The analysis is anonymised, and schools or parents will not be told which children are "at risk". Schools will decide how to spend the money allocated.
The decile system is based on the area from which a school draws its students. But attaching a number to each school has meant some parents avoid lower-decile schools, mistakenly linking a decile number with the quality of education.
That has led to school rolls falling at some schools, particularly those that border schools with higher deciles.
"This is the shift that we need to make as a nation, to get parents thinking about what is the achievement information of a school, what is happening at the leadership level...not what the income is of particular parents," Kaye said.
"We want to change the conversation as a country to be not about socioeconomic income of neighbourhoods, to be about the teaching and learning that is going on at individual schools."
Replacing the decile system is just one part of the government's wide-ranging education reforms.
The Government is already testing a version of "at-risk" school funding. Schools this year did not get a general increase to operations funding to cover inflation. Instead, schools get $92 in extra funding for each student from a long-term, welfare-dependent background.
There is widespread support for scrapping deciles. Principals in middle decile schools complain they do not get enough money, despite having students from poor backgrounds along with those from better-off families.
However, Labour's education spokesman Chris Hipkins said despite assurances the "at risk" model could still lead to schools being stigmatised.
He understood the desire to improve on the current decile system but said the main problem was the government was trying to re-carve an insufficient funding pie. Labour would significantly increase the funding to schools and ECE centres.
Green Party education spokeswoman Catherine Delahunty said "hyper targeting" brought with it no guarantee children who were actually at risk would be helped, and risked stigmatising children.
She was uncomfortable with every child in the country being assessed according to their ethnicity and personal family information, even if the analysis was anonymised.
What was needed was a big boost in funding for schools, Delahunty said, and scrapping deciles wouldn't end the stigma unfairly attached to many schools.
New Zealand First education spokeswoman Tracey Martin said assurances were needed that the advisors behind the change were not a "handpicked crop of National-leaning educators".
"We know former Education Minister Hekia Parata trialled the model last year, and many children missed out on getting more support."
The primary and intermediate union NZEI said any decile replacement would be "merely shifting the deckchairs on the Titanic", unless in was accompanied by an immediate and large increase in funding.
What you need to know
What is the school decile system?
Decile ratings were introduced in 1995 to promote equality, with low-decile schools receiving more funding. Decile 1 schools are the 10 per cent of schools with the highest proportion of students from low socio-economic communities. Decile 10 schools are the 10 per cent with the lowest proportion of such students.
Deciles are worked out by combining students' address information with Census data for small geographical areas to examine factors including household income and crowding, and parents' occupations, qualifications, and benefit support.
Why is the Government replacing it?
Education Minister Nikki Kaye says the decile system is too blunt, and advances in data and analytics mean it is possible to use information about each child or pre-schooler, rather than the neighbourhoods in which they live.
Another key argument is too many parents wrongly equate a higher decile with better quality education, leading to lower decile schools being stigmatised.
What will take its place and when?
As early as 2019 a new system will be used that allocates extra money according to how many "at risk" children schools have enrolled.
A statistical "risk" index will be used to estimate for each student how at risk of educational underachievement they are. What data it uses is yet to be confirmed, but will likely include around 16 factors including ethnicity and a mother's average income. All data will be anonymised, and schools and parents won't be told which children are judged "at risk".
