Dashboard, four open questions

Everything else about this screen is ruled. These four are not, and each is baked into the build, so they are cheaper to settle now than after. Every option below is rendered in the real styles at the real sizes, and every one carries its cost as well as its case. Reply with four letters.

1 · The queue has two names

The tile counts the queue. The section lists the same queue. After the KPI ruling they are called different things, and the project convention is one concept, one surface.

ABoth read "Pending review"recommended
Pending review
6
One has waited 3h 20m

Pending review

3comment replies to approve before they sendoldest 3h 20m
2DM replies to mark right or wrong, so the bot learnsoldest 1h 02m
1hidden comment to delete for good, or put backoldest 44m

One name for one thing. The tile becomes an obvious summary of the section rather than a second concept to learn. Cost: the section heading loses the warmth of "Waiting for you", which was the one place the screen addressed you directly.

BTile counts, section addresses you
Pending review
6
One has waited 3h 20m

Waiting for you

3comment replies to approve before they sendoldest 3h 20m
2DM replies to mark right or wrong, so the bot learnsoldest 1h 02m
1hidden comment to delete for good, or put backoldest 44m

This is what ships today if nothing changes. Defensible on purpose: a KPI strip is a measurement register, a work list is an instruction register. Cost: two names for one queue, and the reader has to work out they are the same six things.

CBoth read "Waiting for you"
Waiting for you
6
One has waited 3h 20m

Waiting for you

3comment replies to approve before they sendoldest 3h 20m
2DM replies to mark right or wrong, so the bot learnsoldest 1h 02m
1hidden comment to delete for good, or put backoldest 44m

Also one name, taken from the other direction. Cost: it partly reverses this morning's ruling, since "Waiting for you" is the one tile label you rejected as too loose. Shown because consistency could reasonably outrank that.

2 · The caption register under formal labels

The labels are now formal. The captions under them are not. The caption is the only part of the tile that explains what the number means, so the choice is how much it says.

AConversational
Inbound contacts
128
Busier than usual, a Friday is normally near 90
Automation rate
71%
Up from 68% last week · 74 of 104
Spend today
$2.14
Higher than usual, normally near $1.60 by now

What ships today. Says the comparison in words anyone can check. Cost: "Busier than usual" under "Inbound contacts" is two registers in one tile, 30px apart.

BMeasuredrecommended
Inbound contacts
128
Above normal, Friday median 90
Automation rate
71%
Up from 68% last week · 74 of 104
Spend today
$2.14
Above normal, $1.60 by this hour

One register throughout, and it still names the baseline, so the claim stays checkable by hand. Shortest of the three without losing a number. Cost: "Friday median 90" asks slightly more of the reader than "a Friday is normally near 90".

CBare comparison
Inbound contacts
128
+42% vs Friday median 90
Automation rate
71%
+3pt vs 68% last week
Spend today
$2.14
+34% vs $1.60 by this hour

Densest, and the closest to a finance dashboard. Cost: "+34%" on spend reads as good until you notice the colour, and colour alone carrying the meaning is the thing the token sheet forbids everywhere else.

3 · What counts against the automation rate

This one is not wording. Each option is a different denominator and prints a different number on the same day, so it is worth thirty seconds. All three are honest; they answer different questions.

The same real day, three ways. 104 turns: 74 the bot answered, 8 it queued for you, 18 it deliberately did not answer, 4 fail-safes where the system, not the model, produced the outcome.
AEvery turn countsrecommended, and what is built
Automation rate
71%
Up from 68% last week · 74 of 104
turncountin the rate
answered74yes
queued for you8no
chose not to answer18no
fail-safe4no

Answers "of everything that arrived, how much did the bot handle". Hardest to game and the only one that cannot flatter itself: if the bot goes quiet, this number falls. Cost: a turn the bot was right to skip still drags it down, so it never approaches 100%.

BOnly turns that needed an answer
Automation rate
90%
About the same as last week · 74 of 82
turncountin the rate
answered74yes
queued for you8yes
chose not to answer18excluded
fail-safe4excluded

Answers "when an answer was owed, how often did it come from the bot rather than you". The truest measure of the review burden. Cost: a bot that silently skips everything scores 100%. The failure mode is invisible in the number.

CAnything that did not reach you
Automation rate
88%
Up from 85% last week · 92 of 104
turncountin the rate
answered74yes
chose not to answer18yes
queued for you8no
fail-safe4no

Answers "how much of the day did I not have to touch". Treats a deliberate skip as a success, which it often is. Cost: it counts silence as work done, and silence is exactly what nobody notices going wrong.

If you want one answer: A. It is the only one of the three that gets worse when the bot goes quiet, and going quiet without anyone noticing is the failure this product has already had once.

4 · What this screen is called

You ruled that the Dashboard IS Home, so there is one screen at one route. It still needs one name. Two rulings currently point different ways: the shell keeps its current nav, which says Home, and the mockup's own title says Dashboard. One word, in three places at once.

AHome, everywhererecommended
Home
Inbox
Contacts
Analytics

Home

checked 30 seconds ago

browser tab reads Home · Channels Manager

Follows your own shell ruling without amending it, and nobody has to relearn where anything is. "Dashboard" stays what we call it while building it, which costs nothing. Cost: the word Home says where it is, not what it does, and this screen is now a briefing rather than a landing page.

BDashboard, everywhere
Dashboard
Inbox
Contacts
Analytics

Dashboard

checked 30 seconds ago

browser tab reads Dashboard · Channels Manager

Says what the screen does, and matches the KPI register you just chose for the tiles. Cost: it amends the shell ruling made this morning, and it is a rename in the nav, the route's title and the tab, so it is three edits rather than none.

CHome in the nav, Dashboard on the page
Home
Inbox
Contacts
Analytics

Dashboard

checked 30 seconds ago

browser tab reads Dashboard · Channels Manager

This is what the two existing rulings produce if neither is touched, which is the only reason it is here. Cost: you click Home and land on a page headed Dashboard. Shown so it is chosen rather than arrived at.