The Evidence Standard Is Coming for EdTech. Childcare Software Should Meet It Too
On August 20, 2026, the U.S. Department of Education told schools something the software industry did not want to hear: technology must prove it works. The guidance targets K-12 classrooms, but the standard behind it applies just as well to the software running your childcare center. Here is what it says, and how to use it the next time a vendor shows you a demo.
What the Department of Education Actually Said
The guidance arrived as a Dear Colleague letter from the Department's Office of Elementary and Secondary Education (press release). Its core message: technology use in classrooms should be purposeful, instructionally sound, and contributing to measurable outcomes. To get there, it says every education technology product should be able to answer five questions:
- What learning problem does it solve?
- When should it be used?
- For whom should it be used?
- For how long should it be used?
- What evidence demonstrates that it improves student learning?
And it does not stop at purchase time. The letter tells schools to review effectiveness regularly, and it is blunt about what happens when the answer is bad: when evidence shows a tool is not improving learning, schools should change course, and when shortcomings persist, remove it altogether. Coverage in K-12 Dive and EdSurge has framed it the same way: evidence is becoming a procurement requirement, not a nice-to-have.
Why This Matters Outside K-12
To be precise about scope: this guidance is written for K-12 schools and instructional technology. Childcare centers are not the audience, and childcare management software is operational, not instructional. Nobody in Washington is grading your check-in flow.
But the dynamic the Department is pushing back on will be familiar to any director who has sat through a software demo. For years, the industry has been rewarded for shipping features, announcing AI, and growing adoption. None of those are the same thing as making your Tuesday morning easier. A product can have impressive technology and still fail to change a single outcome you care about: whether you are in ratio, whether every child is accounted for, whether the claim went out complete, whether a parent got the message.
The federal government just raised the bar for one education market. Directors can raise it for theirs today, for free, by asking the same five questions. Here they are, translated for a childcare center.
The Five Questions, Translated for Childcare
1. What operational problem does it solve?
Not "what does it do" but what specific problem disappears. Name the moment: knowing your ratio status at 7:30 AM without walking every room, a check-in that takes seconds with an audit trail, a CACFP claim that is already reconciled on claim day. If the vendor cannot name the moment, the feature is decoration.
2. When should it be used?
Good childcare software lives inside the day: at the door during arrival, in the classroom during meals and naps, at the desk when a parent calls. Software that only gets opened after the children leave is admin homework with a subscription fee. Ask to see the product at the busiest moment of the morning, not in a quiet demo.
3. Who should use it?
A director-only tool moves paperwork; it does not change the center. Ask what teachers do in it daily, what families see, and whether a lead teacher can run their room from it without training. Then ask the harder version: who should NOT use each surface? Role boundaries are a feature, not a limitation.
4. How long should it take?
Time is the honest currency of a childcare center. Every workflow should be measured in it: setup measured in days not months, check-in measured in seconds, a billing run measured in minutes. If a vendor cannot tell you how long their own workflows take, they have never measured them.
5. What evidence shows it works?
The Department tells schools to look past vendor claims to third-party evidence and local outcome data. Your version of that: does the product show you, on screen, the numbers that prove it is doing its job? Live ratio status per room. Attendance counts that reconcile. Claim totals with the math visible. Overdue balances that update themselves. If the product cannot show its own impact, no case study will.
How to Use This in Your Next Demo
You do not need a procurement office to run this playbook. Three habits get you most of the value:
- Ask the five questions out loud, in order. Vendors are prepared for feature questions. They are less prepared for "what evidence shows this works." The quality of the answer tells you more than the demo does.
- Pilot against a number, not a feeling. Before a trial, pick two or three measurable targets: minutes at morning check-in, hours to produce your CACFP claim, days of tuition outstanding. Measure the week before and the last week of the trial. Keep the numbers.
- Be willing to change course. The Department's hardest advice applies to centers too. If your current software is not moving your numbers, that is evidence. Our guide on switching childcare software covers how to move without losing your data or your families.
One caution as AI arrives in this industry: "AI-powered" is a description of technology, not of impact. The same five questions apply. An AI feature that cannot name the problem it solves, the moment it is used, and the number it moves is engagement, not evidence. We have written about how we think AI should behave in childcare settings in our childcare AI safety guide.
Holding Ourselves to the Same Standard
We build childcare software, so this standard points at us too. Here is our commitment: we will always answer the five questions about our own product, and the fifth one with numbers you can check yourself.
That is why Neztio is built to show its work on screen. Room ratios update in real time as children check in and out, so "are we in ratio" has a live answer with a timestamp. Attendance carries a full audit trail. CACFP meal counts accumulate as teachers log meals, and the claim summary shows the totals behind the reimbursement estimate. Billing shows outstanding balances computed from the invoices themselves. During a 30-day trial, all of that is your local outcome data: measure your before, measure your after, and decide on the evidence.
We would rather be evaluated on that bar than on a feature list. If the whole industry gets held to it, directors win, and so do the centers that children walk into every morning.
The Bottom Line
The Department of Education just told K-12 schools to demand evidence from education technology: five plain questions, reviewed regularly, with the willingness to walk away when the answers fail. Childcare directors do not need to wait for anyone to issue that guidance for their market. The questions travel, the standard is free, and the vendors who welcome it are the ones worth trialing.
Start with our guide on choosing childcare software, or put Neztio to the test.
Related
How to Choose Childcare Software: A Director's Guide
Related
Childcare AI Safety: How AI Should Behave Around Children's Data
Glossary terms in this article
Childcare Management Software
Digital platforms that help childcare centers manage enrollment, attendance, billing, communication, and compliance.
Staff-to-Child Ratio
The number of caregivers required per group of children, set by state licensing regulations based on age.
CACFP
The Child and Adult Care Food Program, a federal program that reimburses childcare providers for qualifying meals and snacks.