Leona Group Schools · Marketing ↔ Data Team
How the report suite in the ALL SCHOOLS workbook counts students — which clock each report runs on and why, the counting rules, and the data it all depends on.
Adopted with the data team on 9/30/2026. The principle: what date makes sense depends on what the data is for.
Every student in Synergy carries two meaningful dates, and they answer different questions:
Add Date Time Stamp) — the moment the family converted. A family's engagement with us peaks here, not on the first day of school. This is the clock for marketing attribution: leads, ads, tours, conversion windows.Enter Date) — when the student is in a seat and the state starts funding them. This is the clock for counting new students, pacing, cohorts, and anything that should reconcile with enrollment reality.Why it matters: families often pre-register months before school starts. Colegio Petite's kindergarten families register in January–June for an August 1 start; the alt high schools pre-register all winter and spring. Counting those students at their registration date filed them into the previous school year — 34 of Colegio's fall-2024 kindergartners were invisible in the fall-2024 column before this change, and Alta Vista alone had 111 hidden fall starters. Synergy's own Original Enter Date field agrees: the student's entry is the day they start, not the day the paperwork was created.
| Report | Clock | Why |
|---|---|---|
| New Students Report · Enrollment Pacing · Monthly YoY · Dashboard new-student counts | Started | "New by this date" means students in seats. A student counts in the school year they actually start — so spring pre-registrations land in the fall they attend, and a record created with no enrollment yet counts in no year until the student starts. |
| Marketing Retention · Marketing ROI · Lead Insights · tour reporting | Pulled in | Marketing attribution — credit belongs to the moment the family was won. Lead windows, conversion rates, and spend joins all key on lead/record creation dates. |
| Monthly YoY (planned toggle) | Either | A March registration for an August start is a March recruiting win and an August enrollment. A control cell will show either view; until the script update installs, the live tab still counts on the pulled-in clock. |
| Cohort Retention · Retention Insights · attrition reports | Started | Retention clocks run from when a student actually attends. Departure rules are unchanged; cohort entry follows the start clock. |
| Marketing Retention's In-Session Days · 90-day blocks | Both | The lead side uses the marketing clock; the day count deliberately borrows the funding clock (Enter Date + each school's calendar) so blocks track what the state pays for. |
Preschool never counts as new — and that's a funding statement. PS isn't government funded, so a preschooler isn't a funded enrollment yet. The student becomes New to Leona in the year they start KG (or their first non-PS grade) — the first time the state funds them. PS rows are excluded from every new-student count; they don't create newness and a PS year doesn't block it later.
Students who enroll and later withdraw still count as new. The enrollment happened; the retention and attrition reports carry the drop. New-student counts never look at leave codes — codes only matter for attrition.
Rollovers and returners never re-count. A student counts once, at their first-ever funded enrollment across every year of data. A record created this year for a student with any prior-year enrollment is a returning student, whatever the code says.
What changed on 9/30/2026: new-student counts moved from the pulled-in clock to the started clock, and PS was excluded. Every "new by date" level rose (the pre-registered students surfaced in the right years, consistently, so year-over-year comparisons stay apples-to-apples), and the SY26-27 pacing read moved from slightly behind to slightly ahead. Documents dated before 9/30 use the old rule.
The script installed 9/30/2026. Each report picks up the new rules on its next RUN. One cause behind every shift: students count when they start, and PS never counts.
| Report | What you'll see change | Why |
|---|---|---|
Monthly YoY | Fall months roughly double historically; spring months shrink; KG entry columns rise at elementaries; a new Count by dropdown appears in B5. | Spring registrations move into the fall they attended; PS→KG kids now count as new KG. The B5 toggle keeps the old registration view one click away. |
Enrollment Pacing | Every year's level rises ~450–500 district-wide; SY26-27 reads ahead of last year instead of behind. | The pre-registered fall starters surface in the year they attend. |
New Students Report | Fall date-ranges gain students; spring ranges lose them; PS disappears. | Same clock change, applied to whatever date range is set. |
Cohort Retention | Cohorts re-form around start months; early checkpoints read slightly higher. | The retention clock no longer burns days between registration and the first day of school. |
Retention Insights · Attrition Patterns · Lead Insights · Event Planner | Tenure and survival figures tighten slightly. | All clocks now start at the first day in a seat. |
Attrition · Attrition YoY · Summer vs In-Year | No change. | The LOST rules were never touched. |
Marketing Retention · Marketing ROI | No change. | They run on the pulled-in clock on purpose — attribution belongs to when the family was won. |
Reference restatement — Monthly YoY, August New-to-Leona, all schools (the numbers to expect after the first rerun):
| Count by | Aug 2023 | Aug 2024 | Aug 2025 | Aug 2026 |
|---|---|---|---|---|
| Started (default) | 1,866 | 1,760 | 1,253 | 1,558 |
| Pulled in (B5 toggle) | 861 | 790 | 571 | 572 |
Why the change makes the analysis better:
• Pacing is an early-warning system now. Roughly 500 real fall starters per year were invisible in the current-year column because they'd been counted at spring registration — budget decisions were steering against a number that couldn't move when families actually showed up.
• Seasonality is true. The old curve's "spring enrollment" was paperwork. The restated curve shows when families arrive — and it immediately surfaced real intel: August intake dipped to 1,253 in 2025 and recovered to 1,558 in 2026.
• "Registered but never started" is now a countable group — the warmest re-engagement audience there is.
• PS→KG is a measurable funnel: preschool families converting to funded KG now appear as new-student wins in the KG year.
• Retention clocks are honest, so cohort checkpoints and event timing reflect real at-risk windows, not registration artifacts.
• One language with the data team: new-student counts now reconcile with the ADM view by construction, while the Monthly YoY toggle still answers "when did we win them?" whenever the marketing view is the right one.
Used by the New Students Report, Monthly YoY, Enrollment Pacing, and the entry-grade columns. Runs on the Started clock.
The rules in words. Merge every enrollment row across every Raw Data tab, group by student, and read each row's Enter Date as the start of an enrollment spell. Drop PS spells entirely. The earliest remaining start date is the student's New-to-Leona moment: it decides which school year they count in, which school gets credit (the first one), and their cohort. If Synergy's Original Enter Date is more than 45 days before the earliest visible spell, trust it — the student has history older than the archives (a ghost returner or late-processed record).
What never changes the answer: leave codes (new-student counts ignore them entirely), a rollover record created for the new year (returning student), duplicate spells across tabs (merged by student + date), or a withdrawal after starting (the student still enrolled; retention carries the drop).
Day-true pacing. Because schools start on different dates each year, the pacing workbook also compares years at the same number of school days in, using each school's own Linear School Days calendar — "by day 40" is honest across years in a way "by Sept 23" isn't.
Used by the Attrition, Attrition YoY, and Summer vs In-Year reports and the Event Planner. Everything keys off the student's final record.
Session boundaries. A school's year runs from its calendar first day to its last day (the Linear School Days tabs; estimated from earliest Enter Date for years without one). Anything dated outside that window — plus every No Show — lands in the 7/1 Summer bucket; everything inside counts In-Year on its actual date. Recent months are always provisional: a "lost" student who re-enrolls drops out of the count on the next data refresh, by design.
Who gets the credit. A New-to-Leona student is credited to the first school they joined (their entry grade counts there too); every later campus counts them as New to School, which is what the Transfers In column captures. A Lost-to-Leona student is credited to the last school they left. An internal move debits the leaving school (Left School) and credits the receiving one (New to School), netting out as Moved Within Leona — so school-level columns can move while network totals don't. Marketing-source attribution (the "how did you hear about us" tracker, ad platforms) is a separate system and deliberately not part of these counts.
The Marketing Retention report connects GoHighLevel leads to Synergy enrollments and tracks how long each marketing-sourced student stays — counted in in-session days, the clock the state funds. Runs on the Pulled in clock by design: attribution belongs to the moment the family was won.
Reading it honestly. Matching is name-based, so the Match column states the evidence per row (student name · parent last name · manual · other school). A first-name-only hit counts only with last-name evidence — the student's last name appears somewhere in the lead's or contact's name, or the lead's last word sits inside a compound last name ("Jazmine pacheco" → Castillopacheco); a first-name match whose last name simply disagrees ("juan rangel" → Juan Cuervo) is not a match. Ambiguous leads go to a review list that does the homework: each row shows the tied candidates — Perm ID, name, grade, start date — plus hints when the lead's stated grade fits exactly one candidate, or when every candidate started the same day (likely one family enrolling together). One paste in GHL Match Overrides settles it permanently: the right Perm ID, several comma-separated Perm IDs when one inquiry enrolled several siblings (each child counts), or SKIP. Nothing merges automatically — the candidate pool is all of Synergy, not just GHL leads, so the human ruling is what keeps the report GHL-only. "No enrollment found" is mostly families who didn't convert — the denominator, not an error. And the report covers GHL-touched students only: organic enrollments never appear, so it reads as "retention of the marketing pipeline," never "marketing caused all growth."
The day clock. Each row carries three dates: First Lead (GHL), Enrolled (the Synergy record's creation — the conversion), and Started (the Enter Date — the first day of actual membership), plus the school's own School Start for this year. In-Session Days count actual school days — each date on the school's own Linear School Days calendar, so weekends and breaks contribute nothing — from Started (never before the school's first day) to the leave date or today. A family that converted in February for a July start earns zero days until school opens, which is exactly how funding sees it, and a student there since day one shows the true count of school days held so far. 90-Day Blocks divide that count. (Fixed 10/28/2026: the count previously included weekends and breaks inside the session window, so day totals restate lower — a full school year is ≈180 school days, not ≈290 calendar days.) Leona Connected is year-round, so its students count straight days from Started with no calendar clamp. Schools appear fully expanded on each run; collapse them with the −/+ margin controls.
Tour reporting uses the same spine: tour signups come from GHL (the Tour Pipeline plus written-off leads whose student enrolled anyway; duplicates for the same enrolled student count once), and "enrolled" always means a verified Synergy match — never a pipeline stage. Leona Connected is excluded (online school, no campus tours).
All tabs live in the ALL SCHOOLS workbook. Reports find them by name pattern, hidden or not.
| Source | What it provides | Maintained by |
|---|---|---|
Raw Data + Raw Data YYYY-YYYY | Enrollment rows (one per student, per school, per year). The current export only carries history for recent students, so the archived year tabs preserve the students who have since left — both are always read together. | Data team / archive button |
Naming Normalcy | Which schools count (scope), their abbreviations (used to match the calendar tabs), and their group (Elementary‑Middle / ALT HS / TRAD HS sections). | Data team |
HS Grads YY-YY | Graduate Perm IDs — excluded from all attrition counting, whatever their records say. | Marketing (yearly paste) |
SY## Linear School Days | Each school's calendar — every school day it holds, per year. Defines "in session", powers day-true pacing, and is the day list behind Marketing Retention's In-Session Days (currently SY25-26 and SY26-27; earlier years are estimated from the earliest Enter Date). | Data team |
GHL Enrollments | Raw GoHighLevel opportunities export (all pipelines, all time, list view). The lead side of Marketing Retention, Cohort Retention's % GHL, Lead Insights, and Marketing ROI. | Marketing (monthly paste) |
GHL Match Overrides | Permanent per-lead rulings: Opportunity ID → Perm ID, or SKIP to exclude a bad match. Always wins over the automatic matcher. | Marketing (as needed) |
Ad Spend | Google Ads "Billing per campaign" rows (Month · Campaign · Cost). Campaigns map to schools automatically; Support Office and closed campaigns are dropped; Leona Connected Texas is kept separate. Feeds Marketing ROI only. | Marketing (monthly paste) |
The columns that matter — matched by header name, so columns can move but must not be renamed:
| Column | Role in the logic |
|---|---|
Perm ID | Student identity. All rows across all tabs merge on it (numeric vs. text formats are normalized). |
Enter Date | The Started clock — the first day of actual membership. Decides the school year a student counts as new, anchors Marketing Retention's In-Session Days, and defines enrollment spells. |
Add Date Time Stamp | The Pulled-in clock — when the enrollment record was created, i.e. when the family converted. Marketing attribution keys on it; it no longer decides which year a student counts as new. |
Original Enter Date | Cross-check on "new": if it's more than 45 days before the first visible spell, the student has history older than the tabs, and this date wins. |
Original Enter Grade (M) | The grade a student entered Leona at — source of the KG / 6th / 9th entry columns (PS entrants take their first funded spell's grade instead). |
Grade (O) | The leaver's grade — used for the grade-12 and top-grade completion checks in attrition. Also identifies PS spells (in the year-scoped archives). |
Leave Date + Leave Code | The "lost" clock and the reason (same W-code taxonomy as the weekly tracker's Key Information). Never consulted by new-student counts. |
FullYear / Organization Name | The row's school year · which school gets credit. |
Cohort Retention plus the Enrollment Metrics → Insights submenu. All reuse the exact NEW / LOST rules, so their numbers always reconcile with the core reports.
| Report | The question it answers | Worth knowing |
|---|---|---|
Cohort Retention | Of the students who enrolled in month X, how many were still enrolled 60 / 120 / … / 360 days later? | B3 = start year, B4 = cohort width: 1 lines up month-for-month with Monthly YoY's New to Leona; 2 smooths small schools. Checkpoints only count students old enough to be judged. |
Retention Insights | Which entry grades stick, how long do leavers stay, who comes back, and how much loss is silent? | Win-backs are real: roughly 30% of leavers re-enroll somewhere in Leona. "Silent" losses are the July-1 non-returns — no withdrawal was ever processed. |
Attrition Patterns | Where do school-changers go, why do students leave, and when in each school's calendar? | The transfer matrix reads row → column. Calendar timing uses each school's own Linear School Days first day. |
Enrollment Pacing | Is each school ahead of or behind last year at this same date? | Runs on the started clock: cumulative first-starts from July 1. The day-true view compares years at the same school-day count. The Δ column is the early-warning flag. |
Lead Insights | Which pipelines and schools produce leads that convert and stay — and how fast? Do GHL stages predict reality? | "⚠ enrolled anyway" marks dead-looking stages holding real enrollments — the proof behind never stage-filtering the export. |
Marketing ROI | What does an enrolled — and a 180-day-retained — student cost, per school? | The only tab with dollars (retention elsewhere stays days-only). Spend joins on the lead's created month; a partial month reads artificially expensive until it settles. |
Four kinds of tab. Report tabs are rebuilt from scratch on every run — never type into them, because anything hand-entered is erased on refresh. The yellow control cells and the paste tabs are the only intended inputs.
| Kind | Tabs | Rules |
|---|---|---|
| Reports (script-built) | Enrollment Metrics Dashboard · Enrollment Metrics Data · New Students Report · Monthly YoY Report · Attrition Report · Attrition YoY Report · Summer vs In-Year Attrition · Retention Event Planner · Marketing Retention · Cohort Retention · Retention Insights · Attrition Patterns · Enrollment Pacing · Lead Insights · Marketing ROI | Each has a RUN button and a menu item. Yellow cells (month, years, start year, cohort width) persist across runs and are the only cells meant to be edited. |
| Sources the reports read | Raw Data · Raw Data YYYY-YYYY archives · Naming Normalcy · SY## Linear School Days · HS Grads YY-YY · Linear Dates · Block Start and End Dates · Historical ADM Reference | Read-only to the script (the last three serve the dashboard/ADM side). |
| Inputs you paste | GHL Enrollments · GHL Match Overrides · Ad Spend | Marketing-maintained; cadence on the Upkeep tab. |
| Data-team tabs | ADM Logic Preview · Projected Gross ADM History · In-Progress Enrollments · Trends · Potential Grads by Block Counts · Projected vs Real ADM… and other working tabs | Shannon's funding-view side. The reports never write to them — except Projected Gross ADM History, which receives a snapshot when the menu's capture button is used. |
The data team’s “26-27 Enrollment Data” workbook tracks enrollment activity by week; these reports count students. Same Synergy export underneath, two deliberate lenses — here’s how to read one against the other so the numbers never look like a disagreement.
The reassuring part first: both systems follow the same two-clock agreement, arrived at independently. Every event in the weekly tracker is bucketed into the school year it belongs to (Synergy’s FullYear — the Started side) but plotted on the calendar week its record was created (the Pulled in side). That’s why the tracker’s “26-27” weeks begin in January 2026: a family who registers in spring belongs to the fall they registered for — the same fix the pacing reports made. The tracker also locks each finished year’s results so they can’t drift as the export ages; these reports get the same protection from the hidden Raw Data YYYY-YYYY archives. Cross-checks between the two are fair game.
| Weekly-tracker term | What it counts | Closest term here | Why the totals differ |
|---|---|---|---|
| Rollover Beg of Year (code E1) | Returning students re-registered for the new year | Nothing — deliberately | A rollover never makes a student new. The tracker’s “Total Added” includes rollovers, so it will always dwarf any new-student count on this side. |
| New this Year (E3, E4, E6…) | First entry to that school year, per school | New to School (per-year) | The tracker re-counts a student each year and includes moves between Leona schools. New to Leona counts a student once per career, at their first funded (non-PS) arrival — always the smaller, marketing-true number. |
| Re-enrollments (R1) | Left and came back within the year | Re-enrolled (R1) columns on the Attrition & Attrition YoY reports (added Oct 2026) | Same code, same definition — adopted from the tracker because it’s a retention signal worth watching (~580–600 students a year district-wide). The tracker places the event on its add-stamp week; this report counts it on the return’s Enter Date. Expect agreement within a few percent, not to the digit. |
| Withdrawals (by W-code) | Coded withdrawals, by week | Left School / Lost to Leona | These reports also count year-completers who never return (the July-1 “No Show” bucket — roughly half of true annual attrition) and treat W6/W7 as graduates; the tracker counts coded exits only and files W6 under Others. Its withdrawal columns read lower than real attrition — by design, not by error. |
School ≠ ALL SCHOOLS first, or use the per-school tabs.
Which to reach for. Week-to-week campus activity, enter/withdraw code detail, this-week-vs-same-week-last-year — the weekly tracker. Marketing attribution, funding views (in-session days, 90-day blocks), New-to-Leona YoY, pacing, and anything tied to GHL — these reports. A worked example of both being right at once (October 2026): the tracker’s 26-27 “New this Year” stood near 3,000 while New to Leona over the same stretch was 2,272 — per-year school intake versus first-time-to-the-network families. Different questions, one dataset.
The future-reference checklist. Each item names the dependency and what to do when it changes.
Leave Date, Add Date Time Stamp, Original Enter Grade, etc. silently drops that field. If a Business Objects query changes, keep the headers.Raw Data YYYY-YYYY tab; a missing archive silently shrinks history. (The reports toast a reminder each Jul–Sep.)HS Grads YY-YY tab.Paste the graduate query (Perm ID column keeps its header). Any tab starting with "HS Grads" is honored automatically.SY## Linear School Days tab.Same linear format — header row of school abbreviations, dates below. Years without one fall back to the earliest-Enter-Date estimate (which the SY26/SY27 calendars confirmed matches in practice).GHL Enrollments paste.Replace the tab's contents with a fresh all-pipelines, all-time opportunities export (list view). The report only knows what's pasted; recent cohorts firm up as leads and enrollments accumulate. Resolve review-list rows once in GHL Match Overrides — a Perm ID to force a match, SKIP to kill a bad one; both stick forever.Ad Spend tab.Add the newest month's Google Ads "Billing per campaign" rows under the existing ones (Month · Campaign · Cost — the export's own header lines are ignored). Campaign names map to schools automatically; anything unmappable shows under "Spend not tied to a school" until it's added to CAMPAIGN_ALIASES in the script. Support Office and closed campaigns are dropped; Leona Connected Texas stays its own school.Add Date Time Stamp) — the marketing clock.Enter Date) — the enrollment clock; decides the year a student counts as new.