The SMART Tool: $24 Million Later and Still No Wisdom

by Rebecca Thomas

Always winter. Never wisdom. The thaw has begun.

“Put an end to the new SMART Tool,” says Ginny Andersen.

And I will admit it. That sentence helped me shed a layer of despair today.

Just in case Joe Public isn’t sure teachers assess children every hour of every day I would like to paint you a picture. Teachers notice the child who can decode but not comprehend. The child who understands maths with equipment and tools and fingers, but freezes when it sits on a screen. The child who can explain their thinking beautifully beside you, then clicks random buttons the moment a platform starts watching them.

Assessment matters to us in every moment, that’s how we respond. But this? SMART? This has not felt like ‘assessment’. It has felt like another expensive national experiment dropped into schools with a press release, a delayed deadline, and a prayer to win votes, and an excuse to send money overseas, buy products and ‘keep the economy’ back on track.

Before the starting gun even fired, teachers/leaders were asked in the last two days of Term 1 to upload their students to the system. Two days’ notice. Or else.

I failed that first test. Not because I am incapable, but because I was exhausted and foolishly thought, “I have a holiday coming. I’ll do it then.” Five days later, I opened the system and found it shut for the whole holiday. I had to wait until the first day back.

On that day, apparently, SMART was ready.

I was frustrated, but still hopeful. Perhaps this shiny new tool would tell us something we did not already know about our tamariki. The leadership and learning focus reports sounded like a great opportunity for insights especially with ERO, mid year reports and board reports to be made. Perhaps, after all the money, all the messaging, all the “clear and consistent information” promised to parents and whānau, we would receive something useful. It might help us grasp how far our kids were away from the new curriculum and what we needed to do to plug some gaps.

Then came the PLD.

Except the PLD was not really about using SMART. Not how it worked. Not how to operate it. Not how to troubleshoot it. Not how to read the data. Not how to support our learners through it. It was more like old-school assessment for learning PLD with the new mandated reporting language sprinkled on top like policy confetti.

And we had to do the PLD in Term 2. Or else.

So we moved around other PLD. We shifted the things we were already doing. We made room, because schools always make room for the next thing being done to us.

Then the assessment window opened. Our kids had a go. Our teachers were anxious. I had stopped counting the posts in teacher Facebook groups about glitches, set-up issues, confusion, log-in problems and the hundred small fires that happen when a national tool meets the reality of real classrooms.

None of this should have been surprising. Back in April, when concerns about SMART were being dismissed as a “political hit job”, educators were already raising questions about confidence, readiness and implementation. The profession wasn't trying to score points. It was trying to wave a warning flag.

In the trial, the maths assessment had a calculator, protractor and ruler. We showed our students those features to reassure them beforehand, and we tried to set them up for success.

Then the real test arrived. No calculator. No protractor. No ruler. “Whaea, where is the calculator?” Gone, apparently. Vanished into the same mysterious place as teacher trust, implementation planning, and common sense. Perhaps it slipped through the back of a wardrobe.

The children treated the platform like a gaming habit. When they got stuck, they spammed the keyboard. Some spammed their way to the end and hit submit in five minutes. Nobody flagged a question. Nobody went back to check their answers. Nobody suddenly became a reflective learner because a platform had a button for it.

The writing assessment was even more depressing. It did not give many children much opportunity to show actual writing. For too long, they were selecting multiple choice answers about which sentence “looked” correct. Then, when finally allowed to write, some jotted a few words in the box and clicked through. The AI marker responded with the digital equivalent of a shrug. Not enough writing to mark. No detail to share. Well, that was illuminating.

Meanwhile, teachers were trying to deliver the tool properly. They were doing what they had been asked to do. They were following instructions. But screens froze. Children who did not click quickly enough were locked out. The stress in the room rose. Dysregulated teachers meant dysregulated children, because classrooms are human ecosystems, not assessment laboratories.

And then we waited.

We waited for the high-level detailed reports. The insights. The revelations. The promised clarity.

Four weeks after the testing window now closed, and two weeks before school reports were due to go home, already written, there was still no meaningful high-level report for our learners. Nothing useful for leadership to analyse. Nothing timely enough to support teaching. Nothing that justified the disruption.

Helpful? Not really.

Looking back, that might be the most frustrating part. Many of the concerns raised before a single child sat the assessment were not predictions at all. They were professional observations from people inside the system. The first mandated assessment window has now been and gone, and much of what was dismissed as resistance has simply become lived experience.

New Zealand has committed around $24 million over five years to Janison for this tool. That is not a small amount of money. That is teacher aides. That is learning support. That is release time. That is trained humans beside children who need someone to notice the story behind the score.

Instead, we got a platform that, in its first mandated window, gave us stress, delays, confusion, missing tools, thin writing data, vague AI marking, and reports that did not arrive in time to help the very reporting cycle it was supposed to strengthen.

Treasury warned there were implementation risks. Teachers could have told them that for free. Because the problem was never that teachers were against assessment. The problem is that teachers are tired of being handed expensive solutions that do not solve the thing they claim to solve.

SMART did tell me some children scored low. The problem is that I already knew that. I knew it before the assessment window opened. Their teachers knew it. Their whānau knew it. In many cases, the children knew it too. What I needed was something that helped explain why they struggle with learning.

Even obvious human ‘whys’ that don’t need expensive AI to do so:

Why the child who missed half the term because of chronic absence scored differently from the child who attended every day.

Why the child still learning English interpreted questions differently.

Why the child with dyslexia could explain their thinking beautifully but struggled to demonstrate it on a screen.

Why the child who freezes in formal testing situations produced a result that looked nothing like the learning I see every day.

And before anyone tells me no assessment tool can explain those things, I agree. That is precisely the point. Those answers come from teachers, whānau, attendance data, observations, learning support specialists, relationships and professional judgement. They do not come from a screen. Yet SMART was sold as a way of providing deeper insights into learning. What many schools have received so far feels more like confirmation than revelation.

SMART measured performance on a particular day, in a particular format, under particular conditions. What it did not do was help us understand the causes behind the result.

Schools do not improve achievement by knowing a child scored low.

We improve achievement by understanding what sits underneath that score and responding to it.

The promise of SMART was not simply that it would generate numbers. Schools have been generating numbers for decades. The promise was better information, deeper insights, and data that would help teachers make better decisions. Yet months into the rollout, many teachers are still waiting for reports detailed enough to provide those insights. We are in June 2026, the whole ‘deal’ was done under a start for PLD in it last October 2025.

The result is that we disrupted classrooms, invested countless hours, shifted professional learning, navigated technical issues and compliance deadlines, and largely ended up confirming things we already knew.

Signing up was never the same thing as believing. Now we finally know that using it wasn't the same thing as valuing it either.

That is the difference between data and understanding. Data tells me a student is behind. Understanding helps me know what to do next. And if an assessment system cannot reliably move us from one to the other, then what exactly did we spend all that money on?

If the Government truly wants "value for money", as it keeps saying, it might start by asking whether $24 million has bought anything close to wisdom.

Because from where I sit, SMART has not made assessment smarter. It has made assessment more expensive, more complicated, and more removed from the people who know children best.

After eight months of preparation, a mandated assessment window, disrupted professional learning, technical issues, delayed reporting and countless hours of teacher time, I am still struggling to identify what problem it solved.

Which leaves me with one question:

Progress or procurement?