvements.
Rank	Category	Week 12	Week 24	Change	Relative	Direction
1	New Buddy metro area	9,713	14,014	+4,301	+44.3%	up, an improvement
2	Connietown metro area	7,393	10,964	+3,571	+48.3%	up, an improvement
3	Hartmannboro metro area	7,339	10,386	+3,047	+41.5%	up, an improvement
4	Mariamtown metro area	5,751	8,226	+2,475	+43%	up, an improvement
5	New Lorenzo metro area	9,231	7,384	-1,847	-20%	down, a decline
6	Sethfort metro area	8,359	10,045	+1,686	+20.2%	up, an improvement
7	Cranston metro area	8,010	9,558	+1,548	+19.3%	up, an improvement
8	Vacaville metro area	2,826	4,286	+1,460	+51.7%	up, an improvement
9	Christiansenstad metro area	6,683	5,256	-1,427	-21.4%	down, a decline
10	Morarport metro area	7,485	8,706	+1,221	+16.3%	up, an improvement
11	Brendacester metro area	7,163	8,166	+1,003	+14%	up, an improvement
12	Moore metro area	3,594	4,510	+916	+25.5%	up, an improvement
13	Oak Park metro area	3,761	2,983	-778	-20.7%	down, a decline
14	Lake Shea metro area	8,104	8,730	+626	+7.7%	up, an improvement
15	Reillystad metro area	1,885	2,438	+553	+29.3%	up, an improvement
16	South Fae metro area	5,323	4,846	-477	-9%	down, a decline
17	Montebello metro area	1,014	1,459	+445	+43.9%	up, an improvement
18	Lubbock metro area	8,346	8,663	+317	+3.8%	up, an improvement
19	Fort Raymond metro area	5,609	5,294	-315	-5.6%	down, a decline
20	South Geraldine metro area	9,046	9,286	+240	+2.7%	up, an improvement
21	Madison metro area	2,217	2,316	+99	+4.5%	up, an improvement
22	East Lelastad metro area	6,750	6,688	-62	-0.9%	down, a decline
23	The Hammocks metro area	229	282	+53	+23.1%	up, an improvement
24	Fort Frank metro area	209	171	-38	-18.2%	down, a decline

status="ready" with two rows sharing an id · the structural check refuses to draw and names the id, because a duplicate id silently detaches a row from its readout and its tab stop.

Checkout conversion by market

Share of carts that reached payment · Q1 vs Q4

This comparison can't be drawn

Two rows share the id "de"; ids are what keeps a row attached to its readout and its tab stop.

status="error" · onRetry wired — Try again swaps in the ready payload.

Checkout conversion by market

Share of carts that reached payment · Q1 vs Q4

Couldn't load this chart

The data source didn't respond.

onSelectItem wired · rows become role="button" and answer to click, Enter and Space; sort="end" ranks by the latest value instead of by change.

Operating margin by product line

FY23 vs FY24 · negative means the line lost money

FY23FY24improvementdecline

Rows ordered by their latest value, largest first. Higher is better, so an improvement is drawn as a rise and carries the up arrow. Bar length is the change on one scale shared by every row; the hollow dot is FY23, the filled dot FY24. Arrow keys move between rows.

Operating margin by product line: every category at FY23 and FY24, with its change. Rows are ordered by their latest value, largest first. Rising values are improvements.
Rank	Category	FY23	FY24	Change	Relative	Direction
1	Support	21.4%	24%	+2.6%	+12.1%	up, an improvement
2	Services	12%	9.5%	-2.5%	-20.8%	down, a decline
3	Cloud	-3.2%	6.8%	+10%	+312.5%	up, an improvement
4	Hardware	-8.4%	-2.1%	+6.3%	+75%	up, an improvement
5	Legacy licences	4%	-6.5%	-10.5%	-262.5%	down, a decline

Nothing selected yet — click a row, or Tab into the chart and press Enter.