The False Economy of a Shoe-String Data Strategy

My dear old mum uses an adage that I have seen proven over and over again: “Buy cheap, pay twice”.

I need to caveat this insight into my mum’s wisdom and be clear from the outset that while this particular adage is a good one, she is no Nostradamus. When it comes to axioms and proverbs, to be fair, she is more of a blunt spoon than a surgical scalpel. Occasionally she’ll hit the mark, but usually she raises more questions than she answers. “More haste, less speed”, for example. What the hell does that mean? And “a watched pot never boils”. Sorry mum – but it does. Providing the pan is heat-conductive and there is a sufficient heat source, it will boil. It’s simple physics.

But the fact remains – “buy cheap, pay twice” is a good one.

Here is an example of it in practice. A friend of mine bought my son a drone for Christmas. I was annoyed about this, thinking they had spent a lot of money on it (and that it trumped the Victorian hoop and stick that I had given him). But it turned out that it was not expensive. In fact it was ridiculously cheap. So cheap, that within 5 minutes of playing with it, it had broken. Money had been spent on this thing, but it was a complete false economy.  

So where does that leave us? Well, if my son is ever to have a drone, we’ll need to start again and disregard the investment that has already been made.  The options will be to buy him another cheap one – which will break just as quickly – or to buy him a more expensive one that will not break (a third option, of course, is to be more forceful in selling him the desirability of the Victorian hoop and stick). The cheap purchase was, in material terms, meaningless. It was, simply, flushing money away.

The subject of false economising has been particularly pertinent to conversations I have been having with a number of Multi Academy Trusts, around how they manage their data strategy. When it comes to matters of budget and balancing the books it can be difficult to see the wood for the trees. This is more so, when money is tight. In the current economic climate, within education, it is tempting to make a saving wherever the opportunity presents itself – while not always recognising how a calculated investment in one area could actually save significant money in other areas.

Currently, data analysis within school groups (in the UK, at least) is often limited in its scope and this makes it difficult to act on the wider picture. Stretched resources – in terms of both budget and personnel – mean that there are compromises at every step of the data journey. Getting hold of the data from across the estate, cleansing the data, consolidating and aggregating it and then reporting on it often amount to more than a full-time job. In this culture, there is hardly the room for meaningful data analysis. Data becomes little more than static content to report on – it becomes the decomposed cadaver in a postmortem.

What should be the most important part of the data journey – the dynamic and meaningful analysis of data, to enable timely intervention and agile decision making – is too often seen as a luxury that we’ll “get around to at some point”. Stretched resources mean that we are always playing catch-up and the rich mine of data available from across the Trust, simply isn’t being used to its full potential.

The interesting thing is that (like any organisation) when the MAT draws breath and steps back – and works with the data available to it – strategic opportunities for improving the health of the budget and making the most of limited resources reveal themselves.

Let’s consider a few very real examples of this.

An increase in Pupil Premium funding. Through well-presented data, a Multi Academy Trust noticed that despite an overall similar student demographic, one school had a significantly higher level of Pupil Premium funding. Further investigation into this showed that the school had excellent parent engagement and supported parents with FSM and related applications. The trust was able to lever the success of this school to replicate systems and processes across the Trust for greater parent engagement and finance support.

Improving Attendance. In the West Midlands, a Trust identified two schools that had significantly better attendance figures than all other schools within the Trust. Conversations with these schools revealed that both had a “walking bus” that picked children up and dropped them off at the end of the day. The “walking bus” was rolled out across the Trust’s schools and tangible improvements were seen.

Supporting RWM more strategically. In all of the noise around End of Key Stage assessments, with so many variables involved, it can be tricky to spot meaningful patterns. Through aggregated data visualisation, one MAT immediately saw two stark patterns. Writing at KS2 was suffering significantly more than any other area; and it was the boys that were having the greatest impact here. Seeing this so plainly, as being globally consistent across the Trust, the MAT was able to lay on a consolidated programme of Improving Boys’ Writing for all schools across the Trust.  Vital funding was directed to exactly where it was needed and economies of scale meant that the money went that much further.

Better learning opportunities for students based on characteristics. In a Trust near Manchester, the Academic Director noticed that Pakistani girls in one school performed significantly better in English than elsewhere across the Trust. A conversation with the school uncovered that cultural sensitivities were at play – with adjustments having been made both in curriculum content and the way classes were grouped – leading to greater confidence and engagement from the students. This approach was shared with English faculties from schools across the group, with a view to improving standards in English teaching and learning across the Trust for different ethnic groups.

There are many other examples of where Trust-wide data has been used so effectively, but in each of these cases, we see the Trust use data to raise pertinent questions and drive improvement while making the most of limited resources. It is also worth noting that in these cases (I have been deliberate in picking them!), the dynamic of the Trust was levered to its full potential – with good practice being identified and then shared with other schools, engendering a fully collaborative Trust-wide culture. Crucially, in each case, the Trust had invested in a dynamic data tool to make all of this possible.

What we see in these Multi Academy Trusts, is the possibility for executive leaders to see real-time data from all schools across the Trust and to act on that. In each case, the data was already there for them, was available for scrutiny and was often already surfaced in intelligent ways that would aid insight and support inquiry. The immediacy of the data was crucial not only for gaining insight, but to enable it to be acted on quickly. The pain of collecting the necessary data – relying on human processes to obtain, cleanse, aggregate and present it – simply was not a factor. Data collection, aggregation and presentation was automated and so the data experience for the executive teams was fully dynamic. From wanting to understand what was happening across the Trust to drawing up hypotheses and conclusions, there was no barrier to interacting with the data.

This can be compared to a whole plethora of different data collection strategies at play in Trusts across the UK. In some cases, Executive Teams have log-ins to many different systems and have to collect and make sense of the data themselves. Not much time for data analysis there. In other cases, stoic Data Managers work hard to service the many different requests from across the Executive Team, pulling together different data sets as required and presenting them as best they can. The work involved in this – not to mention the sheer logistics of the thing – mean that interaction with Trust data is anything but dynamic. In other cases still, a “budget approach” to working with data means that while a MAT-level system is in place, it only does some of what is needed (often focusing on assessment). In such cases, a tool comes to be relied upon and seen as a source of truth - but in fact, gives a very narrow perspective of the Trust's rich data.

It is fair to say that when a Trust cobbles together a budget solution, money is definitely being saved. With no investment in an intelligent data analysis system, no such system will feature as a line on the budget. 

However, there is a trade-off. And this is where my mum would be nodding sagely.

Instead of a powerful tool managing data collection and integration – with all that is needed for different executive stakeholders to dip in and out to inform their understanding of ongoing Trust performance across a range of areas – what these Trusts rely on is vast amounts of man-hours to make the data available and get it in to shape. Not only does it take time and money to do this, but often the people doing this are the people whose expertise would best be spent analysing this data. The sad irony is that once they’ve knocked the data in to shape, cleansed it, consolidated it and presented it…there simply isn’t the time to do any meaningful analysis. Oh – and they’ve probably lost the will to live.

And that’s why there are so many decomposing cadavers out there undergoing postmortems in the boardroom.

So we can safely say that it certainly isn’t necessary for a Multi Academy Trust to invest in a dynamic system that will take away the headaches around Trust data. You don’t need such a tool to collect the data, or to aggregate it or to present it in clean, useful ways.  People can do this. It just takes a while. And the data starts to go out of date pretty quickly. And what ends up getting presented is liable to be fairly uninspiring. But that’s ok. In technical terms, too, no doubt there are many ways to collect and display data – though admittedly, the elegance of this will probably be wanting and any such system is liable to need a lot of maintenance and health-checking.

But there are great systems out of there – for example, Groupcall’s MAT Analytics and the powerful Assembly platform – that do the donkey work and present the data on plate, in truly insightful ways, to those people whose job it is to make decisions based on that data. 

The cost of these systems is nothing, compared to the savings and efficiencies they represent.  Of course, as with all things in life, savings can be made by refusing to invest in such a system. On balance, though, opting out of such an investment demonstrates a false economy as workloads go up, efficiencies go down and the agility to solve immediate problems is lost the four winds.