AAA × CatalyzeCX — Unified Reporting Rule Book (v1.9 · MASTER)

Purpose: single source of truth for the three Claude reporting sessions — Command Center (dashboard / build.py), Daily Slides (analyze.py / build_deck.js), and Performance Review. Every number in all three must follow the rules below.

Data window: 6/29/2026 → last complete day (Central). v1.9 generated 2026-08-04 (supersedes v1.8). This file is the master — sessions may keep their own working copies, but on any conflict this master wins.

Standing duty for all three sessions: treat this document as a living, shared rule book. (1) Whenever a rule, definition, value, or list size changes in any session, update the master in the same pass, notify Janina, and bump the changelog so all three re-sync. (2) Continuously watch for and report data anomalies — during current work and in future runs. Full detail in Section H.

---

A. Locked decisions (the 7 alignment calls)

1. Productive hours — include Ring

Productive hours = Talk + Wrapup + Hold + In-Review + Ring, summed per call (caps in rule 2), ÷ 3,600,000 (ms → hours). One hours number; labor cost / CPS uses these same hours (no separate "occupied hours").

2. Per-call time caps — Talk 90 min, all others 30 min

Cap each component per call before summing: Talk = 90 min (5,400,000 ms); Wrapup / Hold / In-Review / Ring = 30 min each (1,800,000 ms). Values above the cap are clamped.

3. Sales — count "Sale" (and its aliases), no talk-time gate

Sales = rows where Disposition is Sale, Renewed, or Account Renewed (the last two are aliases of Sale — a closed renewal). No minimum talk time. 259 (thru 7/21) / 269 (thru 7/22). Net Contact keeps its own >30s rule; a Sale always counts as a Net Contact regardless of talk time.

• Do not confuse with "Already renewed" — that is NOT a sale. It means the customer had already renewed elsewhere; it is an ineligible contact (Section D), never a sale.
• Alias note: Renewed / Account Renewed have 0 rows in the current 20260722 extract, so sales are unchanged (259/269). The rule is forward-looking — if these aliases appear in a future export they must count as sales. Change needed — Daily Slides & Performance Review: add Renewed and Account Renewed as sale aliases (Command Center already does).

4. Contacts — gross, no talk-time gate

Contacts = any call reaching a person by disposition (eligible + ineligible tiers — see Section D), inbound and outbound. No talk gate. Gross Contacts = ~14,702 (6/29–7/21) / ~15,376 (6/29–7/22). Includes Auto, Invalid number, and Not interested; excludes the recall/callback family. Confirmed against Command Center (15,374) and Daily Slides (15,360), which match within a few calls. (The earlier ~14,215/14,871 figures were wrong — they dropped the trailing-space 'Not interested ' label; see cleaning step 6 and Section G.) Net Contact is the separate talk-gated metric — do not merge the two.

5. Calendar date & day-of-week — US/Central

Convert Created At (epoch ms) → datetime, then take the date and weekday in US/Central. Applies to day bucketing, the window, and the cutoff.

6. Weekends — included; no sparse-day threshold

Keep all days, weekends included. No minimum-dials-per-day rule. Only the trailing incomplete day is dropped (plus stray next-day rows cumulative exports carry).

7. List sizes — from the List Health report (authoritative)

One authoritative Callable-Leads table for all three (drives List Penetration and List Completion):

One authoritative Callable-Leads table (drives List Penetration and List Completion). Cycle = launch month, assigned explicitly (not date-derived).

Cycle 7 (in scope, launched):

Campaign	Code	Callable Leads	Cycle
Salvage	ACA00001	5,974	7
Salvage V2	ACA00002	14,430	7
PRX-30 AR Decline	ACA00003	7,995	7
PSX-1 DB-1	ACA00004	10,268	7
PRX-15	ACA00006	11,973	7
PRX-45	ACA00007	3,846	7
PSX AR Decline DB	ACA00008	6,763	7
PSX-30	ACA00009	5,994	7

Cycle 8 (added 2026-08-03; "Not Received", list_size 0 until lists arrive):

Campaign	Code	Name	Dial window	Cycle
Salvage 45-post	ACA00010	ACA00010_REC_PSX_Salvage (45 day Post)	7/22–8/23	8
PRX-30 AR Decline	ACA00011	ACA_REN_PRX-30_AR Decline	8/4–9/2	8
PSX-1 DB-1	ACA00012	ACA_REN_PSX-1_DB-1	8/6–9/5	8
Salvage 45-post	ACA00013	ACA_REC_PSX_Salvage (45 day Post)	8/15–9/28	8
PSX AR Decline	ACA00014	ACA_REN_PSX_AR Decline	8/18–9/16	8
PRX-15	ACA00015	ACA_REN_PRX-15	8/18–9/1	8
PRX-45	ACA00016	ACA_REN_PRX-45	8/18–9/16	8

In scope now = all 15 (ACA00001–00004, 00006–00016). ACA00005 (PSX-15 DB) remains "Not Received", out of scope until it has data. ACA00007 and ACA00008 are now in scope (were excluded in v1.6). Cycle-8 campaigns carry list_size 0 so Penetration/Completion correctly show "—" until real lists are entered.

Program → cycle pairing (for cycle-over-cycle comparison): Salvage 45-post = {C7: 00001+00002 combined, C8: 00010+00013 combined}; PRX-30 AR Decline 00003→00011; PSX-1 DB-1 00004→00012; PSX AR Decline 00008→00014; PRX-15 00006→00015; PRX-45 00007→00016; PSX-30 00009→(no C8 yet).

Matching: the engine maps a call to a campaign by finding the ACA##### code anywhere in the Name field (exact → normalized → regex fallback). Because each new campaign's ID is embedded in its name, calls auto-map even when the descriptive wording repeats across cycles (e.g. 00003 and 00011 are both "PRX-30 AR Decline" — the ID keeps them distinct).

---

B. Shared rules (all three match exactly)

Source & row cleaning (in order)

1. Use the latest cumulative call-detail CSV. Pass the prior clean file (Customer_Call_Details_20260702.csv) as --prior for de-mislabeling.
2. Drop the summary/total row (blank Call Id, or Call Id under 10 chars).
3. Exclude test/pilot agents — remove their rows entirely:
   • cesar.ayala@catalyzecx-sv.com
   • jason.barry+focal@nextiva.com
   • jcarlos.angel@catalyzecx-sv.com
   • sara.hernandez@catalyzecx-sv.com
   • kent.ingram@catalyzecx-sv.com
   • paul.garcia@catalyzecx-sv.com
   • federico.ramirez+focalpoint@nextiva.com (added v1.7)
   • claudia.orozco+focal@nextiva.com (added v1.7)
   • KEEP sara.villeda@catalyzecx-sv.com — a real agent, not excluded. Match logic: an agent is excluded if their address exactly equals, or contains, any string above (lowercased substring match). Federico/Claudia use their +focal/+focalpoint @nextiva.com addresses — the exact strings Janina supplied. Kent Ingram and Paul Garcia have 0 rows in the 20260722 extract, so this exclusion has no effect on current numbers — it's a forward-looking guard. Confirm their exact Agent identifiers so future extracts match reliably.
4. De-duplicate by Call Id only (keep first). This is the sole dedup key — do not add secondary keys.
5. Drop rows with an unparseable timestamp.
6. Normalize disposition text — trim leading/trailing whitespace before any tier matching. The source has labels with trailing spaces (confirmed: 'Voicemail ' and 'Not interested '). Exact-string matching without trimming silently drops those rows — this is what caused Performance Review to undercount contacts and net contacts (see Section G). Strip whitespace on every string field used for matching (Disposition at minimum; ideally Agent and Type too).

Reporting window

Start floor 2026-06-29 (exclude earlier + any Test/Pilot rows). Run through the last complete day (Central); drop the trailing partial day and stray next-day rows.

Units

All duration fields (Talk, Hold, Wrapup, Handle, In-Review, Ring, Queue, Abandon) are milliseconds.

Campaign tagging & de-mislabeling

• Campaign = first ACA##### pattern from the Name field.
• De-mislabeling: a REN_PSX_30 / ACA00009 row (Name matches REN_PSX-30, REN_PSX_30, or ACA00009) is real ACA00009 only if (a) Call Id is not in the prior file and (b) date is ≥ 2026-07-02. Otherwise re-tag to ACA00001.
• Salvage split: any ACA00001 row dated ≥ 2026-07-01 moves to ACA00002.
• Rescued sale — REMOVED (v1.3). The hard-coded sara.villeda 6/30 sale is absent from the current extract (her sales start 7/2). Do not hard-code it; match the extract. If genuine, fix upstream in the export. All sessions land on 259 / 269 without it.
• Campaigns in scope (15): ACA00001–00004, 00006–00016 (see Section A.7 for the full list, list sizes, and cycles). ACA00005 out of scope (Not Received).

Call-type scope

• Dials = outbound only. Inbound never counts as a dial.
• Contacts, Net Contact, Sales include both inbound and outbound.
• Disposition "0" / blank = non-call: excluded from contacts, net contacts, sales. Exception: the inbound funnel's "calls received" layer counts them.

---

C. Metric definitions (identical in all three)

Metric	Definition
Dials	Count of outbound calls (in-scope campaigns, in window).
Contacts (gross)	Calls reaching a person by disposition (eligible + ineligible tiers), inbound + outbound. No talk gate.
Net Contact	An eligible-tier call (except "Already renewed") with Talk Time > 30s, OR any Sale.
Sales	Disposition == "Sale". No talk-time gate.
Conversion	Sales ÷ Net Contact.
Productive hours	(Talk + Wrapup + Hold + In-Review + Ring), per-call caps from rule 2, ÷ 3,600,000.
DPH / CPH / SPH	Dials / Contacts / Sales ÷ Productive hours.
CPS	(Productive hours × $16.75) ÷ Sales.
Contact %	Contacts ÷ Dials (can exceed 100% — contacts include inbound).
Net Contact %	Net Contact ÷ Contacts.
List Penetration	Unique outbound leads dialed (distinct last-10-digit To) ÷ list size, capped 100%.
List Completion	Dials ÷ (list size × 6-attempt cap). Cap changed 8→6 on 2026-08-03 (v1.7); completion % rises for every campaign vs the old 8-cap with no change in dialing.

Billed rate: $16.75 per productive hour (CPS), all three.

---

D. Disposition tiers — FINAL (v1.4, all 25 dispositions classified)

Tier	Dispositions	Contact?	Net Contact eligible?
Eligible	Sale (+ aliases Renewed, Account Renewed), Will renew later, Interested / Not Ready, Not interested, Hang up, Decision Maker Unavailable, Lost Connection – Immediate Callback	Yes	Yes (if Talk > 30s; Sale always)
Ineligible	Do not call, Escalation, Wrong number, Deceased, Language barrier, Already renewed, Auto, Invalid number	Yes	No
Not a contact	Voicemail, Answering machines, "0"/blank, Ring / No Answer, Completed, Transferred, Recall Request, Recall No Specific Time, General Callback, Dropped Call/Callback	No	No

Classification rulings (from Janina):

• Renewed / Account Renewed → SALE (v1.5). Aliases of the Sale disposition; count as sales and sit in the Eligible tier. Distinct from Already renewed, which is an ineligible contact and never a sale. (0 alias rows in the current extract — forward-looking.)
• Invalid number → ineligible CONTACT. Counts toward gross Contacts (~2,200 raw rows), never toward Net Contact or sales. Applied in all three.
• Recall / Callback dispositions → NOT a contact. This covers Recall Request, Recall No Specific Time, General Callback, and Dropped Call/Callback — a promised call-back is not a reached person.
• Exception — "Lost Connection – Immediate Callback" → ELIGIBLE contact. Despite "Callback" in the name, this one is a contact (a live call that dropped). Do not sweep it into the non-contact callback bucket above. It can become a Net Contact if Talk > 30s.
• Auto → ineligible contact (system/automated outcome, not a live conversation).

If a new disposition appears that is not in this table, flag it (Section H2) before deciding a tier — do not silently bucket it.

---

Net-contact outcomes (v1.8): the Dispositions-tab "Net-contact outcomes" view counts each net contact under its canonical disposition name via the alias map — raw variants fold into one outcome. Rule: "Sale" = "Account Renewed" = "Account Renewed (Sale)" — one outcome. Same for all variants (e.g. "Interested/Not Ready" → "Interested not ready"). The only possible outcomes are the 7 eligible-tier dispositions.

Executive three-bucket grouping (v1.8): for summary presentation, net-contact outcomes group as — Positive (Account Renewed (Sale) + Will renew later + Interested not ready), Declined (Not interested), Lost mid-call (Hang up + Lost Connection); Decision maker not available = neutral (shown in the mix, outside the three buckets).

E. Inbound funnel — REDEFINED v1.7 (queue-based)

The funnel now measures service against the in-queue population (callers who genuinely waited), not against everyone who dialed in. This replaces the v1.2 "answered ÷ received" definition.

Column / stage order (funnel flow): Voicemails · Received · Response % · IVR drop · Abandoned · In Queue · Answered · Answer % · Inbound Sales · Conv %

Definitions:

• Voicemails left = outbound Voicemail dispositions with talk time > 0 (v1.9: zero-talk voicemail rows mean no message was actually left and are excluded from the base).
• Received (inbound_raw) = all inbound rows (includes disposition 0 / abandoned / IVR).
• IVR drop = inbound, disposition 0/blank, Abandon Time = 0 (left in the menu, never queued).
• Abandoned = inbound, Abandon Time > 0 (caller hung up).
• Quick hang-up = reached queue but < 10 s wait (queue_sub10) — barely entered the queue.
• In Queue (inbound_queue) = Received − IVR drop − quick hang-ups(<10 s) = callers who reached the queue and genuinely waited (≥ 10 s). Equivalent to "Total Queue Time ≥ 10 s".
• Answered (q_answered) = reached queue (≥ 10 s) and has a real disposition. This is a strict subset of In Queue. Note: distinct from inbound_answered (any inbound with a disposition, no queue gate), which is broader and must NOT be used as the Answer-% numerator (doing so produced impossible >100% rates).

Rates (all campaign/floor views — identical everywhere):

• Response % = Received ÷ Voicemails
• Answer % = Answered (q_answered) ÷ In Queue — how many of the callers who actually waited got answered. Always ≤ 100%.
• Abandon % = Abandoned ÷ Received
• Conv % = Inbound Sales ÷ Answered (q_answered) — consistent with the Answered column. Edge case: a sale from a sub-10 s call not in q_answered could rarely push this slightly over 100% on a tiny campaign; acceptable, flag if seen.

Threshold: the "reached queue" / quick-hang-up gate is 10 seconds (queue_min_secs), reused from the existing Abandonment analysis — no separate 30 s gate.

Day inclusion (v1.8): inbound views include a day if it had ≥20 outbound dials or any inbound activity (isInboundDay). Outbound/floor metrics keep the ≥20-outbound-dial gate (isDialDay). Rationale: weekends/partial days with inbound callbacks but light dialing were being dropped from inbound tables, under-counting queue/answered (e.g. ACA00009 137 vs true ~174). The inbound tab may therefore span more days than outbound tabs — by design.

Consistency scope (v1.7): these definitions are applied identically across the Command Center inbound funnel cards, the inbound trend chart, the inbound-by-campaign table, and the inbound-by-campaign-over-time table. Agent-level inbound views intentionally still use inbound_answered — queue fields (In Queue, q_answered, IVR drop) are computed per campaign only, not per agent, because a call that abandoned or dropped at IVR never reached any specific agent. Agent "answered" = what that agent actually handled.

Change needed — Daily Slides: adopt the same queue-based Answer %/In Queue/Conv % definitions for inbound so the deck matches Command Center.

---

F. Coaching & ranking (Command Center only — documented, not shared)

SPH-coaching flag: agent flagged when SPH < 70% of team-average SPH (team rate = total sales ÷ total hours). Agents with < 20 dials excluded from ranking. Agents below 5 qualifying calls get a low-confidence caution flag. True conversation (Engagement) = talk time ≥ 40s.

---

F2. Performance goals — NEW v1.7 (Command Center; documented for all)

Monthly floor targets, changed 2026-08-03. The three rate goals are a matched set (SPH and Conversion move together; CPS is derived from SPH).

Goal	Value	Prior (v1.6 era)	Notes
SPH	0.50	0.90	Sales per productive hour.
Conversion	11%	20%	Sales ÷ Net Contact. Set as the arithmetic pairing of the 0.50 SPH (0.50 ÷ ~4.5 net-contacts/hr ≈ 11%).
CPS	$33.50	$18.61	Derived: $16.75 ÷ SPH goal. Not an independent target — moves with SPH.

Display convention — "% to Goal" (not variance): the Goals tab shows percent to goal, not a raw +/- variance. For higher-is-better metrics %-to-goal = actual ÷ goal; for CPS (lower-is-better) = goal ÷ actual. So 100%+ always means "at or beating target" (green) for every metric, and below 100% is short (red).

Attempt cap = 6 (drives List Completion; see Section C). Changed from 8 on 2026-08-03. campaigns.csv also carries a per-campaign attempt_cap column (some legacy 5/8 values); the global default and all new campaigns use 6.

Goals may vary (v1.9 rule): targets can differ by month, by campaign, or by any other dimension the business defines. The monthly mechanism (effective_from) exists today; per-campaign goals are a requirement for the platform rebuild (see PRD §5).

These goals currently live in Command Center dashboard code, not a reference file — a future enhancement is to move them to reference so they roll monthly without a code edit.

---

G. Expected numbers after alignment

(Figures below are the v1.6 6-campaign baseline thru 7/22; they predate the Cycle-8 additions and the cap/goal changes. Update on the next full reconciliation run against current live data.)

Metric	6/29–7/21	6/29–7/22
Dials	~49,855	~52,734
Contacts (gross, incl. Auto + Invalid number + Not interested)	~14,702	~15,376
Net Contacts	~2,731	~2,877
Sales	259	269
Productive hours	~1,328	~1,426

Small dial differences (~30–50 calls) are input/extract noise, not a rule difference.

Reconciliation status (v1.6 — all three checked thru 7/22):

• Command Center: Dials 52,722 · Sales 269 · Contacts 15,374 · Hours 1,423.9 — matches (within noise). ✅
• Daily Slides: Dials 52,681 · Sales 269 · Contacts 15,360 · Net 2,876 · Hours 1,423.3 — matches (within noise). ✅
• Performance Review: Dials 52,710 · Sales 269 · Hours 1,424.7 — match; Contacts 14,864 and Net 2,464 are LOW by ~512 / ~413. Cause: it drops the 'Not interested ' disposition (trailing space) — 505 rows (412 with talk > 30s) in-window. Change needed — Performance Review: apply cleaning step 6 (trim disposition whitespace) so 'Not interested ' is caught; contacts and net contacts will then rise to ~15,376 / ~2,877.
• History: the old "~12,777" (Command Center/Daily Slides, pre-Invalid-number) and "~14,215/14,871" (a whitespace-buggy recount) are both superseded by the corrected ~14,702 / ~15,376.

---

H. Standing duties for every session (living document + anomaly reporting)

H1. Keep this master current

• Any change to a rule, definition, formula, threshold, rate, campaign scope, or list size is written into this master in the same session that makes it — never left as an undocumented tweak in one session.
• Notify Janina on any change so the other two re-sync. A rule that lives in only one session is a divergence, not an improvement.
• Log every change in the changelog (date, what, why, sessions affected) and bump the version.
• Re-read this master before a run so the session works from the latest agreed rules.

H2. Report data anomalies — now and ongoing

Actively watch for and flag (don't silently absorb): numbers that stop reconciling across sessions beyond input noise; new/unexpected dispositions not in Section D, or a sharp shift in disposition mix; new campaigns (ACA00005/07/08 going active) or a list size changing in List Health; corrupt/capped-out timers (frequent 90/30-min clamps, impossible durations); volume cliffs or spikes by day/agent/campaign; sub-30s "Sales" and other likely misdispositions; de-mislabeling drift (REN_PSX_30 count moving, prior-file coverage off); duplicate Call Ids, unparseable timestamps, or missing days in a new extract; test/pilot agents beyond the two known ones.

Rule of thumb: if a number looks surprising, say so in plain language with the figure and the likely cause — don't wait to be asked.

Open anomalies (carried):

• The 6/30 sara.villeda sale that existed in the 20260702 extract is absent from the current extract — escalate to whoever owns the export to confirm whether it was a real sale.
• Source data has trailing whitespace in disposition labels ('Voicemail ', 'Not interested '). Handled defensively via cleaning step 6 (trim before matching). Watch for new whitespace/casing variants in future exports; if a disposition's count suddenly drops to near-zero, suspect a label-format change first.

---

Changelog

Version	Date	Change	Sessions affected
v1	2026-07-24	Initial unified rule book: 7 alignment decisions, Auto = ineligible contact, List Health list sizes, standing duties (Section H).	All three
v1.1	2026-07-24	Performance Review aligned: weekends included, ACA00001 5,479→5,974, Auto moved to ineligible. Flagged the missing 6/30 rescued sale.	Performance Review; all (window/contacts note)
v1.2	2026-07-24	Inbound "Answered" = disposition alone (no ≥10s queue gate); Command Center rebuilt to match.	Command Center; Daily Slides
v1.3	2026-07-24	Rescued-sale rule removed (all land on 259/269). Contacts-tier work began — but the three diverged: Invalid number counted by Performance Review + Command Center, rejected by Daily Slides; Contacts figures split ~14,215 vs ~12,777.	All three (partial / conflicting)
v1.4	2026-07-24	MASTER reconciliation. Full 25-disposition tier table finalized (Section D). Invalid number = ineligible contact (locked, all three). Recall Request / Recall No Specific Time / General Callback / Dropped Call/Callback = not a contact. Lost Connection – Immediate Callback = eligible contact (explicit exception to the callback rule). Gross Contacts corrected to ~14,215 / ~14,871 via independent recount — supersedes the stale ~12,777. Changelog and version unified across the three copies.	All three
v1.5	2026-07-24	Renewed + Account Renewed = sales (aliases of Sale; distinct from the ineligible "Already renewed"). Kent Ingram and Paul Garcia added as excluded test/pilot agents; sara.villeda explicitly kept. Dedup confirmed = Call Id only (no change). Both agent- and alias-rules have 0 rows in the current extract, so numbers are unchanged — forward-looking guards. Change needed — Daily Slides & Performance Review: add the two sale aliases and the two agent exclusions (Command Center already has them).	All three
v1.9	2026-08-04	Post-review refinements (Janina's plain-English read-through). (1) Voicemails counted only with talk time > 0 — zero-talk VM rows (1%) excluded from the Response-rate base at BOTH counting sites. (2) Rule added: goals may vary by campaign / month / any dimension. (3) Time-cap clarification: over-cap time is counted at the cap (talk 90 min, others 30), never discarded; time above caps is now measured and shown on Data Quality split into plausibly-real (cap–2h) vs timer glitches (>2h) with billed-$ equivalent. On data thru 7/09 in-scope: 6.4 plausible hours ($107) vs 127 glitch hours — caps protect the numbers rather than forgo billing. Change needed — Daily Slides & Performance Review: adopt the VM talk-gate.	All three
v1.8	2026-08-04	Process + inbound-day fixes. (1) Inbound views use isInboundDay (≥20 outbound dials OR any inbound) — outbound keeps isDialDay; recovers inbound on weekends/partial days (Section E). (2) Inbound funnel + by-campaign table + trend now all aggregate from the same per-campaign source and the full queue field set (two display-aggregation bugs fixed). (3) Goals moved to reference/goals.csv (sph 0.50 / conversion 0.11 / cps 33.50, effective 2026-08-01); dashboard reads DATA.goals; CPS independently settable (blank = derived $16.75÷SPH); goal-month label dynamic. (4) Server boot-time rebuild: startup reprocesses the archived CSV with the current engine. (5) Footer shows data-refreshed timestamp + data-thru date (replaces filter-scoped call count). (6b) Net-contact outcome folding: outcomes keyed by canonical disposition via alias map; Sale = Account Renewed (one outcome). Change needed — Daily Slides & Performance Review: adopt the inbound day-inclusion rule and read goals from reference/goals.csv (do NOT hardcode).	All three
v1.7	2026-08-03	Major session update. (1) Attempt cap 8→6 — List Completion denominator (Section C); completion % rises floor-wide. (2) 7 new Cycle-8 campaigns ACA00010–00016 added (Section A.7), status "Not Received", size 0; ACA00007/00008 now in scope; total 15. (3) Cycle concept = launch month (Cycle 7 = 00001–00009, Cycle 8 = 00010–00016), assigned explicitly; program→cycle pairing for the new Cycle-Compare view. (4) Two more excluded agents — Federico Ramirez (+focalpoint@nextiva.com), Claudia Orozco (+focal@nextiva.com). (5) Goals changed (new Section F2): SPH 0.90→0.50, Conversion 20%→11%, CPS→$33.50 (derived); Goals tab now shows "% to Goal". (6) Inbound funnel redefined (Section E): In Queue = Received − IVR drop − quick hang-ups(<10s); Answer % = queue-answered ÷ In Queue; Conv % = sales ÷ queue-answered; applied consistently across all campaign/floor inbound views; agent-level unchanged. Change needed — Daily Slides & Performance Review: attempt cap 6, new campaigns/cycles, the two agent exclusions, goals, and (Daily Slides) the queue-based inbound definitions.	All three
v1.6	2026-07-24	Whitespace fix + contacts correction. Added cleaning step 6: trim disposition whitespace before tier matching — the source has 'Voicemail ' and 'Not interested ' with trailing spaces. Corrected gross Contacts to ~14,702 (7/21) / ~15,376 (7/22) and added Net Contacts (~2,731 / ~2,877); the prior ~14,215/14,871 were undercounts from the same whitespace bug. All three reconciled thru 7/22: Command Center ✅ and Daily Slides ✅ match; Performance Review is LOW on Contacts (14,864) and Net (2,464) because it drops 'Not interested '. Change needed — Performance Review: trim disposition whitespace (step 6).	Performance Review (apply trim); all three (corrected expected numbers)