How schools are using pupil premium to close the attainment gap in 2026

Hello everybody, thank you so much for choosing us today. We've only got 15 minutes, we're going to make every single second count. My name is Paul, I'm a former maths teacher and I lead on secondary at Birthspace Learning. I'm delighted to be joined by Chrissy, who is Head Maths and Associate Assistant Principal at Bourneville, and Natalie, who's the Academy Improvement Officer for Maths. Advisor for Maths and Numeracy at Fairfax Maths, of which Bournville is a part. So we have a problem, and it's the problem that brings us all together today, and that is that the attainment gap between disadvantaged and non-disadvantaged is not closing fast enough. We know that we may want to put intervention in place to support our students, and we know the best way to do that is personalized intervention. But personalised intervention means individual. It doesn't mean necessarily small groups. It means personalised to that particular student, where they're identifying their misconceptions and getting the support that they need. That's the gold standard. But gold standard things are very expensive, and this is where the intervention tends to fall down. Sound familiar? So, Kristi. Could you talk us through the challenge that you have as a school?

Yeah, so as you can see, we've got 65% disadvantage, and we are a relatively small school. But as you guys will know, the difficulty is always what kids are on target for intervention.

There are simply too many.

There's one of me, you know. Do I have the staff that are skilled enough to deliver intervention? And I'm always, you know, putting a small drop in the ocean really in terms of the impact that those interventions are actually having. And so that's kind of the main problem. We've also got high EAL, um, the starting points of our students are also significantly below national average, um, when they come in at Key Stage 2. So our average APS is 98 to 99 depending on the year group. Um, so yeah, these— then on top of that is what interventions are we going to put in place, but then what is the drawback of the workload issue that that's going to have on myself as head of maths or on the other staff that I may or may not have available.

Absolutely. So we call this the rationing trap. So all those issues result that we often end up with focusing on a very small cohort, so maybe 8 to 10 fixed places a term because it's expensive, and those students are often exam students, maybe Year 11. But this means that we end up with everybody else getting left. But the disadvantaged cohort is growing, so what do we do about that? The model is often a fixed session model, so more pupils means more cost. You've got your 10 students, if you want 11, it's going to cost you more money. If you want to swap somebody else in, it's going to cost you more money. So the budget ends up deciding who gets support, not the need. Now, wouldn't it be lovely if we didn't have these constraints? That's where Sky comes in. So our company has been— we have loads of former teachers working for us combined with the expertise of engineers. We've created Sky. So Sky is an AI voice maths tutor. It's designed for one-to-one support. It's not a homework help, it's not a worksheet that marks itself, it's in-school scheduled sessions with the support of a teacher. We've got spoken dialogue, high frequency feedback, structured lesson design, strong curriculum alignment, and it's very careful scaffolding. So it's not just about bringing down those cohorts to small groups. What we want here is personalised intervention for every single person in the room. What does this look like in practice? So, Natalie.

So when we first started looking at is we've got 4 schools in our trust, 3 of them have got 65%+ PP, so have a significant amount of PP. So when we first looked around, we were looking at like one-to-one tutoring and it was just cost restrictive, we just couldn't put it in place in the way we wanted to. So when we first saw the AR Tutoring preferred space, it was a little bit of an exciting time because for the first time we thought actually we can unmask support lots and lots of students where it didn't matter if you wanted to swap students around because it was one cost that we could just have lots and lots of students with different year groups if we needed. But the exciting thing as well meant we weren't increasing teacher workload massively. And if anyone ever used ThirdSpace website ever to analyze maths information, they give a wealth of data. So I knew that actually I had a lot of maths teachers who were designing and planning this and we could pick out the misconceptions. So we chose this and we've got it across all 4 of our schools. Bournville have probably implemented it the best and having the most impact at the moment. It was a really interesting time where no longer were we going to go, well, only those 10 students can have access to it. So many of the students that we have now got on our intervention plan probably wouldn't have received intervention at all. So it's the first time we've been able to en masse, give lots and lots of intervention to students.

That's one of the things we're so excited about. It's not just about those 8 to 10 students. You can have as many as you want, whenever you want. It's unlimited tuition. Sort them out if you want to prepare students for an exam or an assessment. You can bring in different cohorts, different year groups, bring in a Year 7 cohort all for one price. So you add more students, it doesn't cost any more. So how does this work? So what we want is personalized support. So let's take two different pathways. Pupil A, very secure. Let's take Pupil A. So Pupil A starts off with a diagnostic question and they get it right. That's great. So what we don't want to do now is spend lots of time modeling and going through things that they already know. So they're going to skip and go down one particular pathway that takes them straight to independent practice. Now Skye's on hand throughout this process. She's there to support. You can ask her questions. She will throw, and if they get it wrong, she will model. So we're identifying misconceptions here, even if they've got the diagnostic question correct. When they're feeling confident, we're going to move on, challenge question, and then we'll have a check-out question. So we get this bookending of data. Check-in data, check-out data. However, people B Same classroom, so People B. People B maybe might struggle a bit. They're not very secure, maybe they're not as confident, so they're going to get that diagnostic question wrong. That's fine. That's what Skye is saying. These are the students that we're really trying to reach. So what we're going to do is go down a different scaffolded pathway. Skye's going to model the question. She's going to do guided practice using a follow me, your turn approach. She's going to ask questions, she's going to support them, she's going to build that confidence. And only when they're ready, they're going to move on to independent work. And then a check-out question, so we've got data so that we can use, the school can use, to assess what happens next. So a question might be, what does the scaffolding look like? So as I say, it's been designed with teachers who've got decades' worth of classroom experience poured into teaching Sky how to deliver these lessons. So we start with an acknowledgement. So get the question wrong, that's absolutely fine. Sky does not say, "That's wrong," or, "Try again." That's not— no good. What Sky's going to do is acknowledge it, identify what they've done well, and then build on it. So it might be, "Oh, you have correctly identified that the tens column in the place value chart, lovely, but we need to go a bit further. We need to look at writing your answer as a decimal. So we're going to take that knowledge then, and then Skye's going to identify the misconception. With that misconception then, she's going to slowly guide, build, encourage until the student's feeling more confident. What did this look like then? When you first heard about this, what did you think, and how's it been working?

So me and Abi had a meeting one day, and she went, You bought this package. I was like, okay, sounds great. I said, amazing, I don't need to buy it, great. So you bought this package and you're going to get 15 kids in a room and they're all going to do individualized lessons. And I was like, what do you mean? It felt like Christmas Day. And I thought, okay, we'll see how it goes. So we had to set it up. So I said, focus on Year 11, that's our starting point, because let's be honest, that's where we are this time of year. We started in October, so we've only been on it about 10, 10 10-ish weeks for now. Um, and I went, let's just get 15 Year 11 kids in. But it meant that me myself— so I just run it for now, just I took control, just took ownership of that. Okay, so it's just me. Well, I can get 15 kids. And I went, right, well, I'll have those, those 5, they're struggling to get a 4 at the minute, they just are. I've got 5 kids over there who just need a little push to get a 5, and I've got 5 kids that, that need a push to get a 7. And I'm like, amazing, I can, I can target all 3 of them straight away. And they, they come in every day of the week, only 4 days a week because I can't take them out of assembly, unfortunately. I tried. But they come in every morning, we have form time in the morning which really helps, half an hour, they log on, they put on their little gaming headsets which is still a novelty to them, um, and they just crack on with their session. And straight away, I mean, I was just in a session about live marking and I'm imagining in my classroom I'm teaching a topic And first of all, even on that topic, some of them might already know what to do. So that diagnostic question, well, they're jumping straight to the independent practice straight away. They're not waiting for me as a teacher. But I've seen that live marking, and where, you know, I'll go for— uh, it's Emma in the corner— where Emma might have got that question wrong, but I'm stuck over here with these 4 or 5 or 6 pupils, and it's been 20 minutes until I can go and tell her that she's wrong and tell her where, where that misconception is. So what's been really nice is that in only half an hour a day, the students are opening up a new skill, figuring out if they can do that skill or not, and if they can't, they've got teaching input straight away to help them get there. And then at the end of the session, they're checking out and they're going, 'Couldn't do it at the start, can I do it now? Yes or no?' And then they come in the next day, and if they still didn't get the skill from the previous day, they carry on that skill. So they're just carrying on what they're doing, all that's ticked off, and then move on to the next one. And what I did to set it up— I mean, the setup was really nice and quick and easy. I just emailed Thirdspace, stick all of my Year 11s on please. I then clicked the kids that I wanted, and I chose a target grade for them. So knowing your kids and where they're currently at— so I just use mock data— that every kid is either a target of 3 or 4 or 5 or 7 or 6, are they? 3, 4, 5, 6, or 7. So each— so if you're on a 4, sorry, you're on a 3, you're on a 5, and you're on a 6, they're all coming in the room in the morning and they're all doing completely different things depending on what's been set up. Um, and for me that's just been like a game changer. Like, the alternative of that is me running interventions every morning trying to think, okay, these 15 kids What do they know? What don't they know? What am I going to pick to teach today? And I like to think I'm a pretty good teacher, I think, but I can't get anywhere near the impact myself as a lone person that this can. And that's the thing for me that I'm finding really, really exciting. So some data. So this is kind of the, the exciting point. At the end of each session, the kids rate how they found the session. And we've got an average satisfaction of 3.8, which I think is pretty good. I think if the kids rated me on a daily basis, I might not get 3.8, I might get 2 or 3, some days a 5. Um, but then this, for me as a mathematician, is my biggest one. So 80% of kids are able to correctly— well, I'll flip that over. They come in on Monday morning, they get a question, they get it wrong. So they can't do that skill. Half an hour later, 80% of our kids can then do that skill. And for me, that's kind of the biggest indicator that this is, is going to have real, real impact. Okay, we're only 10 weeks in, so we haven't got any actual exam dates or anything like that. We are in the middle of mocks, which I'm excitedly marking, and I'm really intrigued to see, well, what impact has that had? They did November mocks kind of roughly when this started. I think they've done 4 weeks now. And they did the November mocks, and then they've been in the program now, and now they've done the next set of mocks. So if anyone did want to contact me kind of in a couple of weeks, I will— or I can let you know what improvements have been made. And I say that positively because I am feeling positive. I, I'm confident there's going to be improvements. Is it going to work for every kid? No, it's not. But it's certainly having way, way more impact than myself or any of my maths teachers could do between us at the minute. Um, and then the last one is just whether they're improving their confidence. And for those, for those of us that know with maths teaching, if the kid is confident in maths, they will naturally make progress. So even if it's not just about learning that specific skill every day, they're coming in the morning, they're doing a bit of maths, they're learning something new, and then they're going on about their day with that improved confidence. Or they're coming to me, period 3, and going, 'I did this this morning, I can do it now because I've had my intervention, I know what to do.' So we're seeing some really positive indicative data that kind of makes us quite excited for the future.

We are super excited about this and following what's going on with very, very close interest. And before opening it up to any questions, we'd love to hear the questions. There's two things I just wanted to mention which we get asked about quite a lot. So one of them is safeguarding. So this is an AI, and safeguarding is our number one priority for teachers. So all sessions are recorded. We have a red flag monitoring system in place. It's super, super safe, but it's better to hear from people using it. How do you find the safeguarding?

Yeah, so when we first start looking into the system, we all— we had to fill in a data impact assessment as you do in GDPR. And one of the things we're really proud to look at is safeguarding. So one of the first things that actually first spice then did when they spoke to us about the system was talk about the safeguarding systems in place with our local authorities. They spoke about the fact because students are responding and the AI responds to them, um, that actually if any of the students are speaking about anything that they shouldn't be, that there is a flag system that then comes to us so then we can act on it appropriately. Now that's in place regardless of whether they're doing the sessions inside school because this can be used outside of school by students in their own home as well, and that safeguarding system isn't just when they're on school site but whenever they use the system off-site as well, so we can address any concerns there. And the other big one, as we know, is things around open AI systems and the issues that we might have around those, but Firstbase Learning is a closed AI system, so So they've been helping to find students' voices, data, information, floating around on the internet, which is a really, really important element for us.

And yeah, I mean, we found that. And one of the funniest things was someone had to try it, obviously. And I— we did like one-hour session where we explained how this was going to work, and I said, listen, it's closed AI, you can't talk to it like you talk to ChatGPT. And they went, watch me. 'What's your favorite film?' This is them in their headsets, by the way. And then they're like, 'Miss, Sky's not responding to me.' And I'm like, 'Stop asking any questions, it's not going to answer.' All it does is just loops back around. So it just assured me really that they were going to stay on task, and that meant I could then assure the parents, because naturally there was some kind of, 'Hang on a second, what do you mean that they're talking to some AI person at home or in school?' And I said, 'No, no, explain how it works.' And that kind of gave them the reassurance that The scary bit about AI isn't there in this system. It just doesn't exist.

And the second thing I wanted to raise as well is, you know, it's wonderful to have it in a classroom. You know, if we're in this classroom now, everyone would have a laptop and a headset. You'd have a variety of different lessons going on. But how do we scale this? You know, there's so many disadvantaged students. How do we get this across of trust? What are the challenges associated with that? What goes well? What hasn't gone Could you speak on that?

Yes, so we have got it running in all 4 schools. 2 schools have been more successful with the implementation, so Bourneville's probably our most successful, and then Smithswood Academy over in Solihull, if anyone knows it, is also the one that kind of opened, ran with it fairly quickly. One of the biggest things was having staff buy-in from the 2 heads of department in those 2 schools that I massively brought in first. And probably have some of the more disadvantaged students and really could see the benefit of being able to get students to get on a system. They probably also have more concerns around high quality access to teaching, so they could see, really see the benefit of having something in place there. And they're also in that lucky position where Balfour Head Department do not have tutor groups, so they could straight off go, well, we'll run it, we'll get it in place, and we have both those heads of departments running it every single morning. In fact, Smiths would also do it with Year 10, so they have a morning where they do Year 10 as well, so they really scaled it out more. And they also were both in a lucky position where there was a computer room available in the mornings that didn't have to be used by a tutor group, because they've got spare classrooms in their school, so they could straight straight off get that done. One of our other schools has had more of an issue because it's such a large oversubscribed school is I think we all automatically thought, well, they'll have some computer rooms they can use, but they're all used by student groups and trying to swap those out was a bit of a logistical issue. They also didn't have available a bank of laptops because they'd already been nabbed by Literacy so they've already been used by Reading Wise literacy package. If we'd perhaps had this in September, we probably could have got hold of those a bit sooner and booked them out. So when, if you do want to implement it, you have to think about what can we got available. Have we got a member of staff? It does not have to be a maths teacher. It's helpful but doesn't have to be. And going forward, maybe using some of our TAs in the morning again they were all allocated out to literacy, but if we can get there early, we should be able to get one, and making sure there is access to the technology. So to overcome it, the one school that has not got the technology available, they actually have less disadvantage, so they're going to roll out the package— have started rolling out the package outside of school instead. So they'll have more buy-in from parents in the first place, and they're going to roll this out outside school, so So even if there's some barriers inside school, it is still available to use outside of school. So we've just started to see that being implemented then.

Yeah, and that's the thing with us and scaling up. It was very easy with me to kind of get it rolling, but now I'm already thinking, well, what about September? And really, we can double, triple, quadruple the impact we can have because if I can just get a couple of TAs in the morning, I need one bank of laptops, 15 in each room, all right? And you just be creative with your spaces. Your IT facilities and have a quiet word in people's ears. And you want to think about what— getting that infrastructure in place first, nabbing some support staff that might be available during those times. And that was the thing with form time, is it's a bit easier to get that. Um, and then I'm looking to have at least 3 kind of lots of sessions running. I have Year 11s, I have Year 10, I have Year 7. Uh, that's exciting for me today because I haven't got any Year 7s on there yet. So I'm thinking, right, well, how can I start making an impact in Year 7 using this program that we've already paid for? It's just sat there on shelf. How can we have impact there? Which means I'm not fighting fires in Year 10 and Year 11 anymore. Um, I'll come back to you on the Year 7 point.

Smithswood have just recently gone through Ofsted. Um, I don't know if any have gone through Ofsted yet, but I do ask about closing gaps of foundational knowledge. So I'm thinking ahead of Year 7s perhaps using and talking about this, being able to use this to close foundational knowledge gaps in our Year 7s. I think it would be a really interesting move forward.

Thank you so much, everybody. We would absolutely love to hear any questions.