Turning Practice Data Into Evidence Your Principal Trusts

You will get about eleven minutes on a leadership agenda, and "the kids love it" is not evidence; it spends all eleven without buying anything. What renews a program is a usage number, a growth number that survives a follow-up question, and an honest sentence about what your data cannot show. In that order.

TL;DR: Lead with usage consistency, the share of students active in at least 8 of 11 weeks. Then median words per minute with accuracy beside it, per student, then two dated writing samples. Name your ESSA tier yourself, concede the volunteer sample, and ask for one grade-level team, not a rollout. School pricing is quoted on request, so attach the written figure.

Most coaches lead with growth because growth is the exciting part. Administrators discount growth first, because they have been handed a lot of it.

The three numbers leadership acts on

Start with consistency. A program that ran four days a week in eight classrooms for eleven weeks exists; a program with a beautiful average and 31 percent participation happened to a few children. LearnPlatform by Instructure's EdTech Top 40, built from 12.6 million K-12 users inside Canvas, found that individual students and teachers interact with an average of four tools, while the same research group reported that districts accessed an average of 2,982 distinct edtech tools in 2024-25. Your principal lives on the wrong side of that gap and knows it. Proving that something is genuinely used is not a modest claim in that environment.

Number	How to state it	Why it lands
Usage consistency	Sessions per student per week, plus the share of enrolled students who hit that in at least 8 of 11 weeks	It cannot be inflated by enthusiasts and it predicts what year two looks like
Accuracy-adjusted growth	Median change in words per minute with accuracy reported beside it, per student, never a bare class mean	Speed without accuracy is the claim administrators have learned to distrust
Work-product change	Two dated writing samples per student, typed, same task type, start and end	It is the only evidence in the packet a principal can read without taking your word for anything

Report the median, not the mean, and say which you used. One child who went from 9 to 46 words per minute will drag a class average somewhere your class never went, and the first person to notice is often the one you most needed to convince. A median with accuracy beside it is a number your principal can repeat to a superintendent without exposure.

What to leave out

Leave out the fastest student. Leave out percentage gains computed from tiny starting numbers, because a child moving from 6 to 12 words per minute is a 100 percent improvement and also still a child who cannot write a paragraph. Leave out screenshots of leaderboards.

Leave out, above all, the vendor's headline statistic, including ours. Across Meta Typing Club's whole user base, learners practicing daily average roughly 10 words per minute of gain per month, and that figure is not segmented by age, starting point, keyboard layout or language. It is a company-wide average dominated by adults and older students. If you put it in a memo about second graders, the first question will be which of those things it controls for, and the honest answer is none of them. Use your own eleven weeks of data instead. It is smaller, and it is yours.

Name your evidence tier before someone else does

Districts increasingly speak in the vocabulary of the Every Student Succeeds Act, and you can borrow it to your advantage. The Department of Education's guidance Using Evidence to Strengthen Education Investments sets out four tiers: strong, moderate, promising, and demonstrates a rationale. As Ohio's ESSA clearinghouse records the statutory wording, promising evidence requires "at least one well-designed and well-implemented correlational study."

Your pilot has no comparison group. It is Tier 4, and saying so in your own memo is the most persuasive sentence you will write, because it shows you know the difference between what you observed and what you proved.

There is a further limit worth conceding in the same breath. The teachers who ran your pilot volunteered, so your adoption figure is measured on the friendliest possible sample. Say it, then propose the fix: one grade-level team next term that did not raise its hand. Conceding it costs you nothing your principal had not already noticed.

What the platform gives you, and what you assemble yourself

Practically, the reporting exists at per-student level. In Meta Typing Club, homework can be assigned to a class, several classes or named students with optional due dates, and the teacher view returns per-student completion, accuracy, words per minute and stars. Speed is reported in words per minute only, with no characters-per-minute alternative, which is worth knowing before you promise leadership a metric that does not exist. Lesson progression sits across more than 2,400 structured lessons, so "where the class actually is in the sequence" is a real answer rather than an estimate.

What no dashboard produces is the writing samples. Those you collect yourself, on paper, in week one and week eleven, same prompt type, dated. It costs two class periods across a term and it is consistently the exhibit administrators read first.

Digital Promise's pilot planning guidance asks districts to draft "a statement that describes what success would look like at the end of the pilot" before implementation. If you skipped that step, write it now for the next cycle rather than reverse-engineering one from your results. Everyone in the room can tell the difference.

The one-page memo

One page. Numbers, limits, ask. Adapt the shape:

Between [dates], [n] students in [n] classrooms used a structured keyboarding program for [x] sessions per week. [n] of [n] students met the usage threshold in at least 8 of 11 weeks. Median words per minute moved from [a] to [b] with median accuracy moving from [c] to [d], reported per student; the full table is attached. Attached also are start and end writing samples for every participating student. Limits: participating teachers volunteered, there was no comparison group, and this evidence demonstrates a rationale under ESSA rather than an effect. Teacher time cost was [x] minutes per week, reported by participating teachers. Ask: extend to [grade-level team] next term, including at least one class that did not volunteer. Written per-student annual quote for that extension: attached.

The ask at the end is deliberately small. Nobody scales a program off eleven weeks, and requesting that they do is how a good pilot gets refused.

Key takeaways

• Lead with usage consistency, not growth. It is the number that cannot be inflated by enthusiasts.
• Report median words per minute with accuracy beside it, per student, and say you used the median.
• Two dated writing samples per student outperform any dashboard export as evidence a principal can read directly.
• Do not quote vendor averages, including our roughly 10 words per minute per month, which is a whole-user-base figure with no age or starting-point controls.
• Name your ESSA tier yourself. Local pilot data demonstrates a rationale; it does not demonstrate an effect.
• Concede the volunteer sample problem and propose the non-volunteer class as the fix.
• Ask for one extension, not a district rollout, and attach the written per-student quote; school pricing is on request, so request it the week you start the memo.

Frequently asked questions

What if our usage numbers are mediocre?

Report them and diagnose them. Low usage caused by a scheduling collision is a fixable finding and reads as competence; low usage hidden behind a growth average reads as advocacy the moment anyone opens the file.

How long should a pilot run before I write this memo?

Long enough to cross a grading period and at least one disruption, which usually means ten to twelve weeks. Shorter than that and the usage number is measuring novelty.

What does this cost per student per year if we expand?

There is no published school price for Meta Typing Club. Free covers personal use: individuals, families, and an independent teacher running one class of up to five students on their own account, which is an evaluation, not a pilot. A pilot across several of your school's classrooms is school use, which is commercial and priced on request. Send the quote request through contact the week you draft the memo, so the written per-student figure arrives before the memo does.

Will leadership ask about student data privacy?

Yes, usually in the same meeting. Have the short version ready: which data the platform holds, who created the student accounts, and whether the district's data privacy agreement has been executed. Our platform collects no age and no date of birth, and account deletion is by email request rather than self-service. Both belong in your answer.

Can I use per-student data in a public presentation?

Not with names attached. Aggregate or anonymize for anything leaving the building, and keep the identified per-student table as an attachment for internal review only. The reporting views are built for teachers, not for public display.

The program does not earn its next term by being liked. It earns it by producing a page that a principal can defend to a superintendent without you in the room.