Afterschool is still very manual.

Like, very very manual.

We asked 142 aides, clerks, coordinators, and site leads on California Expanded Learning teams how attendance actually gets taken and how long it costs them every week. They answered before we trained them, so nobody was describing our software. Here is all of it.

728hours a week

of manual attendance admin, across 142 people, on top of running the program

That is the total of what they told us they spend on taking attendance and on the reporting that follows, added up across everyone who answered.

18 jobsWhat 728 hours a week comes to at 40 hours: eighteen full-time people, doing paperwork that already exists somewhere else.

26,208 hrsAcross a 36-week school year, from this group alone. Nobody budgeted for it and no purchase order shows it.

Share of these staff giving up three hours a week or more. This is not a few overloaded outliers. It is the job.

Every person we asked, and what their week costs them

One dot is one person. They are grouped by how many hours a week they said the manual work takes, adding their answer for attendance to their answer for reporting.

The reporting is heavier than the attendance

The same people, answering about three parts of the job. Each bar is everyone who answered that question, split by how long they said it takes.

1–2 hours a week 2–4 hours 4–6 hours

It lands hardest on whoever owns the record

The same answers again, sorted by job title instead of by task, using the longest band each person picked. Sorted by how heavy the group's typical week is.

tops out at 1–2 hours tops out at 2–4 hours reaches 4–6 hours

What they wrote about, unprompted

Two questions had no answer choices at all: what is most painful, and what you want most. These are the themes across all 142 written answers, with what we take from each.

How we see the problem

The charts are what staff said. This is what we think they add up to, and it is the argument behind what we build.

The record gets made twice, and software did not stop that

Almost every site already had something digital. It did not remove the second pass. Attendance is captured once where the children are, on a sheet, a clipboard, a phone, a memory, and then entered again somewhere it counts. Buying software solved the storage problem and left the capture problem alone.

That gap is where all 728 hours come from. It is also why so few people said their tools were the problem. Their tools are fine. The handoff between them is not.

The door is an identity problem, not a form problem

What staff describe at pickup is not slow typing. It is asking a stranger who they are, matching an adult to a child, reading a signature written in a hurry, catching the parent who signed the wrong line. Faster forms do not fix any of that. Knowing who is standing there does.

It is why we keep coming back to a PIN and a name the staff member can see at the door, rather than another field to fill in later.

Reporting is reconciliation, and it lands on one person

The reporting hours are not spent writing a report. They are spent making three records agree: a paper sheet, whatever the system holds, and what a funder or auditor needs to see. The work is comparison, not composition. When those three sources are made by different people at different times, they will not match, and someone spends a Thursday finding out why.

That someone is usually a clerk or a coordinator, which is exactly the shape of the third chart.

Attendance is the safety system nobody calls a safety system

When we asked about attendance, some of what came back was not about attendance at all. A child who skipped check-in and went to the playground. A headcount that broke because someone was in the bathroom. Not knowing who is on campus during an emergency. The attendance record is the only live list of which children are on site.

That reframes what the record is for. A roster that is accurate the next morning is a compliance artifact. A roster that is accurate right now is a safety tool, and it has to be right at 3:15, not at 5:00.

If a correction is hard, staff keep a paper backup forever

The fastest way to keep paper alive is to make fixing a mistake require somebody else. Staff told us about corrections they could not make without calling an administrator, and about systems that logged them out and lost the entry. People who cannot fix their own errors write things down twice, on purpose.

So an edit trail is not a nice-to-have feature. It is the thing that lets a site put down the clipboard.

The people doing this work usually did not choose the tool

Software gets picked by a cabinet, a director, or a business office. It gets used by an aide holding a clipboard in a cafeteria doorway. Those two groups have never once described the job to us the same way. Leadership talks about one audit-safe number. Front-line staff talk about the parent in front of them and not losing a child.

Both are real. But if the second group cannot get through pickup with it, the first group will not get their number either. That is why we survey the whole team before launch and not only the person who signed.

What this proves, and what it does not

We would rather you trust the parts that hold than take the whole thing on faith.

What it shows

The burden is real, measurable, and concentrated on the people who own the record.
Existing software does not remove it. Most of these sites already had some.
Reporting costs more than taking attendance, and the cost arrives after the children leave.
Staff describe safety problems when asked about attendance, without being prompted.

What it cannot show

It is not a statewide estimate. These are teams that chose to work with us.
Hours are self-reported in bands, and 4–6 was the highest we offered, so the tail is cut off.
It measures the before, not the after. It cannot prove any tool lowers these numbers.
Survey versions asked slightly different questions, so registration has a smaller base.

How this fits Attendly

We build for the people doing this work, not only the people who buy the software. Ending administrative agony only counts if the record stops taking a second pass after the kids go home. What staff tell us before launch is what we ship against, and we ask them again afterwards.

How we learn this

Before Attendly launches with a district partner, we ask the people who will use it every day how things are going. What takes time. What hurts. What they want gone. Same questions every time, so the answers stack up.

Many front-line staff do not get to pick Attendly. Someone else does. So we listen to them and build for them.

This page is the rollup of those answers, including from a district whose pilot later ended. Their afternoons were as real as anyone's. It updates as more replies come in. We publish the pattern. We do not publish names, who said what, or the products they named.

Source: Attendly pre-launch surveys of staff on California Expanded Learning teams, distilled 8 September 2026. 174 submissions, less 18 with no answers recorded, 1 internal test, and 13 repeat submissions where we kept the most recent, leaves 142 people. 139 answered on attendance, 139 on reporting, and 71 on registration, which not every survey version asked; percentages are of those who answered. Weekly hours use the midpoint of each band a person picked, 1.5 for 1–2 hours, 3 for 2–4, and 5 for 4–6, added across attendance and reporting, the two questions everyone was asked. 4–6 hours was the highest band offered, so totals understate anyone above it. The school-year figure assumes 36 weeks and is arithmetic on those same answers, not an observed measurement. Job groups come from the free-text title people typed; 9 people gave titles outside the four groups charted and are still counted in every total. No names, no districts. Updates as more replies come in.