Ian — the Bilateral Core Constraint Model is exactly the right frame for why the procurement dysfunction I'm describing isn't accidental. The B2G sales dynamic and institutional risk aversion aren't bugs in the system; they're load-bearing walls. The $17 Revolution essay I wrote in response tries to locate where the edge moves while the institution deliberates — not as a replacement for the structural fix, but as what happens in the meantime. The students who are ready don't have the luxury of waiting for the procurement model to reform. Appreciate you putting the structural diagnosis on the table so clearly.
Well, dang! That's one solid tree from the seed I offered.
I've got a few more industry-side perspectives in the queue to pull John Meynard Keynes and Milton Friedman into the current EdTech kerfuffle. There were calls from every President and mountains of money appropriated by Congress based on the "Students Need to Be Taught the Latest Technologies so That They’re Prepared for the Jobs of the Future" orthodoxy.
Thank you — and I'd genuinely like to hear your perspective. The equity dimension looks different depending on where you're standing. The cost-effectiveness data I'm citing is almost entirely drawn from U.S. and OECD contexts. The implementation conditions that make intelligent tutoring systems work — reliable connectivity, device access, trained teachers — aren't evenly distributed even within wealthy countries, let alone across the global South. What are you seeing on the ground?
African Edtech Innovator commented on your post The $165 Billion Question: What The Economist Got Right (and Terribly Wrong) About Education Technology.
Great rebuttal article! The article raises an important question about the $165B invested in EdTech and whether it is translating into learning outcomes. That scrutiny is healthy for the field.
One issue I’ve been writing about in my Impact Reality blog series is what I call Evidence Debt. Many products are validated at a single point in time, but the product, implementation, and classroom context evolve quickly. When the evidence doesn’t evolve with them, a gap forms between what the research once proved and what is actually happening in classrooms today.
Another challenge is that the debate often treats “EdTech” as one category, when outcomes vary widely depending on the instructional use case. The real question isn’t whether technology works, but under what instructional conditions it improves learning.
That’s why the next phase of impact measurement needs to move beyond traditional ROI toward what I’ve called ROI², Return on Instruction. The critical signal isn’t just whether a tool was purchased or adopted, but whether it changes the instructional routines that drive learning: feedback cycles, formative assessment, and targeted practice.
The future of EdTech accountability will depend less on one-time validation studies and more on ongoing evidence tied to real classroom implementation. In a field evolving this quickly, static evidence quickly becomes outdated. Continuous evidence will be the standard the sector ultimately moves toward.
Melissa, 'Evidence Debt' is exactly the right term. It highlights the intellectual dishonesty of using outdated pilot studies to justify massive, stagnant contracts. I love the shift to ROI² (Return on Instruction). It moves us away from the binary 'does tech work?' and toward 'under what conditions does it empower the teacher?'
We need to stop treating EdTech as a product and start treating it as a process. If it doesn't shorten the feedback loop or sharpen the practice, it's just a digital paperweight
John — yes, and I'd push one step further: the training problem is downstream of the procurement problem. Districts buy first, train second (if at all), and the pedagogical framework never gets built because the vendor's incentive ends at the sale. The IXL case in Kansas illustrates it precisely — the software has documented efficacy in controlled conditions. It failed at McPherson because implementation support was essentially nonexistent after purchase. The tool wasn't the problem. The handoff was.
Schools are highly idiosyncratic, which explains why educational reforms don't transfer well. Before entering education (from finance), I would have thought that schools budgeted for implementation and training. Most major EdTech tools (e.g., SIS, LMS) require significant implementation and training. I think schools don't take "smaller" EdTech tools seriously enough.
I've evaluated educational technology programs for the past 25 years or so, and this article tracks with nearly everything I've seen. A really great piece!
Please share the link when it's up. The field has no shortage of critiques of what doesn't work — what's genuinely scarce is concrete, practitioner-tested descriptions of what does. If you have examples of implementation that held together past the pilot phase, that's worth documenting in detail.
My modest offer to the discussion is The Bilateral Core Constraint Model of Institutional EdTech. https://www.linkedin.com/pulse/postmodern-era-edtech-united-states-ian-mccullough-bwvwc/#:~:text=what%20it%20is.-,Seeing%20the%20Bigger%20Picture,-Zoom%20out.%20Look
Ian — the Bilateral Core Constraint Model is exactly the right frame for why the procurement dysfunction I'm describing isn't accidental. The B2G sales dynamic and institutional risk aversion aren't bugs in the system; they're load-bearing walls. The $17 Revolution essay I wrote in response tries to locate where the edge moves while the institution deliberates — not as a replacement for the structural fix, but as what happens in the meantime. The students who are ready don't have the luxury of waiting for the procurement model to reform. Appreciate you putting the structural diagnosis on the table so clearly.
If you haven't clicked on the image or checked the endnotes yet, I did a bit of work with my $240 year Claude subscription to try to get some context for what poured out of my brain. https://claude.ai/public/artifacts/de2fec07-287e-4ff1-9e4b-04c36a666023
Interesting. The essay inspired me to write this. The $17 Revolution Nobody Is Waiting For https://open.substack.com/pub/learningengineering/p/the-17-revolution-nobody-is-waiting
Well, dang! That's one solid tree from the seed I offered.
I've got a few more industry-side perspectives in the queue to pull John Meynard Keynes and Milton Friedman into the current EdTech kerfuffle. There were calls from every President and mountains of money appropriated by Congress based on the "Students Need to Be Taught the Latest Technologies so That They’re Prepared for the Jobs of the Future" orthodoxy.
When successive Presidents of the United States issues calls and Congress appropriates funds for specific purposes… yeah, businesses are gonna respond and pursue that money. https://drive.google.com/file/d/1XKD8b4x6t5_hC5uQnGh79LcdKrVqpCTi/view?usp=sharing
Interesting read.
Thank you — and I'd genuinely like to hear your perspective. The equity dimension looks different depending on where you're standing. The cost-effectiveness data I'm citing is almost entirely drawn from U.S. and OECD contexts. The implementation conditions that make intelligent tutoring systems work — reliable connectivity, device access, trained teachers — aren't evenly distributed even within wealthy countries, let alone across the global South. What are you seeing on the ground?
Nik Bear Brown - Computational Skepticism
African Edtech Innovator commented on your post The $165 Billion Question: What The Economist Got Right (and Terribly Wrong) About Education Technology.
Interesting read.
Great rebuttal article! The article raises an important question about the $165B invested in EdTech and whether it is translating into learning outcomes. That scrutiny is healthy for the field.
One issue I’ve been writing about in my Impact Reality blog series is what I call Evidence Debt. Many products are validated at a single point in time, but the product, implementation, and classroom context evolve quickly. When the evidence doesn’t evolve with them, a gap forms between what the research once proved and what is actually happening in classrooms today.
Another challenge is that the debate often treats “EdTech” as one category, when outcomes vary widely depending on the instructional use case. The real question isn’t whether technology works, but under what instructional conditions it improves learning.
That’s why the next phase of impact measurement needs to move beyond traditional ROI toward what I’ve called ROI², Return on Instruction. The critical signal isn’t just whether a tool was purchased or adopted, but whether it changes the instructional routines that drive learning: feedback cycles, formative assessment, and targeted practice.
The future of EdTech accountability will depend less on one-time validation studies and more on ongoing evidence tied to real classroom implementation. In a field evolving this quickly, static evidence quickly becomes outdated. Continuous evidence will be the standard the sector ultimately moves toward.
Melissa, 'Evidence Debt' is exactly the right term. It highlights the intellectual dishonesty of using outdated pilot studies to justify massive, stagnant contracts. I love the shift to ROI² (Return on Instruction). It moves us away from the binary 'does tech work?' and toward 'under what conditions does it empower the teacher?'
We need to stop treating EdTech as a product and start treating it as a process. If it doesn't shorten the feedback loop or sharpen the practice, it's just a digital paperweight
The problem is a lack of training, limited best practices, and no pedagogies built-in
John — yes, and I'd push one step further: the training problem is downstream of the procurement problem. Districts buy first, train second (if at all), and the pedagogical framework never gets built because the vendor's incentive ends at the sale. The IXL case in Kansas illustrates it precisely — the software has documented efficacy in controlled conditions. It failed at McPherson because implementation support was essentially nonexistent after purchase. The tool wasn't the problem. The handoff was.
Schools are highly idiosyncratic, which explains why educational reforms don't transfer well. Before entering education (from finance), I would have thought that schools budgeted for implementation and training. Most major EdTech tools (e.g., SIS, LMS) require significant implementation and training. I think schools don't take "smaller" EdTech tools seriously enough.
I'm going to do a blog post about what good EdTech looks like (johnfaig.medium.com)
I've evaluated educational technology programs for the past 25 years or so, and this article tracks with nearly everything I've seen. A really great piece!
Please share the link when it's up. The field has no shortage of critiques of what doesn't work — what's genuinely scarce is concrete, practitioner-tested descriptions of what does. If you have examples of implementation that held together past the pilot phase, that's worth documenting in detail.