When I learned that SESIS, the special education data system, was being discontinued by New York City’s education department, I felt conflicted.

I’d spent thousands of hours inputting information into that system. On the one hand, SESIS has major technical flaws, requiring educators to either check or uncheck a Kafkaesque litany of boxes to enter information accurately. But I worry that SESIS is just the easiest thing to blame for the real problem: inadequate support being provided to students who need it the most.

We can replace SESIS with something new, but far more important issues will remain.

For one, creating Individualized Education Programs for students is demanding no matter what system is used. Creating high-quality, useful IEPs is an even more intensive process. But teachers often don’t get the training or time they need to do it well. Teachers are too frequently working in isolation to develop IEPs, rather than with a team — an inefficient system that results in IEPs that are of poorer quality.

When I first began creating IEPs as a newbie teacher, I struggled to craft a narrative from a few points of general data. I still have my records from back then, and I’ve taken a short snippet from an IEP for Jerry (not the student’s real name) as an example:

Jerry writes independently in complete sentences and with excellent handwriting, and he will write as a means of self-expression without any prompting by the teacher. However, Jerry requires assistance with his spelling, grammar, and the organization of his thoughts.

I’m frankly embarrassed by how meager this was. Many students struggle with spelling, grammar, and organization! What was it about what Jerry was struggling with that required an individualized program? This matters, because teachers, service providers, and parents need to know exactly what challenges a student like Jerry is facing in order and what tools and accommodations can help.

The problem was not simply that I didn’t have sufficient assessment data. It was that I was writing my IEPs mostly on my own, struggling to draw the connections between my daily interactions with the student and whatever information I had.

I tried to address this isolation by working with my principal to form a special education team that met regularly. This helped, but the more I learned, the more I realized still more needed to happen. At my next school, where I was the special education coordinator, I had the great fortune of having a staff willing to use a grade-level meeting every week to focus on the needs of individual students with disabilities.

Essentially, instead of working by myself to write IEPs, we worked as an entire team to craft the documents. By having a rich discussion, grounded in student data and work, and then pairing that discussion with follow-ups with the students and their parents, our IEPs became far more specific and, ultimately, more useful to educators and students alike.

Here’s a short snippet about a student’s writing ability from one of my last IEPs (again, not using a student’s real name):

When Ken shares his thinking either verbally or in writing, he may make connections that are initially inaccurate or that may refer associatively to what he is doing in the process of writing or thinking in that moment. He frequently gets “stuck” when trying to verbally articulate his thinking, and requires frequent prompting and sentence starters to begin or to complete his thought. Ken is observed to perform better when 1-on-1 support is provided or explicit, literal, visual support is provided. Graphic organizers and other visual scaffolds are important tools for Ken to gain access to class tasks and topics.

Ken shared that opportunities for peer review and feedback of his writing are beneficial to him: “When we did an ELA essay, when they peer reviewed it … once I read what the peer had said, I tried to add a thesis statement, and as many body paragraphs, when I was reading my peer’s handwriting, I have to think about what I can do better next time.”

See the difference? It’s not because I developed greater insight or analysis into student needs myself. It’s because I was working with a team, which included conversations directly with the student.

Of course, none of this has anything to do with SESIS. This process occurred outside of SESIS, with educators typing in a Google Doc. I would then just copy and past that information into SESIS and fine-tune it to make it into a cohesive narrative.

But as intensive as all of that was, it still wasn’t enough. You can end up with the greatest IEP in the world, but the bigger question is: Did that student’s outcomes improve? That’s a much harder thing to accomplish.

So sure, bring in a new system. Throw away SESIS and all of its arcane checkboxes and fax cover sheets. I welcome any improvement with open arms — especially one that allows teachers to spend more time actually improving IEPs or helping students directly rather than wrangling with an online data system.

A new system could even be designed to better allow for dynamic, ongoing team conversations and collaboration. Getting there will require a shift in mentality and in-school practices, though, not just a shift in software. The reason SESIS was so cumbersome, bogging down teachers with boxes to check, is because it was a system designed to ensure explicit adherence to every stipulation of the law. Far more important is whether every adult who interacts with a student with an IEP knows that student’s needs and is working daily to address them.

If we can build a special education data system that allows the student, the student’s family, and all the adults that work with the student to collaborate, then that would be a system worth investing in. But let’s also not fool ourselves into thinking any system will solve the human problems that lie behind poor student outcomes.

Mark Anderson is an instructional lead for ELA for the New York City Department of Education’s Bronx Borough Office. Before that, he served as a special education teacher.