How and when to scale systems within your MAT

By | June 17, 2018

This post is a re-blog of a report of a presentation I gave at an event organised by Arbor Education. I thought its content might be interesting for people who hadn’t come across it elsewhere. I also thought that having done the hard yards of preparing and then giving a presentation to a group of MAT leaders (tough crowd!), I might as well get a free blog post out of it. You can see the original on Arbor’s site here.


In our last blog “The common barriers to scaling a MAT”, we looked at some of the key areas of debate surrounding the biggest challenge to scaling faced by MATs today (this was also the theme of our first MAT conference in London last week).

One of the biggest challenges MATs face is getting the right infrastructure and systems in place to support growth. We invited Dominic Norrish, Group Director of Technology at United Learning, to speak about his experience of deciding how and when to scale systems within a MAT. We’ve summarised his presentation below.

How to decide when to scale systems within your MAT
Dominic talked about how the degree of MAT centralisation vs. school autonomy at your trust is one of the biggest challenges to deciding how to scale systems. Exactly where your MAT sits on the scale of full autonomy vs. full centralisation, or at least where people perceive your MAT to be, is the product of your values and culture. The problem this sliding scale creates is to do with the locus of control (the perception of where authority to make a decision sits) between schools and the MAT central team. The locus of control will always be in tension, since both sides have limited views of the others’ reasoning and drivers. This is often what makes it hard to decide how to scale.



To help overcome this, Dominic suggested applying the principle of subsidiarity to all decisions about whether or not to centralise a system. The principle of subsidarity dictates that if a decision can be taken at a local level, it should be. To determine whether this is true, Dominic suggested asking 3 questions when considering whether a system should be scaled across your MAT:

  1. Where is the activity/decision most effectively carried out?
  2. What is the educational benefit of a single approach?
  3.  Are there strong operational or financial benefits?

If the decision is not most effectively carried out by a central team — or, if there is no benefit to all schools in your MAT adopting the same approach, and if there are no operational benefits, the decision can be local. By contrast, United Learning decided to roll out a single assessment system (Hegarty Maths) across all its schools in 2016 because there was an educational benefit to using the same approach across all schools. Similarly, Dominic said that this principle would suggest that core operational systems, such as Finance, MIS & HR, should be scaled centrally.

When to scale systems within your MAT



When you decide to scale systems comes down to the size and age of your MAT. As the number of schools in your MAT increases, it goes through 4 stages: The Honeymoon Period, The Rubicon of Regret, The Difficult Middle Years, and Converging Needs & Attitudes.

The Honeymoon Period
This is the stage where your MAT has c.2-15+ schools. At this point, appetite for centralisation is high, and the cost/complexity of centralising is relatively low. This is what makes “The Honeymoon Period” a good time to centralise:

Your Identity Management solution (e.g. Microsoft 365, Google)
Your Management Information System 
Standards for the classification, collection & storage of pupil data
Affordable yet scalable Finance & HR solutions 

These are the fundamental systems that should be in place for any young MAT as it grows since the cost of changing them at any point in a MAT’s life is disproportionately high (which is why MATs often put off these changes until it’s too late). These requirements are made clear to schools considering joining the MAT.

The Rubicon of Regret
This is the stage where your MAT has 20-30 schools. “The Rubicon of Regret” is the point at which your MAT has not centralised many (or any!) systems, and now regrets that decision since the cost & complexity of centralisation at this stage is high (but not impossible). This in turn makes the appetite for centralisation low.

Before “crossing the Rubicon”, MATs should centralise:
Whole-Trust educational systems (e.g. assessment), because people tend to be really invested in current choices
Common solutions for IT hygiene factors, such as firewalls and Anti-Virus software
Procurement basics – your MAT should have a leasing partner and preferred suppliers
Back office processes – GDPR, SCR, school information

At any time in your MAT’s growth
From 30 schools upwards, the cost and complexity of centralising systems only continues to rise, but so too does the appetite for centralisation as MAT central teams see the value of doing so after the “Difficult Middle Years.”

However, at any time in the life of your MAT there should be a really high barrier for prescribing the systems teachers use to teach. United Learning, for instance, have stopped doing this altogether. This is because the likelihood of consensus forming around a single product/approach is extremely low, whilst the cost of changing current products and practice rarely delivers ROI (rolling out the same smart whiteboards is one example – does it matter whether all your schools use the same one?). In this case, it would be far better to support schools in driving their own digital strategies.

Click here to see Dominic’s presentation in full

We’ll be adding more posts from our conference on scaling culture, strategy, processes, procurement, and governance soon – keep an eye out for them!

Leave a Reply

Your email address will not be published.