What happens on the call

Thirty minutes on Zoom, nothing to prepare.

  • You tell Gina what you are building and what is getting in the way. Three short questions when you book help the call start where you are.
  • Together you name the real problem and one next move, whether or not we work together.
  • If it is a fit, a one-page scope note with the fee arrives within two business days.

Book a discovery call

This week at EdTech Week, held at Teachers College, Columbia University, I spent most of my time listening. Many sessions focused on young children and on AI, but the advice I heard applies well beyond both. If you are building anything people learn from, these are the lessons I would keep.

1. Lead with the relationship your product supports

Kathy Hirsh-Pasek, a Temple University professor and senior fellow at the Brookings Institution, opened her panel with a simple reminder about young children:

“It’s all about human connection.”
Kathy Hirsh-Pasek, EdTech Week 2026

She warned against products built to stand in for people. Later in the session, the panel made the point that it is not hard to build a tool that strengthens social connection instead of dampening it. Brandi Kenner of Bezos Academy, a developmental psychologist and learning scientist, added that the best use of new technology is to build adult capacity, so teachers and parents can do more of what works.

For any learning product, ask who the learner learns with: a teacher, a parent, a manager, a peer. Then design so your product makes that relationship stronger.

2. Evidence first: build on what learning science knows

Hirsh-Pasek pointed out that researchers and product teams work on different clocks. Product teams want to get to market; scientists want to wait and see. Her answer was that decades of research already tell builders a great deal, and she invited product teams to bring researchers to the table. She also pointed to the Learning Sciences Exchange, a program she helped start with New America to prepare people who can work across research and practice. Her own work offers a starting point: a 2015 review she led, “Putting Education in ‘Educational’ Apps,” names four pillars of learning: active involvement, engagement, meaningfulness, and social interaction.

3. Ask across silos and learn each other’s language

Kenner described moving from early childhood classrooms into research and then into technology, and having to learn each field’s vocabulary along the way. Her advice was for researchers, educators, and technologists to get out of their silos and learn to translate for one another. Her closing challenge to the room: ask yourself what you need to learn to be more of a connector. For a small team, that can be as practical as having your educator and your developer review the same design together.

4. Recruit learners as co-designers and critics

A panel with The GIANT Room, the Joan Ganz Cooney Center at Sesame Workshop, the Robin Hood Foundation, Sago Mini, and NYC Public Schools shared what happened when they designed with students and teachers instead of only testing with them at the end. The team went into classrooms with questions instead of assumptions. Students came up with game ideas no room of adults would have produced, and they wanted the job of critiquing the tool. In The GIANT Room’s write-up of the project, critiquing the AI’s output became part of the app itself, with sentence starters that teachers first modeled in class.

5. Name the learning goal before the feature

A Robin Hood Foundation leader on the same panel put it in a line worth remembering:

“The technology mattered less than the sequencing.”
Robin Hood Foundation, EdTech Week 2026

Teachers set the instructional goals, the curriculum followed those goals, and the tools were built around both. The panel also pushed back on the idea that technology means a student alone at a screen. Their classrooms used it to support hands-on work and face-to-face collaboration in small groups.

6. Eyes on every output: keep a person in the loop

In that classroom project, a person reviewed every AI output before it reached students. That is a sound rule for any product that generates content for learners. Decide which outputs need human review, who does it, and how quickly, before you launch.

7. Run pilots with real partners, and bring families in early

The district partner on the panel described what makes a pilot worth agreeing to: a chance to try the product, give feedback, and have a say in its purpose, with a team willing to change what does not work for their students and teachers. One panelist found that most students in their classrooms had first used AI at home, on their parents’ devices. The panel agreed that the families or adults around a learner need to know how the product is being used and how to support it. A good pilot plan includes them.

Build for the LEARNER

The seven lessons, one letter each, with a question to ask your team before the next build or pilot.

  1. Lead with the relationshipWhose learning relationship does this make stronger?
  2. Evidence from learning scienceWhich learning research is this design built on?
  3. Ask across silosHave an educator and a developer reviewed this together?
  4. Recruit learners as co-designersWhen did learners last shape or critique a feature?
  5. Name the learning goal firstCan we state the learning goal before the feature?
  6. Eyes on every outputWho reviews what reaches a learner?
  7. Run pilots with real partnersWhat can our pilot partners change?

Read down the first letters to remember all seven.

Get the LEARNER card (PDF): one page to work through with your team, plus a pocket card to cut out and keep.

An example from my own work

These lessons match what I learned early in my own research. In my doctoral study with first-year special education teachers, I started by listening. The teachers described isolation, limited materials, and a short training period, and the response was a small peer community. It was a small, one-semester study with four teachers, so I hold it as a lesson about process rather than proof. The lesson was the same one I heard on stage this week: the people who will use the thing should help shape it.

What an education product engagement looks like

When StrideForward works with a small team building a learning product or program, the work usually includes:

  • A clear case for why it works: a short research brief and a logic model that ties the design to the science of learning.
  • Co-design sessions and a pilot plan: who takes part, what to measure, and how learners, teachers, or families give feedback along the way.
  • Adoption support: onboarding and short professional learning so teachers or managers can use it well.
  • The partnerships and materials to bring it to schools, funders, or employers, built and handed over to your team.

Signs your product needs more input from the people who use it

  • Most design decisions are made by adults on your team, without learners in the room.
  • You can list features faster than the learning goal they serve.
  • Teachers like the demo but stop using it after a few weeks.
  • No researcher or experienced educator has reviewed the design.
  • The adults around the learner do not know how to support it.

If two or more of those sound familiar, plan a small co-design session before your next build cycle.

Apply it

Try this before your next build cycle

  • Write the learning goal in one sentence before you write a feature.
  • Name the human relationship your product should strengthen.
  • Ask one researcher or experienced educator to review the design.
  • Run a session where learners use the product and critique it, and write down their exact words.
  • Decide which outputs a person will review before they reach a learner.
  • Plan how teachers, families, or managers will hear about it.

Tip: your checks save on this device and land in your notes below. Prefer paper? Print the page. (We will not tell anyone you still print things.)

Your notes

Everything you write in the Your turn boxes, and every step you check off, collects here so you can take it with you.

Nothing here yet. Answer a Your turn prompt or check off a step above, and it will show up here.

Your notes stay in this browser until you send them. Texting and your email app use your own apps, so you choose where the notes go.

Know someone who needs this? Send this field note to a teammate.

Plan the pilot on one page.

The free Pilot Plan walks through seven questions to answer before you build: the first group, what should change, who owns each step, and the decision at the end.

Get the Pilot Plan

Common questions

Does this advice only apply to products for young children?

No. Many EdTech Week sessions focused on young children, but the same lessons hold for courses, curricula, workplace training, and apps for older learners: start with the relationship, build on learning science, design with the people who will use it, and keep a person in the loop.

What does co-design mean?

Co-design means learners, teachers, or other users help shape the product throughout development as partners, instead of only testing a finished version. In practice it looks like sessions where they try early versions, suggest ideas, and critique what they see.

How do I find a researcher to review my product?

Start with a nearby university’s education or psychology department, or an organization that already runs programs like yours. Be specific about what you want reviewed, how much time it will take, and what you will do with the feedback.

Does StrideForward help with this?

Yes. We help small education teams build the research case for a product, plan co-design sessions and pilots, and prepare the materials and partnerships that schools and funders ask for.

Sources and further reading

Keep going with StrideForward

Building something people learn from?

Bring the product or program you are working on. In thirty minutes we will look at the learning goal, who should help design it, and one practical next step.

Book a discovery call