[2026 Re-Post] The Last Mile Problem

Who owns intentionality of implementation?

The Man in the Van

In logistics and communications industries, there’s a thing known as the “last mile” problem. It relates to the challenges of taking the final step of delivery of goods and services to a customer.

In logistics, all of the careful packaging and shipping and warehousing and tracking can fall apart when a box is put into the hands of some guy driving a van, who has to hand-deliver it to someone’s doorstep. Oversight can fall apart during that last mile. A lot of human error or slovenliness or greed or disinterest can come into play—from the van driver to a neighborhood thief to a homeowner who leaves a package out in the rain. The last mile is where systems often surrender to trust.

In communications, the “last mile” involves the final step of connecting a network to an individual user—historically, the running of cable from a central line, managed and maintained by a fleet of workers in trucks, into and through your house, where God-knows-what might happen to it. The cable comes through your wall, and the electric company has no control over what your cat or your baby is going to do to it.

At some point, systems have to surrender to trust, because the end users are not employees and they are not mere receivers; they are users—and the way in which they use a product or a service can be idiosyncratic, or ill-advised, or just dumb. As the purchasers those of goods and services, though, they have every right to be as crazy or stupid about their use as they want to be. They will do what they will do—not, perhaps, what we would wish them to do.

Life in the Egg Crate

We have our own last mile problem in the world of education. Creators of programs and products often refer to it as “fidelity of implementation.” It’s what doctors would call, “usage as prescribed.” We intend for a product to be used in a certain way, because we have some belief (or evidence) that this way-of-using is meaningful and effective. And…then what happens?

I worked for an education technology company that sold a reading program that had decades of solid effectiveness research behind it. We had data from over two million students showing that, when used as recommended, the product could accelerate Lexile (or reading level) scores three to four times greater than expected growth over the course of a school year. That’s a significant and meaningful level of growth for students who are lagging behind their peers. This acceleration rate held true across all demographic sub-groups; no matter what kind of student you were, you could improve and catch up and achieve. It was the most impressive efficacy research I had ever seen, for any company I had worked for. It could change lives.

However. As with any product like this (or any medicine, to go back to the doctor analogy) using-differently-than-directed may lead to sub-optimal results. In our research, we could see that when students didn’t read at least one article in the program each week, every week, or when they didn’t put some effort into answering questions correctly, the growth rates shrank. Some students used the product once a month. Some used it twice a year. Some raced through, selecting “C” for every answer, because they didn’t care. Those students saw no growth at all.

Anyone want to bet on many students used the product haphazardly or occasionally, versus how many used it consistently and to good effect? If you’re a gloomy, cynical pessimist, you’ll win this bet. About 10% of our students were using the program as recommended and getting the maximum possible growth out of it. A majority of users were barely “users” at all. Most of the students did little with the program and got little out of it. Implementation without intentionality made it, to be quite blunt, a waste of time and money.

My former company is hardly the only one dealing with this kind of discrepancy. Peter Greene cites a study in a blog post, showing even worse fidelity numbers for some very well-known products—one as low as 4.7%. That’s right—not even five students out of a hundred were getting the real benefit of the program the school had paid for. And this is a program that is in many American schools.

If a school is going to purchase something that can help students, whether it’s a product for classroom use or a training in a new instructional approach, the teachers need to implement that something with a sense of intention and purpose if it’s going to have any hope of being worth the investment. I like the word “intention” much better than “fidelity.”

But, as Greene puts it:

If you hand me a tool that has been made so difficult or unappealing to use that 95% of the “users” say, “No, thanks,” I’m going to blame your tool design. 

Teachers have their own intentions and purposes and priorities. They owe publishers and software companies no “fidelity.” If they don’t like a product or a new instructional strategy, they’re not going to use it. If they find the kids don’t like it, they’ll stop forcing it on them. Teachers don’t care how much their administrators may have paid for the thing; it’s not coming out of their salaries.

This seems to be exactly what happened recently to Khanmigo, the much-touted AI tutor-bot developed by Khan Academy. “Brilliant!” said its inventors. “Meh,” said the kids. Between the idea and the reality, as T.S. Eliot said, falls the shadow.

This last mile problem we have in education is old and baked into the structure of how we do schooling. For decades, we treated schools like shopping malls, where every room was an entity unto itself, run independently and with little accountability to any central authority. This is something many teachers have found attractive about the profession over the years—the ability to close your door, do as you please, and be left alone.

Some researchers call this an “egg crate,” approach: lots of little eggs, each in its own little protective cocoon. The more we have tried to standardize teaching, the more we have forced accountability measures and mandates and metrics upon our teachers, the less attractive the profession has become to many of them. Mission-driven and creative people don’t go into teaching in order to read a script or follow daily commands. Many find ways to ignore or work around those commands…or they leave the profession.

So, what can we do? How can those of us who make things for use in schools address this last mile problem that undermines the use of what we create?

Well, we’ve tried to provide initial and ongoing training. We’ve tried to bake in as much guidance and help as possible, to make up-front training less and less necessary. We’ve tried to do user research and user interviews to make sure our products are not “difficult or unappealing to use,” as Greene puts it.

Are there garbage products on the market? Of course there are—in every market. Caveat emptor, administrators: do your due diligence and don’t foist crap on your teachers and kids. Some stuff is just bad. Some stuff is well-intentioned but poorly designed or built. That’s not particularly interesting to me.

The challenges faced by good actors, making good product, trying to make a real difference—that’s what keeps me up at night—because I’ve been around for a while, and I know there are a lot of smart and dedicated folks working in this space. A lot. And they spend a lot of time and effort trying to do good things for students, whether it’s creating new products or improving existing ones. And if even the good stuff—if even the best stuff—has only a 10% rate of thoughtful, intentional implementation, then what hope is there?

Plus ça change, plus c'est la même chose

Adoption of any new idea or practice requires a change in habits, and change is something we don’t manage well in schools. To be honest, it’s not something we manage well anywhere.

There are reams of books and articles about the challenges of driving organizational change. Since at least 1965, we’ve had a good and helpful model, developed by Bruce Tuckman. You may have heard the terminology of “norming, storming, forming, and performing,” or seen a graph that looks like this:

What does it tell us? It tells us that whenever a significant change is adopted, initial excitement gives way to the hard part of learning: the frustration and lower performance that come from trying to do something new. As you become more adept and comfortable, you become happier and more successful. Eventually, if the change was a wise one, you are more effective and efficient than you were before you started—and you’re feeling better about yourself and the team, as well. Everybody wins.

What actually happens, though, more often than not, is that when things start to fall apart, people bitch and moan and complain, and leadership usually gives in (or goes on to another job), abandoning the change effort and resetting the graph back to the start. So, what you really get is this cycle: initial excitement; frustration of learning; fear of failure; abandonment; let’s try the next new thing. Round and round it goes, with millions of dollars being spent with every reset, and no one making any real improvement.

Should we blame the people bitching and moaning and complaining? We should not. As Chip and Dan Heath put it, in their book, Switch, what looks like resistance may simply be exhaustion, People have hard jobs, and they didn’t ask to have their cheese moved, as another book puts it. They’re just trying to get through the day.

Change requires a lot of effort and a lot of time. Why engage with it if you know, deep down, that it’s going to be abandoned, just like the last Big Idea was? Change demands vulnerability: nobody likes feeling stupid and clumsy, and you always feel stupid and clumsy when trying something new. If you know that, “this, too, shall pass,” why even try? Why not just keep your head down and wait for the wave to pass you by?

No, don’t blame the troops: blame the leader. Who is making the case for this change? Who is convincing the troops that the old way isn’t good enough anymore? Who is convincing them that the gains will be worth the pains? Who is shepherding the flock to these new and better grazing fields?

If leaders really want to see change, they need to align their people’s intentions with their own. You can’t actually mandate a change; you have to manage it. You have to make the case for it. You have to dig in and show them that it’s important—that you won’t abandon it as soon as things get difficult. You have to be willing to burn the boats and say, “we’re not going back.” You have to be willing to take it to the last mile.

Help is on the way

For those of us who create educational products with a sense of mission and purpose—who don’t work in classrooms but who want to help effect change in the world—the challenge we face—the great task—is keeping purpose and intentionality alive from start to finish: from designers to developers, from developers to marketing and sales teams, from the sales rep to the superintendent, and then to keep going—keep going—to find ways to help our superintendents carry the message, complete and with clarity, to their administrative team, and then to the classroom teacher, and then to each, individual student, so that the fire that lights us, that we use to light our products or programs, carries all the way to the hearts and minds of the students we’re trying to help, like the chain of torches that sends messages from mountaintop to mountaintop in Middle Earth.

It’s hard. You can tell how hard it is by how rarely it works. But in the end, it doesn't really matter what should be done or even what can be done. All that matters is what will be done.