📄 Table of Contents
- Data Foundation
- Season Date Window
- Playing Member Filter — Inclusions & Exclusions
- Member Scopes: All Active vs Registered Only
- Association Roll-up (Clubs & Schools → Parent Association)
- Deduplication — One Person, One Count
- Growth Dashboard
- Retention & Churn Dashboard
- YTD Comparison Dashboard
- Revenue Performance Dashboard
- Association Deep Dive Dashboard
- Participation & Revenue Dashboard
- Analytical Dashboards (v2) — Competitions, Demographics, First Nations, FN Penetration, Churn Deep Dive, Market Penetration, Opportunity
- Full Glossary of Terms
1. Data Foundation
All dashboards pull from the Datawarehouse, accessed via the prod_marts schema in PostgreSQL. No data is manually entered or estimated — every number traces back to live registrations and payment transactions recorded by Netball Connect.
This guide and all dashboards are current to the date below — they include every registration and payment up to and including that date, and nothing after it. The date is the latest data refresh (export) from Netball Connect:
📅 Data current to 16 June 2026Primary Tables & Views
| Object | Type | What it contains |
|---|---|---|
| prod_marts.fact_registrations | Table | One row per registration event. Contains the person's user ID, their affiliate name (association, club, or school they registered through), the product they registered for, payment status, registration date, fees paid, and season year. |
| prod_marts.dim_affiliates | Table | Reference table for all NQ affiliates (associations, clubs, schools). Contains the organisation type (Association / Club / School) and the affiliated_to field that tells us which parent association a club or school belongs to. |
| prod_marts.dim_users | Table | User dimension. The dashboards read only the date of birth (to bucket members into age bands) and the is_valid_dob flag. No member names, emails, mobiles, or dates of birth are ever displayed on a dashboard — outputs are counts and aggregations only. Joined to fact_registrations via user_sk when age-group breakdowns are needed. |
| prod_marts.fact_payments | Table | One row per payment transaction line item. Contains transaction_date, fee_type, net_settlement, source_account (primary / social / hpl), and organisation fields. Used exclusively by the Revenue Dashboard. |
| prod_marts.pbi_v_playing_growth | View | Pre-aggregated season summary per year and per member scope (All Active / Registered Only). Columns: season_year, status_scope, total_members, returning_members, new_members, churned_members, net_change, growth_pct, retention_rate_pct, growth_quota_3pct. Used by Growth and Investigation Report dashboards. |
| prod_marts.pbi_v_playing_growth_assoc | View | Same as above but broken down to association level. Carries both total_members (HOME attribution) and players_served (touchpoint). Used by the Growth Dashboard for the association-level charts. |
| prod_marts.int_playing_registrations | Table (intermediate) | Canonical playing-registration grain — one row per qualifying playing registration after the full playing-member filter (product inclusions/exclusions, payment-status rules, and the role-aware Team Registration rule). Carries season_year (= platform registration_year), external_user_id, the rolled-up association, club_name, and product_cat. Every membership dashboard now derives from this one model, so the playing-member definition is applied identically everywhere. |
| prod_marts.int_player_home_association | Table (intermediate) | One row per (status scope, season, person) resolving each member to a single HOME association (see HOME vs SERVED below). Also flags is_outside_window (registered outside the announced Nov–Oct dates) and the season month of first registration. Source of truth for state and association headline counts across Growth, Retention, YTD and Deep Dive. |
external_user_id) is counted only once per season, regardless of how many registrations or transactions they have on file. external_user_id is an internal, non-identifying key — it carries no name, email, or date of birth.
2. Season Date Window
A "season" in Netball Queensland does not match the calendar year. The announced season window is:
In general terms: Season Y = 1 November (Y−1) to 31 October (Y). The season is named after the year it ends in. This window is used for the season-month numbering on the chart X-axes (see table below).
How a registration is assigned to a season
registration_year label — NOT by filtering on the registration date.
The Nov→Oct window above is an announcement. Real club behaviour varies around it:
- Some clubs open registrations early (before 1 November).
- Some run late or schedule "one last game" after 31 October.
- Non-playing programs (coaching, officiating, volunteer, non-playing memberships) are confirmed to run into the next window as well.
Because of this, a registration's date can legitimately fall outside the announced window. The season a member belongs to is taken from the platform's registration_year label, not from the registration date — so these members stay in the right season. For example, someone who joins the 2025 season on 1 November 2025, just after the announced close, still counts in Season 2025.
Season Month Numbering (YTD & Deep Dive dashboards)
When charts show "month 1, 2, 3…" along the X-axis, they are using season month numbering, not calendar months:
| Season Month | Calendar Month | Notes |
|---|---|---|
| 1 | November | Start of season — early registrations, schools |
| 2 | December | |
| 3 | January | Registration surge begins |
| 4 | February | |
| 5 | March | |
| 6 | April | Peak registration month for most associations |
| 7 | May | Current month as of this report |
| 8 | June | |
| 9 | July | |
| 10 | August | |
| 11 | September | |
| 12 | October | End of season |
| Outside window | Any date before 1 Nov (Y−1) or after 31 Oct (Y) | Late "one last game" / late-admin registrations still labelled this season by registration_year. Shown as a separate bucket so months 1–12 + Outside window = the full season total. |
The "Outside window" count is small but real — e.g. 50 playing members in Season 2025 (all late, dated November 2025). Per-season figures: 2021 = 58, 2022 = 34, 2023 = 60, 2024 = 40, 2025 = 50, 2026 = 0 (season still mid-window).
3. Playing Member Filter — Inclusions & Exclusions
Not every record in the registrations table represents a playing member. NQ has a carefully defined filter to identify genuine playing participants and exclude non-playing registrations (administrators, carnival-only participants, etc.) and failed/cancelled transactions.
What IS included (Product inclusions)
A registration is included as a playing member if the membership product name matches any of the following patterns:
| Pattern | Typical product names covered | Category |
|---|---|---|
| ILIKE '%player%' | Player, Player - Social, Player - 10 Week, Player (Senior), etc. | Player |
| ILIKE '%net%set%go%' | NetSetGo, Net Set Go Membership | NetSetGo |
| ILIKE '%team registration%' or '%team nomination%' | Team Registration, Team Nomination | Team Registration |
| ILIKE '%10 week%' | Player - 10 Week Program, 10 Week Season | Player - 10 Week |
| ILIKE '%social%playing%' | Social Playing Membership, Player - Social Playing | Player - Social |
| ILIKE '%season%membership%' | Season Membership (10-week style product) | Player - 10 Week |
| ILIKE '%little%nett%' | Little Nettys (early years program) | Little Nettys |
| product = 'Annual' AND type IN ('Player', 'Player - NetSetGo', 'Player - Walking Netball') | Annual membership with a player-type subtype — used in some jurisdictions | Player / NSG / Walking |
What is EXCLUDED (Product exclusions)
Even if a product name hits one of the inclusion patterns above, it is immediately excluded if it also matches any of these:
| Exclusion rule | Why excluded |
|---|---|
| ILIKE '%non%player%' | "Non-Player" products — administrators, bench officials, coaches who are not playing members. |
| ILIKE '%non%playing%' | "Non-Playing" products — same rationale as above. |
| ILIKE '%carnival%' | Carnival-specific registrations. These are event-only, not regular season memberships, and would inflate counts. |
| membership_product_type IN ('Carnival/Event', 'Event/Rep Trials', 'Event/Carnival') | Product type-level exclusion for event/rep trial registrations even if the product name itself looks like a player product. |
What is EXCLUDED (Payment status exclusions)
A registration is excluded if its payment_status is one of the following — these indicate the registration did not result in a valid active membership:
| Excluded status | What it means |
|---|---|
| De-Registered | Person has been fully de-registered from the system. No longer a member. |
| Approved De-Registration | De-registration has been approved — equivalent to De-Registered. |
| Cancelled Registration | Registration was cancelled before it was completed. |
| Incomplete Registration | Registration was started but never finalised. No payment received. |
| Failed Registration (with paid_fee ≤ 0) | Registration failed AND no payment was collected. Failed registrations with a partial payment (e.g. paid fee > 0) are still included, as the person has some financial commitment recorded. |
paid_fee_inc_gst ≤ 0 are removed. If a failed registration still has a positive paid fee on record, it is counted — this covers scenarios where a registration failed at a late stage but payment had already been processed.
Role-aware exclusion on Team Registration / Team Nomination products
A Team Registration / Team Nomination row can be created by any one of several people on behalf of a team:
- the player themselves (or their parent on their behalf),
- the team coach or manager,
- a team registerer — typically a club admin or committee member who pays the team-entry fee on behalf of the squad.
Only the first of these is a playing member. The others are administrative roles whose registration is purely to enter the team; they are not playing the sport themselves. Counting them as players would overstate participation by roughly 200–650 people per season.
The source data carries a registration_role column — values are Player, Coach, Manager, Umpire, Team Registerer, Volunteer, Committee, Team Official, or Program Registrant. The role-aware filter applies the following rule:
registration_role = 'Player'. Admin roles holding only a team-product row are excluded.
How many people this affects each season (admin-only registrants excluded):
| Season | Team-only registrants | Player role kept | Admin roles excluded |
|---|---|---|---|
| 2023 | 656 | ~394 kept | ~262 excluded |
| 2024 | 195 | ~132 kept | ~63 excluded |
| 2025 | 421 | ~213 kept | ~208 excluded |
| 2026 | 421 | (in progress) | (in progress) |
4. Member Scopes: "All Active" vs "Registered Only"
The Growth Dashboard and Investigation Report show two separate lines/figures representing the same pool of playing members under different payment status lenses. These are called status scopes.
All Active (the primary headline number)
This is the widest, most inclusive view of participation. It includes all payment statuses that were not explicitly excluded — meaning it includes Registered, Pending Payment, Awaiting Payment, Conditional Registration, and any other non-excluded status.
2026 season figure: 58,498 (live count, as at 29 May 2026 — grows as the season progresses)
Use All Active when you want to understand the full breadth of participation — every person who came through the system and has a valid playing membership on file, regardless of whether they have fully paid yet.
Registered Only (the conservative count)
This is the most conservative view. Only people who have reached the "Registered" final status are counted. Pending payments, conditional registrations, and awaiting approvals are excluded.
2026 season figure: 52,412 (live count, as at 29 May 2026)
Use Registered Only when you want a fully committed, payment-confirmed member count.
The difference explained (live counts, as at 29 May 2026)
| Scope | 2026 Count | What's in the difference (6,086) |
|---|---|---|
| All Active | 58,498 | Includes Pending, Awaiting, Conditional, etc. |
| Registered Only | 52,412 | Payment-confirmed only |
| Difference | 6,086 | Members in the pipeline — partially registered or awaiting payment confirmation |
5. Association Roll-up (Clubs & Schools → Parent Association)
Netball Connect records registrations against the specific club or school that the member registered through — for example, "Mitchelton Netball Club" or "Brisbane State High School". However, for association-level reporting, these must be rolled up to the parent association (e.g. "Metro North Netball Association").
How the roll-up works
The dim_affiliates table contains an organisation_type column (see Organisation Types & Hierarchy below) and an affiliated_to column that names the parent association. The roll-up logic maps clubs and schools up to their parent association:
- If the affiliate is a Club or School (hierarchy level 5): use the
affiliated_toparent association name. If that field is blank, fall back to the raw affiliate name. - If the affiliate is an Association or League (hierarchy level 4) or higher: use the affiliate name directly.
Organisation Types & Hierarchy Levels
The dim_affiliates table (sourced from staging.stg_affiliates) classifies every NQ affiliate by organisation type and assigns a numeric hierarchy level. The hierarchy reflects the governance structure of Netball Queensland.
| Hierarchy Level | Level Name | Organisation Types | Description |
|---|---|---|---|
| 1 | National | National | Netball Australia — the national governing body. |
| 2 | State | State | State/territory member organisations, e.g. Netball Queensland. |
| 3 | Region | Region | Regional bodies sitting between the state body and local associations. |
| 4 | Association / League | Association, League | Local associations and competition leagues. Both organisation types are assigned level 4 and treated identically in the roll-up logic. |
| 5 | Club / School | Club, School | Individual clubs and schools that register members. These roll up to their parent Association/League via the affiliated_to field. |
| 6 | Other / Unknown | Unknown, other | Affiliates where the organisation type could not be determined from the source data. |
The Association Deep Dive dashboard groups all data at hierarchy level 4. This means Association and League are treated as the same level and will both appear as selectable items in the association dropdown. If an organisation is classified as a "League" in the system, it will behave in every way like an "Association" in the dashboard — they share the same roll-up bucket and their member counts are aggregated together under the same label.
Member attribution: HOME vs SERVED (touchpoint) key concept
A single member can register through more than one association in the same season — for example, winter netball at one association and a summer social competition at another. When we report "members per association", we must decide how to attribute that person. There are two valid answers, and the dashboards use both deliberately:
| View | Rule | Does it sum to the state total? |
|---|---|---|
| HOME single-home |
Each member is attributed to exactly one association per season — the one where they have the most playing registrations. Ties are broken by the most recent registration date. A person is never counted twice. | Yes. Σ HOME across all associations = the state headline total, exactly. No double-counting. |
| SERVED touchpoint |
A member is counted at every association they registered with that season. Someone who played at two associations is counted once in each. | No. Σ SERVED > the state total. The excess equals the number of cross-association registrations. |
Σ HOME = state total (always), and Σ SERVED − state total = members who registered at 2+ associations. If an association page shows SERVED above HOME, the gap is exactly the members it shares with other associations that season — not an error.
Which dashboard uses which
| Dashboard | Association headline | Secondary / reconciling |
|---|---|---|
| Growth (association charts) | HOME (total_members) — sums to the state KPI | SERVED shown as players_served |
| Retention (association) | HOME — keeps the member universe identical to Growth | — |
| YTD (association table) | SERVED (matches Netball Connect at the same date) | HOME columns reconcile the table total to the state KPI |
| Association Deep Dive | SERVED (matches what the association sees in Netball Connect) | HOME secondary + the Outside-window bucket |
Both views are computed once in prod_marts.int_player_home_association (HOME = the resolved row per person; SERVED = every qualifying registration from int_playing_registrations), so the definitions are identical across every dashboard.
6. Deduplication — One Person, One Count
A single person can appear in the datawarehouse multiple times within one season — for example, if they register for both a regular Player membership and a 10-week program, or if they transferred between clubs. The dashboards always count each unique person only once per season.
Technical implementation
The SQL uses COUNT(DISTINCT fr.external_user_id) or a DISTINCT ON (season_year, external_user_id) CTE to ensure each user is counted once. The earliest (minimum) registration date within the season window is used as the "first registration" date for that person in that season.
7. Growth Dashboard
File: growth_dashboard.html · Script: scripts/generate_growth_dashboard.py
Answers the question: "Are we growing? Did we hit our 3% growth quota? Which associations are growing and which are declining?"
Data sources
prod_marts.pbi_v_playing_growth— season-level totals for both scopesprod_marts.pbi_v_playing_growth_assoc— same but broken down by association (All Active scope only)
Year Slicer
The dashboard has a row of year buttons at the top (2021–2026). Clicking a year updates the KPI cards, waterfall chart, and association bar chart for that selected year. The full-history trend chart and scope comparison chart always show all years but highlight the selected year.
KPI Cards (top row)
| Card | What it shows | Data field |
|---|---|---|
| Total Active Members | Total unique playing members for the selected season year (All Active scope). | total_members |
| Net Change | Total members minus prior year total. Positive = growth, negative = decline. | net_change |
| Growth % | Net change divided by prior year total, expressed as a percentage. | growth_pct |
| Retention Rate | Percentage of the prior year's members who re-registered in the current season. | retention_rate_pct |
| 3% Quota Status | Shows "MET" or "MISSED" — whether growth reached the NQ target of 3% above prior year. Also shows the numeric target (prior year × 1.03). | growth_quota_3pct |
Bar chart showing total All Active members per season from 2021 to 2026. Bars are red if the 3% quota was missed that year, blue if met. The dotted orange line shows the 3% growth target for each year (calculated as prior year × 1.03).
- X-axis: Season year
- Y-axis: Unique member count
- Selected year bar is highlighted by the year slicer
Shows how the total for the selected year is built up from the prior year. Bars represent: Prior Year base → Returning members kept → New members gained → Churned members lost → Current Year total.
- Green bars = positive contributions (returning, new)
- Red bars = members lost (churned)
- Blue bars = totals (prior year, current year)
Two overlapping line charts across all years. The solid blue line is All Active, the dotted orange line is Registered Only. This lets you see the size of the "in-pipeline" membership (members not yet fully confirmed) each year.
- The gap between the two lines = members pending payment confirmation
- Both lines use the same playing member filter; only the payment status scope differs
Horizontal bar chart showing all associations ranked by total members for the selected year. Bars are coloured by quota status (blue = met 3% target, red = missed, grey = no prior year to compare). Hover over a bar to see exact members, prior year, net change, and quota status.
- Updated dynamically when a year is selected
- Shows up to 95 associations
Two side-by-side horizontal bar charts: the top 10 associations by net gain and the top 10 by net decline for the selected year. Useful for quickly identifying standout performers and areas needing support.
- Net change = current year members minus prior year members
- Only associations with a prior year to compare are shown
Key metrics defined
| Metric | Formula |
|---|---|
| Returning members | Members in season Y who also appeared in season Y−1 (matched by external_user_id) |
| New members | Total members in season Y minus returning members = members not present in season Y−1 (first-time, or returning after a gap) |
| Churned members | Members from season Y−1 who did NOT register in season Y = people lost |
| Net change | New members minus churned members (equivalently: current year total minus prior year total) |
| Growth % | net_change ÷ prior_year_total × 100 |
| Retention rate | returning_members ÷ prior_year_total × 100 |
| 3% quota target | prior_year_total × 1.03, rounded to nearest whole number |
8. Retention & Churn Dashboard
File: retention_dashboard.html · Script: scripts/generate_retention_dashboard.py
Answers: "Which age groups are we retaining? Which associations have the best retention? Where is churn highest?"
Data sources
State-level and association-level retention draw from prod_marts.int_player_home_association (All Active scope), so the member universe and per-season counts are identical to the Growth Dashboard — same playing-member filter, same registration_year season basis. Association retention uses HOME attribution (see HOME vs SERVED), matching Growth. Age-group retention draws from prod_marts.int_playing_registrations joined to prod_marts.dim_users for date of birth.
How retention is calculated — definition & assumptions
- Year-over-year (adjacent seasons). Retention always compares a season against the immediately prior season (Y vs Y−1) — never a multi-year gap. Retention % = members in Y who also played in Y−1, divided by the Y−1 base.
- Same HOME association. For association retention a member counts as "retained" only if their HOME association is the same in both seasons. A member who moves to another association is counted as churned at the origin and new at the destination. So association retention measures association loyalty, not whether the person stayed in netball. (State-level retention is the same idea without the association condition — it measures staying in the sport.)
- Bounded 0–100%. Because returners are always a subset of the prior-year base, retention can never exceed 100% and churn is never negative. This holds for every overall figure — statewide, per association, and the detail table. The one exception is the per-age-band band rate shown when hovering a C5 segment, which can exceed 100% on very small bands — see "Age bands are not fixed cohorts" below.
- "New" is broader than first-time. New = total − returning, so it includes genuinely first-time members, members returning after a gap, and (for associations) members who transferred in from another association.
- The insight strip's "Best retention" age group has no minimum-size floor. A small band (e.g. 65+, only a few dozen members statewide) can top it on a strong year — check the C3 heatmap's member counts before quoting it.
Age Group Definitions
Age is calculated at the time of registration — not at 1 January or any other fixed date. Age = age on the registration date. (One variation: the age-band split inside the C5 bars uses age as at 1 July of the season year, a fixed point so every member of a season is banded consistently. Both are "age at the time", never age today.)
| Label | Age range |
|---|---|
| Under 10 | Age < 10 |
| 10–14 | Age 10 to 14 (inclusive) |
| 15–17 | Age 15 to 17 |
| 18–24 | Age 18 to 24 |
| 25–34 | Age 25 to 34 |
| 35–44 | Age 35 to 44 |
| 45–54 | Age 45 to 54 |
| 55–64 | Age 55 to 64 |
| 65+ | Age 65 and over |
| Unknown | Date of birth is NULL or flagged as invalid (is_valid_dob = FALSE) |
Stacked bar chart per season year showing how the total member count is composed. Three stacked layers: Retained members (returned from prior year), New members (first time), and a visual indicator for Churned members (usually displayed as a negative bar or separate overlay). Lets you see at a glance whether growth is driven by new members or by retaining existing ones.
Dual-line chart across all years. One line shows the Retention Rate (%) and one shows the Churn Rate (%). These two lines should sum to ~100% (with minor rounding differences). Helps identify multi-year trends — is retention improving or worsening across seasons?
- Retention Rate = returning ÷ prior year × 100
- Churn Rate = (prior year − returning) ÷ prior year × 100
Colour-coded grid: rows = age groups, columns = season years. Each cell shows the retention % for that age group in that year. Darker / more saturated colour = higher retention. Quickly reveals which age cohorts are loyal vs which are at-risk of leaving the sport.
- Empty cells (2021 age groups) have no prior year to compare — show as N/A
- Hover over any cell for exact % and member counts
- Read with care: members age between brackets (a 14-year-old becomes 15), so an age band is not a fixed cohort year-to-year. The figure is "this year's age-band members who also played last year" against last year's band size — a directional loyalty signal, not a closed-cohort survival rate.
Grouped bar chart comparing two seasons for each age group. Useful for seeing exactly how retention has shifted year-on-year within each cohort. Driven by the year slicer — select a year to compare it against the prior year.
Every association's overall retention for the selected year, ranked lowest → highest. The length of each bar is the association's overall retention %; the coloured segments split that figure by age band (each segment = that band's retained members ÷ the association's total prior-year base, so the segments always add up to the overall figure). The dot at the end is the verdict: green ≥ 70%, amber 55–70%, red < 55%. Shows at a glance not just how much an association retains, but which age groups drive or drag it.
- Year-over-year and same-HOME-association (see assumptions above). The overall figure is always 0–100% and matches that association's row in the detail table exactly.
- Search box filters to an association by name; the min selector (50 / 100 / 200 / 500 prior-year members, default 50) hides very small associations whose rates are noisy.
- Hovering a segment shows that band's own band rate — on tiny bands this can exceed 100% (see "Age bands are not fixed cohorts" above).
- The age-band split uses age as at 1 July of the season year.
Sortable table listing every association with: prior year members, current members, retained count, new members, churned members, retention %, churn %. Can be sorted by any column. Export as needed for reports.
- retained = prior-year members who stayed at the same association (a subset of prior year) · churned = prior year − retained · new = current − retained (first-time + lapsed + transfers-in).
- retention % + churn % always sum to 100%; both are taken over the prior-year base.
- Only associations with members in the selected season appear (C5 follows the same rule). An association whose members all left — most often one that was renamed in the system — has no row that year, and a rename shows up as churn at the old name and "new" members at the new name. The statewide figures are unaffected: those members are still counted correctly there.
9. YTD Comparison Dashboard
File: ytd_dashboard.html · Script: scripts/generate_ytd_dashboard.py
Answers: "At the same point in the season, are we ahead of or behind last year? Which month drove most registrations? Which product types are growing?"
What "Year-to-Date" means in this dashboard
For example, if today is 29 May 2026, then for every historical season the data is cut off at the equivalent of "29 May" in that season (i.e. the same point relative to season start). This ensures a fair apples-to-apples comparison — we are not comparing a full 2025 season against a partial 2026 season. The YTD comparison stays strictly within the season window on both sides (the "Outside window" cohort is a Deep Dive concept, not a YTD one).
What "New Members" means in this dashboard
In the YTD dashboard, "new members" does not mean "first time ever". It means the first registration date in that season falls in a given month. This measures when people first entered the system each season, regardless of whether they were members in prior seasons. It is a measure of registration timing, not first-ever participation.
Line chart comparing cumulative member counts across all selected seasons (default: 2026 vs 2025), from the start of the season (Nov 1) up to today's equivalent date. The running total shows how quickly each season built its membership base month by month.
- Solid line = current year (2026)
- Dashed line = prior year (2025)
- Vertical dashed line marks the current "today" cutoff point
- X-axis shows season months (Nov, Dec, Jan … up to current month)
Grouped bar chart showing new registrations per season month, comparing two selected years. Each pair of bars represents one month — left bar = Year A, right bar = Year B. Shows which months are performing ahead of or behind the comparison year.
- Bars stop at the current month (no future months shown)
- Height = count of members whose first registration in that season fell in that month
Stacked bar chart per year showing the mix of product categories: Player, NetSetGo, Little Nettys, Player - Social, Player - 10 Week, Team Registration, Walking Netball. Each colour segment = one product category. Helps identify shifts in product mix — e.g. is the growth being driven by NSG or by senior players?
Stacked bar or grouped bar showing the payment status mix for each year at the YTD cutoff date. Statuses like Registered, Pending Payment, Awaiting Payment are shown in proportion. Useful for understanding how much of the membership base is fully confirmed vs still in processing.
- Applied after the full playing filter (excludes cancelled, de-registered etc.)
- Gives a sense of how "clean" the registration pipeline is
Every association with its YTD member count for Year A vs Year B, the absolute difference, and the % change. Sortable. The headline columns are SERVED (touchpoint) so they match what each association sees in Netball Connect at the same date; additional HOME columns are shown alongside so the table reconciles to the state KPI.
- SERVED columns sum to more than the state total (cross-association members counted at each association)
- HOME columns sum exactly to the state KPI (each member counted once)
- The total row shows both, so you can see the overlap explicitly
prod_marts.int_player_home_association / int_playing_registrations on the registration_year basis.
10. Revenue Performance Dashboard
File: revenue_dashboard.html · Script: scripts/generate_revenue_dashboard.py
"How much has NQ earned? Which fee type is driving revenue? How do we compare to prior years? Which associations generate the most NQ revenue?"
source_account = 'social' AND fee_type = 'competition', surfaced via the revenue_stream = 'Competition Hosting' label.Scale: ~$1.075M total, spread across 2022–2026 (2024 largest at ~$390K) — a steady back-catalogue, not a sudden spike.
Data source
Source table: prod_marts.fact_payments. This contains one row per payment line item (not per person). The key fields are:
| Field | Meaning |
|---|---|
| transaction_date | Date the payment was processed |
| fee_type | Category of the fee: 'membership', 'shop', 'competition', 'affiliate', 'deposit', etc. |
| source_account | Which Netball Connect login the row settled through: primary (NQ main), social (NQ's comp-organiser Stripe Connect), or hpl (Heart Premier League — reserved, not yet ingested). |
| net_settlement | Gross amount settled to NQ's account — the platform fee is billed separately, not netted here. Refunds appear as negative rows. |
| revenue_stream | Derived dashboard label: 'Membership', 'Shop', or 'Competition Hosting'. |
| transaction_id | Unique payment transaction identifier (unique within a source_account) |
| competition_organiser / organisation_name | Which association generated the transaction |
Three revenue streams represent money that flows to NQ:
- Membership — NQ's levy/cap fee component from each member registration. The per-head fee NQ charges regardless of the member's association. (
fee_type = 'membership', any source_account.) - Shop — NQ merchandise and online store purchases processed through Netball Connect. (
fee_type = 'shop'.) - Competition Hosting — competition fees NQ collects as the competition organiser, via the Social Competition account. (
source_account = 'social' AND fee_type = 'competition'.)
Fee types that remain excluded: Competition revenue on the primary account (it belongs to the affiliate that ran the competition, not NQ), Affiliate, Deposit, and fees_to_be_withdrawn (internal adjustment — never revenue). The only competition revenue counted is the social-account slice booked as Competition Hosting.
'shop' fee type rather than 'membership', depending on how the transaction was recorded in Netball Connect. These entries are captured in the Shop revenue segment. If you see unusually high Shop revenue for an association in a given period, check whether Team Registration or Team Nomination fees have been categorised as Shop for that period.
Calendar Year vs Season Year
The Revenue Dashboard is unique in that it offers two date groupings:
- Calendar Year (Jan–Dec): Groups revenue by the calendar year in which the transaction occurred. Standard for finance reporting aligned to NQ's financial year.
- Season Year (Nov–Oct): Groups revenue by the Netball season year (Nov 1 prior year to Oct 31 named year). Useful for comparing revenue alongside member count data which uses season years.
Refunds
Refunds are captured as negative net_settlement rows in fact_payments. They are already incorporated into net revenue figures. Do not add the refund_amount field separately as that would double-count refunds.
Stacked bar showing Membership, Shop and Competition Hosting revenue per year, with a secondary-axis line showing the year-on-year % change. Toggle between Calendar Year and Season Year using the buttons at the top of the chart.
- Blue segment = Membership revenue (NQ levy)
- Orange segment = Shop revenue
- Purple segment = Competition Hosting (NQ-as-organiser)
- Grey line = YoY growth %
Colour grid: years on the Y-axis, calendar months on the X-axis. Each cell is coloured by net revenue amount — darker blue = higher revenue month. Instantly reveals seasonality (which months NQ collects most revenue) and year-on-year shifts.
- Jan–Feb–Mar are typically peak months (summer season registrations in SE QLD)
- Hover for exact revenue and transaction count
Line chart with one line per year, all plotted on the same Jan–Dec x-axis. Shows the seasonality pattern and lets you compare the same months across all years. Useful for detecting whether a particular month is consistently strong or weak.
Stacked bar per year showing the split between Membership, Shop and Competition Hosting revenue across all years. Useful for understanding whether the revenue mix is shifting — e.g. how material is Competition Hosting relative to the Membership levy?
Horizontal bar chart showing associations ranked by the NQ revenue they generate. Updated by the year slicer. Note: this is the NQ levy portion only — associations with more members generate more NQ revenue, but this does not reflect the association's own total revenue.
Sortable table showing YoY revenue comparison per association: current year revenue, prior year revenue, absolute change, % change. Includes transaction count. Useful for finance reconciliation and association-level revenue queries.
KPI Cards (top row)
| Card | What it shows |
|---|---|
| Total NQ Revenue | Net revenue (Membership + Shop + Competition Hosting) for the latest year selected. Refunds already deducted. |
| Membership Revenue | NQ levy component only for the latest year. |
| Shop Revenue | NQ merchandise sales for the latest year. |
| Competition Hosting | NQ-as-organiser competition fees (social account) for the latest year. |
| Total Refunds | Total negative net_settlement rows for the year — money returned to members. |
| Avg Revenue per Transaction | Net membership revenue divided by number of unique membership transactions. Tracks the per-registration levy rate over time. |
11. Association Deep Dive Dashboard
File: assoc_deep_dive.html · Script: scripts/generate_association_deep_dive.py
"For a specific association, are they ahead of or behind last year? Which clubs within the association are growing? What's the daily registration momentum? What product mix does this association have?"
Association Selector
The dashboard defaults to the largest association for 2026 (by total members). A dropdown at the top allows you to switch to any of the ~117 associations. Changing the association updates all charts and tables on the page.
Data embedded
All association data for all years is embedded in the HTML at generation time. No database queries happen in the browser — the dashboard works fully offline and on SharePoint once generated. Figures derive from prod_marts.int_player_home_association and int_playing_registrations.
registration_year. See HOME vs SERVED and Season Date Window.
Line chart showing the cumulative (running total) unique member count per season month for all years (2021–2026). Each line = one season year. Allows comparison of how quickly registrations built up across all years for this specific association. The headline line uses SERVED (touchpoint) to match Netball Connect; a HOME comparison is also plotted — see HOME vs SERVED.
- X-axis = season months (Nov=1 to Oct=12, labels as Nov, Dec, Jan…)
- Current year (2026) line is thicker/highlighted
- Lines end at the current month (incomplete seasons are not projected forward)
New unique registrations over the season, comparing the two selected years. Peaks and troughs reflect registration events, school terms, and campaign impacts. Useful for a participation or marketing manager to see "what's driving the numbers" and where acceleration happened — including whether a digital advertising push lifted registrations.
- Daily / Weekly toggle. Daily shows per-day markers with a 7-day rolling average; Weekly shows new players per season-week as grouped bars (this year vs last), which is the clearest way to see at regular intervals whether a campaign drove a lift.
- Both years align on the same season-week / season-day rhythm (week 1 = Nov 1), so the comparison is like-for-like.
- Hover any point or bar for the exact calendar date (and, in weekly, the week-starting date).
- Only the two selected years are shown (daily volume is too large to plot all years at once).
Stacked bar showing how many members registered under each product category (Player, NetSetGo, Little Nettys, Player - Social, Player - 10 Week, Team Registration, Walking Netball) for each year. Helps identify whether a specific association relies heavily on one product type.
Horizontal bar chart showing member counts for each individual club and school within the selected association. Updated by a year slicer so you can compare club-level composition between years. Identifies which clubs are growing, stable, or declining.
- Clubs and schools are not rolled up here — each is shown individually
- Club/school names come from the
club_namefield inint_playing_registrations(joined onaffiliate_sk, not on name — affiliate names can collide)
Bar chart or KPI panel showing this association's retention rate for the selected year: how many of their prior-year members came back. Compares the association's retention against the NQ average for context. Retention uses HOME attribution (consistent with the Retention Dashboard) so the member base is stable year-on-year.
Top-row KPI cards for the selected association and year. The headline is SERVED (matches Netball Connect), with the HOME figure shown as the reconciling secondary value beneath it:
- YTD (to today) — SERVED to the current date, with HOME shown below
- Full Season — SERVED across the whole season, with HOME below
- Prior year, Net change, and Retention rate for context. Retention, New and Lost are measured against the immediately prior season (Year A−1), independent of the Year B slicer.
Worked example — Downey Park, Season 2025: HOME 4,331 vs SERVED 5,165 (full season); SERVED-to-29-May ≈ 4,638. The HOME–SERVED gap (~830) is the members Downey Park shares with other associations that season.
NSG (NetSetGo) handling
NetSetGo registrations are captured separately throughout the dashboard. The product category filter ensures that NSG registrations appear as their own category rather than being lumped into "Player". This is important for pathway reporting — NSG is the junior entry point into the sport.
12. Participation & Revenue Dashboard
File: participation_revenue_dashboard.html · Script: scripts/generate_participation_revenue_dashboard.py
Answers: "As unique participation grows, is NQ earning more per head? Which associations punch above or below their member count for NQ revenue? Where do members concentrate, and how does revenue concentration differ?"
What this dashboard adds
This is the bridge between the membership dashboards (which show people) and the Revenue Dashboard (which shows dollars). Its headline metric is NQ membership levy per unique playing member — how much NQ earns for each head of participation.
Data sources
prod_marts.int_player_home_association— unique playing members per season (HOME attribution, sums to the state total).prod_marts.fact_paymentswithfee_type = 'membership'— NQ levy revenue.- Each membership payment is attributed to the payer's HOME association for that season via the member's internal account key. This keeps members and revenue on the same association key (no payout-name matching) and reconciles to ~99% of the state membership total. The small residual is non-playing membership levies not tied to a playing member.
KPI Cards
| Card | What it shows |
|---|---|
| Unique Members | All Active playing members for the selected season. |
| NQ Membership Levy | Net settlement, membership fee type, for the selected season. |
| Revenue / Member | NQ levy ÷ unique members — the headline yield metric. |
| YoY Rev / Member | Dollar change in $/member vs prior season; positive = higher yield per head. |
| Levy Concentration | Top-10 associations' share of NQ levy for the season. |
Bars = unique members per season (left axis). Line = $/member (right axis). Shows whether revenue growth comes from more members, higher yield per head, or both. Selected season highlighted; the current (in-progress) season is rendered in a lighter shade.
One dot per association. The dashed diagonal is the state-average yield: dots above earn more NQ levy per member than average, dots below earn less. Hover for the exact $/member.
Bar chart of members in each association size band (<100, 100–499, 500–999, 1000+), with the number of associations in each band labelled on the bar. This is a people view — how members spread across associations.
Top-25 associations by NQ levy (bars) with the cumulative-share curve. This is a dollars view — a separate question from participation distribution: a few large associations may earn a disproportionate share of NQ levy, or revenue may track participation closely. The curve tells you which.
Every association with its members, NQ levy, $/member, and the dollar gap vs the state average. Sortable. Use it to identify associations that yield notably more or less per member than peers of similar size.
13. Analytical Dashboards (v2)
Sections 7–12 cover the six core dashboards — the membership and revenue numbers NQ reports on. This section covers seven analytical dashboards (six added in July 2026; FN Penetration added November 2026). They answer narrower, more exploratory questions and are built for a specific role rather than for general reporting.
| Dashboard | What it answers | Counting basis | Main caveat |
|---|---|---|---|
| 🏆 Competitions | What competitions are running, who plays and works in them, how many teams were nominated, and what entry-fee revenue was captured. Built for the Competition Program Manager. | Competition × season. People counted per competition, so a player in two competitions appears in both — this is a SERVED / touchpoint view, not a de-duplicated member count. | Team nominations are read straight from raw.team_registration, which is not yet modelled into staging. Entry-fee revenue is sparse by design — most competition fees are invoiced outside Netball Connect or sit in the undifferentiated shop lump. |
| 🌎 Demographics | Age and gender profile, disability inclusion, and geographic reach of playing members across LGAs and state electorates — both as raw counts and per head of population. | HOME — de-duplicated playing members, one association each. Totals reconcile to the state headline. | From 2024, LGA is derived by placing the member's geocoded home address inside the ABS LGA boundaries (about 97% of members); the remainder, and all seasons before 2024, fall back to the ABS majority LGA for the member's current postcode. That fallback assigns a shared postcode wholly to its largest LGA, which is why small remote and Aboriginal Shire LGAs only appear from 2024. Electorate is season-aware. Disability is self-reported at registration. |
| 🌱 First Nations | Aboriginal and Torres Strait Islander participation: trend, age and gender profile, home associations, LGA reach, and retention compared with all other members. | HOME — de-duplicated playing members. | Identification is self-reported at registration. The field is effectively mandatory so coverage is close to complete, but it remains self-identification. Association and LGA cells under 5 members are suppressed. |
| 🗺 FN Penetration | Where First Nations participation is strong or under-weight relative to the First Nations population, across the 19 Queensland SA4 regions — market penetration on an age-matched 5–19 basis — and, for each region, whether a low FN rate reflects low netball participation overall or a First Nations-specific gap. Every region shows its FN rate beside the non-FN rate for the same region, the ratio between them, and the extra FN players needed to reach the selected benchmark: local parity (the region's own non-FN rate, the default) or the state FN average. A season toggle offers each completed season from 2025. A scatter of FN rate against non-FN rate and a four-way "reading" column sort the regions into: thriving overall but FN behind; low overall and FN lower again; low overall with FN in step; thriving and near parity. | HOME — de-duplicated playing members with a geocoded home address, placed into SA4 by point-in-polygon and split by the self-identified FN flag; each side divided by ABS Census 2021 population of the same Indigenous status, age band and sex. Non-FN population is total persons minus FN persons, so census "not stated" sits with the non-FN members whose status is also not stated. "In step" means the FN rate is at least 80% of the local non-FN rate; "low overall" means the region's non-FN rate is below the state non-FN rate. | The map covers geocoded members only (~95% of FN members), so mapped statewide penetration (3.7% in season 2026, 3.3% in 2025) sits slightly below the full statewide figure (3.8% and 3.5%). The population is Census 2021 in every season. The state-average benchmark only finds regions that lag other First Nations communities; it says nothing about parity, and it reads regions with strong netball overall (Central Queensland, Darling Downs) as having no opportunity when FN young people there play at about 70% of the local rate. Use local parity for any statement about equity. Penetration is not the same number as "FN share of members" or the 4.6% population share — see the three FN percentages below. |
| 📉 Churn Deep Dive | For one selected association: what actually happened to last season's players — returned, moved to a named association, or left netball altogether. Includes the flow diagram and retention by age band. The season still being played is offered as a separate view marked in progress: it shows who has returned so far, who has registered elsewhere and who has not registered anywhere yet. | Association-level. Retention here is within-association (did they come back to this association), which is stricter than the statewide retention figure. | Do not add association cells together to get a state figure — a player who moves associations counts as lost here but is retained statewide. The statewide number is roughly 11 percentage points higher. Use the Retention dashboard (section 8) for the state view. The in-progress view is not a churn figure — "not back yet" shrinks as the season's registrations come in. |
| 🎯 Market Penetration | Netball participation as a percentage of the addressable population by SA2 (ABS statistical area), with a full ranked target list of every SA2 (searchable, top 18 highlighted, switchable to an SA3 regional roll-up — click a region to drill into its suburbs) and a map search that zooms straight to any suburb. | Geocoded members placed into SA2 polygons, divided by ABS population for the same age band and sex. Girls, ages 5–19. | Only members with a geocodable address appear. The benchmark is the 75th-percentile SA2, so it moves as participation moves — it is a relative target, not an absolute one. Adult and social participation is out of scope. The season toggle offers each completed season from 2025; the population is the ABS 2024 estimate in every season, so only the player count changes. |
| 🧭 Opportunity | Every strategy layer on one SA2 map — penetration, school penetration and ICSEA, facilities and court access, approximate drive time to the nearest venue (switchable between competition venues and all facilities), population growth, retention — blended into a single opportunity score with weightings the Participation Manager can tune live, plus a searchable ranked table of every scored SA2. | Same SA2 basis as Market Penetration. The score is a weighted blend of percentile ranks, not of raw values. The season toggle switches the playing members; the school layer (2026) and retention (2025→2026) are the same in every season view. | The score is a prioritisation aid, not a forecast. Because it ranks SA2s against each other, every score is relative — a high score means "high opportunity compared with other SA2s", not a predicted number of new players. Where an SA2 is missing an ingredient, that ingredient's weight is redistributed across the others. |
First Nations — three percentages that are not interchangeable key concept
Three different First Nations percentages circulate in NQ reporting. All three are correct; each divides a different numerator by a different denominator, so they answer different questions and must not be swapped for one another. Figures below are season 2025, All Active, audited against the warehouse and ABS Census 2021 on 2 November 2026.
| Percentage | Formula | Value (2025) | What it answers |
|---|---|---|---|
| A. FN share of members | FN members ÷ all members3,426 ÷ 63,400 |
5.4% (5.5% among 5–19s) |
"How First Nations is our membership?" This is the figure on the First Nations dashboard. The denominator is NQ members — it says nothing about the population we have not reached. |
| B. FN share of QLD population | Census FN persons ÷ all persons237,324 ÷ 5,156,139 |
4.6% all ages 7.9% among 5–19s |
"How First Nations is Queensland?" This is the 4.6% quoted in the QGSO regional report — an all-ages figure. The First Nations population is much younger than the general population, so in the 5–19 band netball actually draws from, the share is 7.9%, not 4.6%. |
| C. Market penetration | FN members 5–19 ÷ census FN persons 5–192,729 ÷ 78,137 |
3.5% (non-FN rate: 5.2%) |
"Of First Nations young people in Queensland, how many play netball?" This is the figure on the FN Penetration dashboard (shown there as 3.3% for season 2025 because the map can only place geocoded members). The honest comparison is against the non-FN rate on the same basis. |
Reference data these dashboards depend on
Demographics, Market Penetration and Opportunity join to external reference extracts (ABS population and boundaries, school participation, the facility master) that live under data/reference/. That folder is deliberately excluded from the code repository, so these three dashboards will not rebuild on a fresh checkout until those extracts are recreated. The rebuild steps are documented in docs/market_penetration_project/01_DATA_PIPELINE_STEPS.md.
14. Full Glossary of Terms
| Term | Definition |
|---|---|
| All Active | The broader of the two member scope filters. Includes all playing members who passed the playing member filter, regardless of payment status (except explicitly excluded statuses like De-Registered). This is the headline membership count used across all dashboards. 2026 = 58,498 (live, as at 29 May 2026). |
| Registered Only | The narrow member scope. Only members whose payment_status = 'Registered' are counted. 2026 = 52,412 (live, as at 29 May 2026). |
| Season Year | The name given to a 12-month playing season. Season Y runs from 1 November of year Y−1 to 31 October of year Y. |
| Season Month | Months numbered 1–12 starting from November (month 1) to October (month 12) to align with the season window rather than the calendar year. |
| Registration Year | The field name in the database (registration_year) — equivalent to Season Year. Derived from the registration date using the season window logic. |
| external_user_id | The unique identifier for a person in Netball Connect. An internal, non-identifying key — it carries no name, email, mobile, or date of birth, so it is safe to use in aggregations. Used for deduplication — a person with multiple registrations is counted once per season by their external_user_id. |
| Unique members | The count of distinct external_user_ids. All member counts in all dashboards are unique member counts, not transaction counts. |
| Returning members | Members who registered in season Y AND also had a registration in season Y−1 (matched by external_user_id). |
| New members | Members in season Y who did NOT have a registration in season Y−1. Brand new to the system this season. |
| Churned members | Members from season Y−1 who did NOT register in season Y. People who did not return. |
| Net change | Current year total − prior year total. Equivalently: new members − churned members. |
| Retention rate | returning_members ÷ prior_year_total × 100. The percentage of the prior year's members who came back. |
| Churn rate | (prior_year_total − returning_members) ÷ prior_year_total × 100. The percentage of the prior year's members who did NOT come back. |
| 3% Growth Quota | NQ's internal target: each season must reach at least 3% growth above the prior year total. Target = prior_year × 1.03. Status is MET or MISSED. |
| Association roll-up | The logic that maps club and school registrations up to their parent association for association-level reporting. |
| affiliated_to | Field in dim_affiliates — the name of the parent association that a club or school belongs to. |
| net_settlement | Revenue field in fact_payments. GROSS — the full amount settled to NQ's account, NOT net of the platform fee (the NetballConnect platform fee is billed separately via fees_to_be_withdrawn). Customer-side and bank-side totals should reconcile near-exactly. Negative values indicate refunds. |
| fee_type | Category field in fact_payments. NQ revenue = 'membership' and 'shop' (any account) plus 'competition' on the social account only (Competition Hosting). 'competition' on the primary account, plus 'affiliate' and 'deposit', flow to other affiliates and are excluded. |
| HOME association | The single association a member is attributed to in a season — the one where they have the most playing registrations (ties broken by most recent registration date). Σ HOME = the state total exactly (no double-counting). Basis for Growth and Retention. See HOME vs SERVED. |
| SERVED (touchpoint) | A member counted at every association they registered with that season. Σ SERVED > the state total; the excess = cross-association members. Matches Netball Connect's own per-association export. Headline for YTD association table and Deep Dive. |
| players_served | Column in pbi_v_playing_growth_assoc holding the SERVED (touchpoint) count per association. |
| Outside window | Members whose registration date fell before 1 Nov (Y−1) or after 31 Oct (Y) but who still belong to season Y by the platform's registration_year. Shown as a separate bucket in the Deep Dive so months 1–12 + Outside window = the full season total. |
| source_account | Column on fact_payments identifying which Netball Connect login a row settled through: primary (NQ main), social (NQ's comp-organiser Stripe Connect), hpl (Heart Premier League — reserved, not yet ingested). |
| Competition Hosting | NQ revenue stream = source_account = 'social' AND fee_type = 'competition'. Competition fees NQ collects as the organiser (NQ Social Competition + Vicki Wilson tournaments). ~$1.075M across 2022–2026. Encoded as revenue_stream = 'Competition Hosting'. |
| revenue_stream | Derived label on the Revenue Dashboard: 'Membership', 'Shop', or 'Competition Hosting'. |
| int_playing_registrations | Canonical intermediate model — one row per qualifying playing registration after the full playing-member filter. Single source of the playing-member definition for every membership dashboard. |
| int_player_home_association | Canonical intermediate model — one row per (status scope, season, person) resolving each member to their HOME association, with the Outside-window flag and first-registration season month. Source of state and association headline counts. |
| YTD | Year-to-Date. In this context, data from the start of the season up to the equivalent of today's date in each historical season. Ensures fair comparison between a complete prior year and a partial current year. |
| prod_marts | The PostgreSQL schema containing all production-ready, transformed data. All dashboard queries use this schema exclusively. |
| pbi_v_playing_growth | Pre-built view in prod_marts. Aggregates season-level playing member counts, retention, churn, and quota status by season year and status scope. Used as the source of truth for growth metrics. |
| Little Nettys | NQ's early years participation program for children typically aged 3–4 years. Captured as a distinct product category. |
| NetSetGo (NSG) | Netball Australia's junior development program for children approximately aged 5–10. Key junior pathway product category. |
| Walking Netball | Low-impact version of netball for older adults and people returning after injury. Separate product category tracked across all dashboards. |
| Team Registration | A registration for a whole team (rather than an individual). Counted as one member in unique member counts when the team is identified by a single user ID, but typically each player on the team has their own individual registration as well. |
| Failed Registration | A registration that did not complete successfully. Only excluded if paid_fee_inc_gst ≤ 0 — those with a positive fee paid are retained in the count. |
| De-Registered | A member who has been formally removed from the system. Always excluded from all counts. |
| Incomplete Registration | A registration process that was started but not completed. No payment, no active membership. Always excluded. |
| is_valid_dob | Flag in dim_users. If FALSE, the date of birth is considered unreliable and the member is placed in the "Unknown" age group for age-group charts. |
| first_reg_date | The earliest registration date for a member within a season window. Used in YTD and deep dive dashboards to assign a member to the season month in which they first registered. |
| DISTINCT ON | PostgreSQL syntax used for deduplication — selects the first row per (season_year, external_user_id) combination, ordered by registration date, ensuring each person is counted only once per season. |
Netball Queensland Data & Analytics · Prepared 29 May 2026 · This document describes the same data pipeline and definitions that produce all dashboards. Live member counts shown are as-at 29 May 2026.