ĒMA Quarterly Impact Report

Static snapshot  ·  Data source: Trellis (Railway PostgreSQL)

Loading dashboard…
Core KPIs - Q1 2026 KPI 1 & 2: 3-month assessment window (system launched December 2025) · KPI 3: track completion date in Q1 · Expand each card for data integrity notes
KPI 1
Family Preservation Rate
86% target: 85%
On Track ✓
Children Preserved ÷ Children with Valid 3-Mo FWA
50 of 58 eligible children preserved  ·  20 children excluded (no valid FWA)
Data integrity notes
What we count
Each child belonging to an active mom counts as one data point, not the mom herself
Preserved means
Children whose Family Preservation Impact falls in the Prevention %: Prevented from CPS Involvement (stayed home, no CW contact), Prevented from Foster Care Placement (had CW involvement but stayed home), or Prevented from Permanent Removal (was in foster care but reunified)
Not preserved
Children in Temporary Removal (currently in foster care/kinship) or Permanent Removal (will not return). These are the Out of Care % and are excluded from the numerator
Who's included
Children whose moms had a completed Family Wellbeing Assessment within the last 3 months as of March 31
Who's not included
Children whose moms don't yet have a current assessment on file. They show up in the FWA Data Integrity tab so coordinators can follow up
Why 3 months?
ĒMA launched its new system in December 2025. There isn't yet enough history to use a 6-month window. This will expand in future quarters.
⚠ Known limitation
A small edit to any part of an assessment updates the assessment date, so "recent" doesn't always mean a full re-assessment was done. Coordinators should confirm full re-assessments were completed, not just edits.
KPI 2
Family Stability Score Improvement
71% target: 70%
On Track ✓
Moms with FSS Improvement ÷ Moms with Valid 3-Mo FWA Pair
37 of 52 moms show composite FSS improvement  ·  23 moms excluded (no valid FWA pair)
Data integrity notes
Who's included
52 moms who had both a starting assessment on file and a current one within the last 3 months - we need both to measure change
What counts as improvement
Any increase in a mom's combined score across 10 life areas from her starting point to her most recent assessment
What's not scored here
The Child Welfare area is tracked separately - it feeds into KPI 1, not this score
Who's not included
23 moms who are missing either a starting or current assessment - they appear in the FWA Data Integrity tab for coordinator follow-up
Why 3 months?
ĒMA launched its new system in December 2025 - there isn't yet enough history to use a 6-month window. This will expand in future quarters.
⚠ Known limitation
A small edit to any part of an assessment updates the assessment date - so "recent" doesn't always mean a full re-assessment was done. Coordinators should confirm full re-assessments were completed, not just edits.
KPI 3
Learning Progress
72% target: 70%
On Track ✓
Completions where post composite > pre composite ÷ Completions with valid pre & post pair
13 of 18 eligible completions show improvement  ·  NPP: 8/10 · EP: 3/5 · RR: 2/3  ·  excludes missing or inverted score pairs
Data integrity notes
What counts as improvement
Any increase in a mom's total score between her start-of-track assessment and her end-of-track assessment - even one point up counts
Who's included
Moms who completed a track in Q1 and have both a start and end assessment on file - each track counts separately
Who's not included
Moms with a missing start or end assessment, and a small number of records with data entry issues where the end date was recorded before the start date - all excluded counts are shown alongside this KPI
Each track is its own score
NPP, EP, and RR each use their own assessment - a mom completing two tracks counts twice, once per track
Not related to FWA
This KPI uses track learning assessments only - completely separate from the Family Wellbeing Assessment used in KPI 1 and KPI 2
Program Status - Q1 2026
Mothers
94 served during Q1
Active During Q1 each mother counted once
61  ·  65% In a Track
matched with an advocate
33  ·  35% Membership Community
not in a track
Q1 Activity event date falls Jan 1–Mar 31
Referrals Received
31
Referral submitted
Engaged in Program
22
Became a client
Track Completions
18
Track completed
Data integrity notes
How we count
Every mom active at any point Jan 1–Mar 31 is included - including moms who completed or exited during the quarter
Reporting use
This is the primary count used for grant and investor reporting
Track completions
18 total - 11 Full Track, 4 without Support Sessions, 3 without Post Assessment
Served vs. active
94 includes every mom active at any point during Q1 - some exited or completed during the quarter, so the number still active at quarter end is lower. End-of-quarter counts are available in the monthly report.
Bucket logic
Every mom appears in exactly one bucket - In a Track takes precedence if she had a pairing at any point in Q1, otherwise Membership Community
No double counting
In a Track + Membership Community = 94 total · no mom appears in both
Children
142 served during Q1
Child Welfare Prevention
28
Children prevented from child welfare involvement
Est. Impact Value
~$1.09M
28 × $38,850 · estimated prevention value
Data integrity notes
Children served
Sum of child profiles linked to moms active during Q1 - only children the mom has legal authority over are counted
Missing counts
Where a mom has no child count recorded, the average child count across all active moms is used as a proxy
Child welfare prevention
Children prevented from CPS involvement or foster care placement, sourced from the Child Welfare area of the Family Wellbeing Assessment
Est. impact value
Prevention count × $38,850 per child, based on average annual foster care costs from ĒMA white paper
Advocates
57 active during Q1
Active During Q1 each advocate counted once · priority order below
5  ·  9% Group facilitators
leads a group track
47  ·  82% Paired 1:1
matched with a mom one-on-one
4  ·  7% Waiting to be paired
approved and available, not yet matched with a mom
1  ·  2% Taking a break
active but temporarily paused, not currently matched with a mom
Q1 Activity event date falls Jan 1–Mar 31
Application Received
14
Applied
Trained
9
Completed training
Approved
7
Passed final review
Became Active
7
Became active
Became Inactive
3
Stepped down
Data integrity notes
Unique counts
Each advocate appears in exactly one bucket - priority order is: Group facilitator → Paired 1:1 → Waiting to be Paired → Taking a Break
Pairing source
Group facilitator and Paired 1:1 counts are sourced from mom's pairing record - this is the truth for whether pairing happened, regardless of what the advocate's own status says
Sub-status fallback
Waiting to be Paired and Taking a Break are sourced from the advocate's own sub-status - only used when no active pairing is found on mom's side
⚠ Known workaround
Some advocates are set to Active status with a Taking a Break sub-status as a system workaround. Mom's pairing record takes precedence - if a pairing exists on mom's side they are correctly classified as Paired 1:1 or Group. Only if no pairing exists do they fall into the Taking a Break bucket.
⚠ Pending
Group facilitator count and sub-status values need confirmation with live data before this section is finalized
Multiple counts
One advocate may appear in multiple Q1 Activity counts - an advocate who trained and was approved in Q1 appears in both. These are independent events, not a funnel.
Became active
Advocates whose active status began during Q1 - not advocates already active before Q1 started
Became inactive
Advocates who stepped down during Q1 only - does not include anyone already inactive before Q1 started
Referral Summary - Q1 2026 Date anchor: referral submitted date falls Jan 1–Mar 31, 2026
Total Referrals Received
38
All sources
Engaged in Program
29
Converted from referral
Conversion Rate
91%
29 engaged ÷ 32 resolved (engaged + did not engage) · pending excluded
Still Pending
6
May still convert - not counted in rate
Did Not Engage
3
Unable to contact, declined, ineligible, or referred elsewhere
Data Integrity Notes
How conversion is counted
Converted = referral status is "Engaged In Program"
Pending
Status is "Prospect" or "Prospect Intake Scheduled" - appointment may still happen. Not a failed referral.
Did not convert
Status is "Did Not Engage In Program" - includes: unable to contact, client declined, ineligible, duplicate referral, or referred to another program. These are closed and excluded from the pending count.
Denominator
Conversion rate = Engaged In Program ÷ all referrals received (excluding duplicates). Pending referrals are shown separately - not included in the rate.
Referrals by Source Referral submitted date in Q1 · each row = one referral source category · numerator/denominator shown per source
Referral SourceThe organization or pathway that connected the mother to ĒMA. This is a free-text field in Trellis, so naming may vary. Referrals ReceivedTotal referrals submitted from this source during the quarter, regardless of outcome. Engaged in ProgramReferrals where the mother's status moved to "Engaged In Program." This is the Trellis referral status, not necessarily the same as completing all intake steps. Conversion RateEngaged ÷ resolved referrals (engaged + did not engage). Pending referrals are excluded from this rate. Still PendingReferrals still in process. May still convert in a later period. Not counted in the conversion rate. % of All ReferralsThis source's share of all referrals received during the quarter.
DCF / Child Protective Services 121091%1 32%
Self-Referral / Walk-In 8583%2 21%
Community Partner / Nonprofit 76100%1 18%
Hospital / Healthcare Provider 5480%0 13%
Other Sources 2 sources with fewer than 5 referrals Show 64100%2 16%
All Sources 382991%6 100%
Data Integrity Notes
⚠ Sample data
The referral source categories shown here are placeholders - real category names will be confirmed before this report is finalized
⚠ Data quality
The referral agency field in Trellis is a free-text input, not a dropdown - live data is expected to contain misspellings and inconsistent entries for the same agency (e.g. "DCF", "Dept. of Children & Families", "Department of Children and Families"). RD will need to normalize these before source groupings are reliable.
Missing sources
Any referral with no source on record will be shown as "Unknown / Not Recorded" - not quietly left out of the count
Track Completions — Q1 2026 Tracks closed Jan 1–Mar 31  ·  one track = one record  ·  a mom completing two tracks counts twice  ·  excludes same-day close/reopen duplicates
Track Completions
18 tracks completed in Q1 2026

Completion Rate

53% 18 of 34 tracks
Tracks Completed ÷ Tracks Closed in Q1
Data integrity notes
What counts
Any track with a closed status recorded between January 1 and March 31, regardless of when that track started.
EP and RR
These tracks launched more recently, so moms who started later in Q1 had less time to finish. Lower completion rates for EP and RR are an expected pattern at this stage, not a concern.
Excluded
A small number of records where a track was closed and immediately reopened the same day are excluded. This is a known data entry pattern used to change a track type, not a real exit.

Delivery Format

1:1 Advocacy
13
completed tracks
Group Setting
5
completed tracks
72% 1:1 Advocacy
28% Group Setting
Data integrity notes

How Tracks Completed

Completed — Full Track
11
Completed — Without Support Sessions
4
Completed — Without Post Assessment
3
18 completed 18
Data integrity notes
Full track
All required curriculum sessions, at least one support session on record, and a post-assessment on file.
Without support
Completed all required curriculum sessions but no support sessions were recorded during the track.
Without post
Completed all required sessions but has no post-assessment on record. These 3 tracks are excluded from the Learning Progress KPI denominator.

Why Tracks Ended Early

Track Requirements Unmet
6
Client Choice
4
Did Not Initiate Sessions
3
Relocated
2
Other
1
16 incomplete 16
Data integrity notes
Requirements Unmet
The most common exit reason. Worth reviewing whether these reflect genuine departures or cases where session documentation is incomplete.
Did Not Initiate
These moms enrolled in the program but never attended a first session. May indicate a gap between when a mom signs up and when her first session is scheduled.

Language Split

60% English
40% Spanish
By Track — Q1 2026 Tracks closed Jan 1–Mar 31  ·  session data for completed tracks only

Share of Tracks Closed by Type

68  ·  48%
Nurturing Parenting
48  ·  34%
Empowered Parenting
27  ·  19%
Roadmap to Resilience

Completion Rate by Track

Track Closed Completed Rate
Nurturing Parenting (NPP) 68 — —
Empowered Parenting (EP) 48 — —
Roadmap to Resilience (RR) 27 — —

Learning Progress by Track

Track Improved Valid Pairs Rate
NPP ———
EP ———
RR ———
Data integrity notes
Formula
Moms whose end-of-track assessment score was higher than their start-of-track score, out of all moms with both assessments on file. Same method as KPI 3.
Excluded
Once live data is connected, this will show counts excluded for: missing start assessment, missing end assessment, end date recorded before start date, or track marked completed without a final assessment.
Sample size
Rates with fewer than 20 valid pairs are flagged for caution. Rates with fewer than 10 are flagged as not yet interpretable. The ideal threshold for a stable rate is 30 or more pairs.

Average Sessions per Completed Track

Deferred to Q2
The majority of Q1 completed tracks are from moms who were migrated from the previous system (PromiseServes) in December 2025. Session recording practices differed between systems, making per-track session averages unreliable for this quarter. This section will be available starting Q2 2026 when all completions are from Trellis-native tracks.
Data integrity notes
Lesson marker
The dashed line shows how many curriculum lessons are in that track. When a bar ends before the marker, it may mean fewer sessions were recorded than expected — worth a documentation check.
NPP note
NPP is in phase-out. Lower counts may reflect program timing rather than disengagement.
ĒMA - Every Mother's Advocate
Our approach to data integrity →