How school districts actually buy software

Most software sold to US schools is designed by people who have never been inside a district. That isn't a criticism. It's a market where the buyer, the user and the payer are three different people, and nothing about your product surface tells you that.

This is the org chart, the money and the calendar. It's the least technical post in the series and probably the most immediately useful, because almost every avoidable failure here comes from not knowing which of those three you were up against.

The shape of the thing

The United States has roughly 13,000 public school districts, and they aren't branch offices of anything.

A local education agency, or LEA, is the district. It's a unit of government with its own board, its own budget and its own purchasing authority. Above it sits the state education agency, which sets standards, runs assessment and collects reporting, but generally doesn't buy your software. Below it sit schools, which adopt but rarely purchase.

The size distribution is brutal and worth sitting with. A handful of districts serve hundreds of thousands of students. The median serves fewer than 1,500. Build for the median and you can't survive the largest. Build only for the largest and you've priced out most of the market.

The technical version of that: a nightly roster diff that's trivial at 1,500 students still has to finish inside its window at 200,000 students, 15,000 staff and several million enrolment rows.

The five people who decide

You'll meet them in roughly this order.

The curriculum director or Chief Academic Officer owns instructional fit. Does it teach what the state requires? Does it fit the scope and sequence already adopted? What's the evidence base? This is the person your demo is for.

The CTO or director of technology owns integration and devices. How do you roster? What happens to the help desk in September? Does it run on the Chromebooks we already own? Usually the most technically literate person in the process, and used to vendors who can't answer basic questions about their own integration.

The data privacy officer owns the review. In a large district this is a full-time job. In a small one it's the CTO wearing a second hat and working from a state template. They want a data inventory, a subprocessor list, a retention schedule and a signed agreement. They are the most common point of death for a deal.

The accessibility coordinator, sometimes the special education director, owns conformance. They want a VPAT, and increasingly they read it.

The business manager owns the fund and the paperwork. Which budget line is this? Are you on a cooperative contract? Will the quote still be valid when the purchase order is actually cut?

The superintendent and school board approve rather than evaluate. Board meetings have dates, agendas close in advance, and a contract that misses one waits a month. Nobody will tell you this in time.

Which pocket is allowed to pay

Districts don't have one budget. They have a general fund plus restricted funds, each with rules about what it can buy. Getting this wrong isn't a pricing problem, it's a legality problem, and the business manager will treat it as one.

Fund	What it's for	Watch for
Title I-A	Schools with high concentrations of students from low-income families	Supplemental services only
Title II-A	Educator quality	Professional learning, coaching
Title IV-A	Student support and academic enrichment	Most common home for tech. No more than 15% of the technology share may go to devices and infrastructure
IDEA Part B	Special education	Tied to services for students with disabilities
Perkins V	Career and technical education	CTE programmes only
E-Rate	Connectivity and internal connections	Does not buy courseware. Suggesting it does marks you as new

One piece of recent history still shapes every conversation. The federal pandemic relief funding had to be obligated by 30 September 2024 and liquidated shortly after. An enormous amount of EdTech was bought on that money. Anything sold against it is now being re-justified against a general fund that never grew to replace it, which is why renewals in this market got noticeably harder from 2025 onward.

Two mechanisms that shorten everything

These are worth more to a young company than most features.

Cooperative purchasing. A district can often buy from a contract somebody else already competitively bid: Sourcewell, BuyBoard, TIPS, NASPO ValuePoint, Omnia Partners. One award can serve hundreds of districts without a new RFP. Teams who don't know this run the same bid repeatedly and describe it as a long sales cycle.

The National Data Privacy Agreement. Districts increasingly use the Student Data Privacy Consortium's standard agreement with a state exhibit instead of drafting their own. Sign it once including Exhibit E, the general offer of privacy terms, and other districts in that state alliance can adopt the same terms without renegotiating.

Beyond that, boring artefacts matter more than they should. A W-9, a certificate of insurance, a quote written to survive the fiscal year, a purchase order number you can invoice against. Having them in the first email instead of the fourth compresses weeks.

The calendar is the roadmap

This is the part that most directly constrains engineering, and the part most product plans ignore.

July to August: rollover. New year, new sections, new schedules. Teachers aren't assigned to classes yet. Students aren't fully enrolled. The roster changes nightly. The first bell doesn't move.

August to September: churn. Schedules settle over about three weeks. Every mishandled transfer becomes a help desk ticket in week two, and week two is when a district decides whether you were a good idea.

October to December: the real usage window. Adoption is won or lost here. Your instrumentation needs to answer "is anyone actually using this" by November, because that's when the question gets asked.

January to March: budget and renewal. Next year is budgeted now. Renewal gets decided on usage data you should already have shown them.

March to May: the testing freeze. State assessment windows. Nobody wants releases, interface changes or new sign-in flows anywhere near a proctored test.

May to June: fiscal close. Purchase order deadlines, a 30 June year end in most states, and the summer build for a year starting in six weeks.

Three things fall out of that immediately:

• Your riskiest deploys belong in October and November. More on that in engineering for the school year.
• Your rostering work has to be finished and rehearsed before July.
• Your support capacity needs to peak in the three weeks when your usage numbers are lowest.

What to do with this

Ask which fund early. "Which budget is this coming from?" isn't rude, and the answer tells you the timeline, the cap and the paperwork.

Publish your integration and privacy answers before anyone asks. The CTO and the privacy officer are both working from a checklist. A vendor who hands over a completed checklist is doing their job for them, and that gets remembered.

Shape the roadmap around the year. Not out of deference. Because a district's tolerance for change genuinely varies by a factor of ten across it, and shipping into a low-tolerance window turns good work into a support incident.