Zim Cares — Dashboard

All jobs at the top, then everything happening inside the program in focus.

Data as of Mon Aug 10, 2026 · 12:27 PM
1 · All jobs — where every program stands
4,337
Signups, all jobs
across 6 active programs
2
Jobs at their school target
no work needed
4
Jobs still short
need more schools
41
Real replies in this program
94 out-of-office filtered out
0.5%
Hard bounce, all sending
pauses at 4.0%
Job Grades Signed up Schools with
a signup
School
target
Fill Still
needed
Deadline Need/day Actual/day Signups near
a store
Signups not
near a store
Status
GUNPLA Academia 339 284 4,500 6.3% 4,216 Dec 1, 2026 37.6 19.9 58 281 FOCUS 53% of pace
Clickeez Classroom Program 744 554 334 166% 0 Sep 20, 2026 24.9 177 567 At target
Jelly Blox Classroom Program 628 455 334 136% 0 Sep 30, 2026 21.6 189 439 At target
Tamagotchi Paradise Program 837 617 975 63% 358 Oct 30, 2026 4.5 40.1 195 642 Open on pace
Kinetic Sand Program 907 628 750 84% 122 Nov 14, 2026 1.3 34.2 260 647 Open on pace
Rubik's Classroom Program 882 654 750 87% 96 Nov 14, 2026 1.0 36.8 235 647 Open on pace
The target is a school count, not a signup count. A program stores schools_needed, so fill, "still needed" and "need/day" are all in schools — there is no per-program signup target in the database, and a sponsor quota for near-a-store versus anywhere-else is not stored either, so those two blue columns report what we have without a bar to clear.
Signups near a store come from each program's own eligibility snapshot, the only place a contact's distance to a store is recorded. A program that has never had that snapshot built shows an em-dash rather than a zero.
Actual/day is signups over the last 14 days. Need/day only appears while a program still has a future end date to work against.
2 · GUNPLA Academia — every campaign

GUNPLA Academia at a glance

19 campaigns, 32,062 emails handed to Postmark. Each campaign uses exactly one email copy, so the copy panel below can pool them.
37
Signed up
284 schools of 4,500 needed
36
Schools with a signup
of 16,267 schools emailed
1.03
Signups per converting school
program cap is 3 contacts/school
7.1%
Click rate
judges the email
1.1%
Visit → signup
judges the website
0.6%
Hard bounce
0.0% blocked, counted apart
Every campaign · click a row to open its detail
# Campaign Copy Sent Deliv. Bounce Open Clicks Click % Visits Signups Visit→
signup
Signups per
1,000 sent
Inbound
46 Global Outreach 5k Contacts Scrapped 8.10.26 - part 5 Gunpla Email Template 8/10/26 - For Scrapped 5,000 2,535 0.4% 15% 186 7.3% 552 0 0.0% 0.0 16 Detail ›
45 Global Outreach 5k Contacts Scrapped 8.10.26 - part 4 Gunpla Email Template 8/10/26 - For Scrapped 5,000 1,875 0.4% 18% 173 9.2% 663 2 0.3% 0.4 19 Detail ›
44 Global Outreach 5k Contacts Scrapped 8.10.26 - part 3 Gunpla Email Template 8/10/26 - For Scrapped 5,000 1,685 0.4% 16% 163 9.7% 650 1 0.2% 0.2 19 Detail ›
43 Global Outreach 5k Contacts Scrapped 8.10.26 - part 2 Gunpla Email Template 8/10/26 - For Scrapped 2,942 2,922 0.0% 21% 338 11.6% 258 9 3.5% 3.1 8 Detail ›
42 Global Outreach 5k Contacts Scrapped 8.10.26 part 1 Gunpla Email Template 8/10/26 - For Scrapped 5,000 935 2.6% 21% 66 7.1% 960 0 0.0% 0.0 52 Detail ›
41 Gunpla Email Request for our signups Gunpla Email Template 8/10/26 - for signups 1,080 1,058 0.0% 48% 203 19.2% 194 17 8.8% 15.7 0 Detail ›
34 Global Outreach 5k Contacts Scrapped 8.8.26 Gunpla Email Template 8/8/26 5,002 4,731 0.0% 19% 211 4.5% 86 1 1.2% 0.2 18 Detail ›
27 Gunpla Cold Outreach - COPY D - Gunpla COPY D — What students actually do — Gunpla 257 257 0.0% 0% 0 0.0% 10 0 0.0% 0.0 1 Detail ›
26 Gunpla Cold Outreach - COPY C - Gunpla COPY C — The honest ask — Gunpla 257 257 0.0% 0% 0 0.0% 5 0 0.0% 0.0 0 Detail ›
25 Gunpla Cold Outreach - COPY B - Gunpla COPY B — Straight to the offer — Gunpla 255 255 0.0% 0% 0 0.0% 7 1 14.3% 3.9 0 Detail ›
24 Gunpla Cold Outreach - COPY D - Build COPY D — What students actually do — Build 257 256 0.0% 0% 0 0.0% 7 1 14.3% 3.9 0 Detail ›
23 Gunpla Cold Outreach - COPY D - Kits COPY D — What students actually do — Kits 251 251 0.0% 0% 0 0.0% 4 0 0.0% 0.0 1 Detail ›
22 Gunpla Cold Outreach - COPY C - Build COPY C — The honest ask — Build 256 256 0.0% 0% 0 0.0% 7 0 0.0% 0.0 0 Detail ›
21 Gunpla Cold Outreach - COPY C - Kits COPY C — The honest ask — Kits 257 255 0.0% 0% 0 0.0% 7 1 14.3% 3.9 0 Detail ›
20 Gunpla Cold Outreach - COPY B - Build COPY B — Straight to the offer — Build 249 249 0.0% 0% 0 0.0% 11 0 0.0% 0.0 1 Detail ›
19 Gunpla Cold Outreach - COPY B - Kits COPY B — Straight to the offer — Kits 255 255 0.0% 0% 0 0.0% 12 1 8.3% 3.9 0 Detail ›
18 Gunpla Cold Outreach - COPY A - Build COPY A — Who we are first — Build 254 254 0.0% 0% 0 0.0% 5 0 0.0% 0.0 0 Detail ›
17 Gunpla Cold Outreach - COPY A - Kits COPY A — Who we are first — Kits 254 254 0.0% 0% 0 0.0% 11 0 0.0% 0.0 1 Detail ›
16 Gunpla Cold Outreach - COPY A - Gunpla COPY A — Who we are first — Gunpla 236 236 0.0% 0% 0 0.0% 17 3 17.6% 12.7 0 Detail ›
All 19 campaigns pooled 32,062 18,776 0.6% 17% 1,340 7.1% 3,466 37 1.1% 1.2 136
Do not pick a winner from this table. A single campaign row is a few hundred contacts and a handful of signups — far too small to separate anything. The panel below pools every campaign that shared an email copy, which is where the decision gets made.
Where each column comes from: sent, delivered, opens, clicks and bounces are the first-time timestamps Postmark's webhook writes on email_sends; visits are distinct people in email_tracking_events; signups are opt-ins to this program created after the campaign reached that contact. Delivered counts a confirmed delivery or a later open/click, since either proves arrival.
Visit → signup above 100% is a tracking gap, not a miracle. Signups come from the opt-in record while visits come from the tracking event, so someone who registered without a visit event firing — an untracked link, a blocked script, a form reached another way — counts in one and not the other.
3 · Which email copy is working

Which email gets people to clickthe email decision

Judged on click rate. A click happens before the website loads, so it measures the email and nothing after it. Each copy pools every campaign that used it.
The answer
Run Gunpla Email Template 8/10/26 - for signups. Drop Gunpla Email Template 8/10/26 - For Scrapped and Gunpla Email Template 8/8/26 and COPY D — What students actually do — Gunpla and COPY C — The honest ask — Gunpla and COPY B — Straight to the offer — Gunpla and COPY D — What students actually do — Build and COPY D — What students actually do — Kits and COPY C — The honest ask — Build and COPY C — The honest ask — Kits and COPY B — Straight to the offer — Build and COPY B — Straight to the offer — Kits and COPY A — Who we are first — Build and COPY A — Who we are first — Kits and COPY A — Who we are first — Gunpla.
“Gunpla Email Template 8/10/26 - for signups” leads at 19.2% clicks and “Gunpla Email Template 8/10/26 - For Scrapped” follows at 9.3%. That gap survives chance (p = 0.00), so it is safe to act on. Gunpla Email Template 8/10/26 - For Scrapped and Gunpla Email Template 8/8/26 and COPY D — What students actually do — Gunpla and COPY C — The honest ask — Gunpla and COPY B — Straight to the offer — Gunpla and COPY D — What students actually do — Build and COPY D — What students actually do — Kits and COPY C — The honest ask — Build and COPY C — The honest ask — Kits and COPY B — Straight to the offer — Build and COPY B — Straight to the offer — Kits and COPY A — Who we are first — Build and COPY A — Who we are first — Kits and COPY A — Who we are first — Gunpla are clearly behind and worth cutting.
Email copy Campaigns Delivered Opens Clicks Click rate 95% range
bars that do NOT overlap are definitely different
Signups per
1,000 sent
Do this
Gunpla Email Template 8/10/26 - for signups 1 1,058 48% 203 19.2%
0%10%20%
16.9–21.7%
15.7 SCALE IT
Gunpla Email Template 8/10/26 - For Scrapped 5 9,952 18% 926 9.3%
0%10%20%
8.7–9.9%
0.5 DROP IT
Gunpla Email Template 8/8/26 1 4,731 19% 211 4.5%
0%10%20%
3.9–5.1%
0.2 DROP IT
COPY D — What students actually do — Gunpla 1 257 0% 0 0.0%
0%10%20%
0.0–1.5%
0.0 DROP IT
COPY C — The honest ask — Gunpla 1 257 0% 0 0.0%
0%10%20%
0.0–1.5%
0.0 DROP IT
COPY B — Straight to the offer — Gunpla 1 255 0% 0 0.0%
0%10%20%
0.0–1.5%
3.9 DROP IT
COPY D — What students actually do — Build 1 256 0% 0 0.0%
0%10%20%
0.0–1.5%
3.9 DROP IT
COPY D — What students actually do — Kits 1 251 0% 0 0.0%
0%10%20%
0.0–1.5%
0.0 DROP IT
COPY C — The honest ask — Build 1 256 0% 0 0.0%
0%10%20%
0.0–1.5%
0.0 DROP IT
COPY C — The honest ask — Kits 1 255 0% 0 0.0%
0%10%20%
0.0–1.5%
3.9 DROP IT
COPY B — Straight to the offer — Build 1 249 0% 0 0.0%
0%10%20%
0.0–1.5%
0.0 DROP IT
COPY B — Straight to the offer — Kits 1 255 0% 0 0.0%
0%10%20%
0.0–1.5%
3.9 DROP IT
COPY A — Who we are first — Build 1 254 0% 0 0.0%
0%10%20%
0.0–1.5%
0.0 DROP IT
COPY A — Who we are first — Kits 1 254 0% 0 0.0%
0%10%20%
0.0–1.5%
0.0 DROP IT
COPY A — Who we are first — Gunpla 1 236 0% 0 0.0%
0%10%20%
0.0–1.6%
12.7 DROP IT
Why opens are greyed out. Apple Mail Privacy Protection opens every message automatically, so open rate mostly measures which mail apps our contacts use, not how good the subject line is. Judging copy on opens can pick a different winner, and it would be the wrong one.
How to read the blue bars. The bar is where the true click rate probably sits (Wilson 95% interval); the dark line is what we measured. Two bars that overlap are not distinguishable — telling them apart would take more sending than the answer is worth, and that volume is better spent on signups.
The website half of this test is not measured. A campaign records which email template it used but not which landing page it pointed at, so copy can be compared and pages cannot. That needs a page reference on the campaign before it can be answered.
4 · Follow-up — who to resend to, and what to send them

The resend queuewhere the next signups come from

Everyone emailed for this program who has not signed up, split by how far they got. Each group stopped for a different reason, so each needs a different follow-up.
The school funnel — how many schools reached each stage
16,267
SCHOOLS EMAILED
 
2,254
SOMEONE OPENED
14% of schools
1,605
SOMEONE VISITED
10% of schools
36
SOMEONE SIGNED UP
0.2% of schools
# Group Schools Contacts Expected signups
from one resend
What we know What to send them
1 Visited the website but never registered 1,569 3,429 37 They read the email, clicked, looked at the site and walked away. Highest intent on the list. Do not resend the same email — send a short "you looked at this" note. 0 of them chose the future list because of grade, so exclude those and target the 3,429 who left with nothing.
2 Opened the email but never clicked 649 1,862 1 Subject line worked. The body or the offer did not. Resend with a different body, same subject family. We already know the subject gets through, so this is the group a new copy should be tested against next.
3 No evidence they opened the email 14,013 15,574 7 The email either never got read, or the open pixel was blocked. Resend with a new subject line at a different time of day. Biggest group by far, lowest intent per person.
Everyone still to convert 16,231 20,865 45 Three sends, three different messages
1,569
Schools that visited but did not register
warmest group — resend first
14,013
Schools where nobody opened
needs a new subject line
16,231
Schools with no signup yet
100% of everyone we emailed
45
Signups if we work all three
on top of the 37 we have
⚠ "No evidence they opened" is the least trustworthy number here. Apple Mail Privacy Protection blocks the tracking pixel, so an unknown share of these 15,574 people did read the email. Treat the group as "no evidence they opened it", not "did not open it" — and never write a resend that says "in case you missed my last email", because plenty of them did not miss it.
How the expected column is modelled. It uses rates this program has actually produced — 7.1% click rate and 1.1% visit → signup — not an industry benchmark. A never-opened list is discounted to 60% on the assumption a new subject line lands less well than the original did. These are projections, not measurements.
Resend order matters. Group 1 is far smaller than group 3 but produces more signups per send, because they have already shown they want it. Work down the list, not up.
Two rules for every resend: exclude the 82 unsubscribes and anything that hard-bounced or was blocked — sending to a bad address twice is how the bounce rate climbs toward the 5.0% that damages the domain. And every extra send to the same person converts worse than the one before, so a third attempt is rarely worth the deliverability risk.
5 · Sending health

Deliverability guardrailthe thing that can stop everything

One definition everywhere: bounce = hard bounces ÷ sent, across all sending, not one program — the sending domain is shared. Blocks are counted separately because they mean something different.
0.5%
Hard bounce
pauses at 4.0% · fatal at 5.0%
0.0%
Blocked
a server refused us — reputation signal
3.46%
If we send the 14,543 catch-alls
assumes 12% of them bounce
0
Spam complaints
the fastest way to lose a sending domain
82
Unsubscribed
suppressed permanently
195
Hard bounces
dead addresses — never retry
251
Soft / transient
retryable, excluded from the guardrail
294
Contacts marked bounced
suppressed from every future send
12,207
Verified "unknown"
NeverBounce could not decide — treat as risky
The catch-all decision: 14,543 addresses accepted by servers that accept everything, so verification cannot confirm them. At a normal 12% catch-all bounce rate, sending all of them takes the blended rate to 3.46%. Send a small batch first, measure, then decide — do not send all 14,543 at once.
Bounce type comes from the webhook's own classification (the Postmark bounce type recorded on the send), so a hard bounce is never confused with a full mailbox or a temporary DNS failure.
Not tracked yet: the auto-pause rule itself. Which window it watches (rolling 7-day, per-campaign, per-list) and whether it halts all sending or only the offending campaign is not configured anywhere — the thresholds above are the ones this page reports against, not ones the sender enforces.
6 · Contacts left to email

The list — what is actually sendablemeasured, not estimated

Everything here is a live count from the database: contacts held, how many have been through NeverBounce, and how many valid ones have never been emailed.
31,534
Valid contacts
NeverBounce-confirmed, safe to send
1,367
Never emailed
valid, active, no send on record
14,543
Catch-all, untested
bounce risk — batch-test first
0
Scraped but unverified
cannot be safely emailed yet
3.15
Contacts per school
the number that gates everything
The pool this program draws from, split by distance to a store
Segment Eligible contacts Share of pool Schools Quota to clear
Within the store radius 14,239 24.1% 5,874 not stored — no sponsor quota in the database
Anywhere else 44,904 75.9% not stored
Whole eligible pool 59,143 100% 18,478
61,949
Contacts in the database
every scraped or uploaded contact
61,963
Verified so far
51% of them came back valid
19,016
Schools with at least one contact
of 105,091 schools held
Cost to verify the rest
no credit price is configured
Contacts per school is the lever, not schools scraped. A school only enters a program if at least one of its contacts signs up, so the more contacts we hold per school, the more schools convert — and deepening the schools we already have is cheaper than scraping new ones. We currently average 3.15.
We have more contacts than we can send to. 61,949 are held but only 61,963 have been through NeverBounce, leaving 0 unusable until they are checked. Sending unverified addresses is exactly what pushes bounce toward the 5.0% that damages the domain, so verification, not scraping, is what stands between the list and the send.
Not tracked yet: a scrape target (how many schools we intend to hold in total) and the NeverBounce credit price, so scrape-completion percentage and the cost to verify the remainder cannot be computed here.