bringmoredata

Great idea to have a poll of SIMS users vikpaw! Intrigued to see what the results show...

Sorry I was slow to respond. 56 of UL's schools were SIMS as of the 2019 autumn census. The others were with Advanced, RM and Arbor. So yes, the losses for SIMS from this are quite significant.

Thanks! When I started writing about this in 2014 I didn't think many people would care. I'm still pleasantly surprised and grateful when people get something from it...

I don't necessarily disagree Paul. As context, I try to limit the blog to commentary based on the hard data, rather than speculating about the future. I want it to be as objective as possible and speculation can veer into subjective bias too easily. But yes, what you say is absolutely plausible. Nobody is sitting at home, trying to squeeze data out of SIMS through a VPN (or not able to access it at all) thinking "yay, locally hosted!".

Thanks Ditto - I appreciate you sharing this (and I really need to change my edugeek handle to bringmoredata now I've left Assembly!). I'm intrigued to see how SIMS Support Units will react in particular. I'm not yet sure if they'll be worried by the news (uncertainty in SIMS HQ) or see it as a positive development (new management should come with a new strategy). Anyone have any views?

Hi, Good to see this conversation... Just to throw in my two cents: - Power BI is definitely gaining traction as the dataviz tool of choice in education. Kudos to Bromcom (with LOTS of input from deshills and the Harris team) for their pioneering work on this front. Oasis were also early to the Power BI party, so props to them too. And Matt40k has been (rightly) advocating for Power BI for years now... - Groupcall are now offering a MAT analytics product powered by Power BI, which might be of interest to non-Bromcom MATs... and Assembly (my day job) are also experimenting with Power BI for MATs with more custom needs than we can serve with our core MAT analytics tool. - My main anxiety with Power BI (or in fairness any BI tool) in MATs is that I've yet to see an implementation that requires zero customer maintenance - and maintaining dashboards can quickly snowball from a small task (fun project for a Friday night!) to a big task (employing one or more full-time Power BI people). E.G. Assembly Analytics (our main analytics product) contains lots of benchmarks built from public data, and we update them behind the scenes whenever required... but it's much harder to do something similar with Power BI. Not impossible... but harder. - A big and important question is also whether you want to power dashboards from an API (like Bromcom do with OData) or a data warehouse (like Groupcall do). Both have strengths and weaknesses - and we're experimenting with Power BI attached to both at Assembly. Short answer: it's great to do as much as possible via an API, but at some point it may become necessary to have an underlying dataset that does pre-aggregations / historization etc. - I also think it's worth avoiding the temptation to use every visualisation the tool offers - particularly on the same dashboard! You can do a lot with bar charts, which are easy to interpret.... But I'm veering in to subjective stuff so I'll shush now. Hope that's helpful context :-)

Also, there's one other thing I should have mentioned: we tried a range of permission options before we arrived at the current recommended set for SIMS. What we found was that narrower permissions led to some unexpected data blockages within the SIMS API. I.E. certain scopes of data weren't being extracted as expected on some versions of SIMS. So we've settled on the current recommended set because they've worked in all cases so far. I hope that helps to explain our approach!

Hi Sarge (I'm director of Assembly btw), It's a fair question. We should probably document why we ask for this better on our support site. It's important to start by explaining that the data that flows to the Assembly platform from your MIS is controlled by you within the Assembly platform. By connecting your MIS to Assembly platform with a MIS login (the step you're currently undertaking), you don't actually send any sensitive data to us; you just set the parameters by which data could be transferred at a later date. We ask for this range of permissions because it covers the wide range of use cases that we expect to see - and we don't want you to have to keep adjusting or adding SIMS permissions whenever you use a new app. BUT, the key thing is that for data to flow, an app needs to make a data access request within the Assembly platform, which a school then authorises. This support guide shows the Data Access Request screen for the Mathematics Matery app. The Data Access Request clearly shows the scopes of data that the app will pull. No sensitive data can flow until you grant your first Data Access Request, and we can only extract those scopes you've authorised. So you can be confident that no sensitive data can flow to Assembly without you first authorising a data access request, regardless of the SIMS permissions you set. We also give you the ability to see what scopes have been authorised for which apps in our admin UI. Another way of thinking of it is that there are two "taps" between the school and an app: one within the MIS (i.e. via the user's permissions) and one within the Assembly platform (via the Data Access Request). Both need to be open for data to flow. But we think it's much easier to control, limit and explain what data is flowing through the Assembly tap. So we ask for you to open the MIS tap, safe in the knowledge that the Assembly tap gives you everything you need to control data flows in a secure and reliable way. I hope that explains it. Do say if you have any further questions - or feel free to email [email protected].

Thanks @vikpaw - nice to be welcomed! That's an interesting suggestion on future modelling, though I'm afraid I don't know an easy way to do it. Also, to be honest, I think there isn't enough trend data to project forward... though I'm sure @matt40k's Jan census FOI request will help to update the picture. I'll add that data into my model when it comes through...

Great that you two are on the case - I'll keep a keen eye out for your FOI. I've met @GREED and had fun conversations but don't think we've ever come across each other? Give us a shout if you're ever in London and fancy popping by our Holborn offices for a coffee...

Having looked at the data in more detail, by my calculations Capita lost 138 schools... but I also think that only 373 schools switched... so the net loss equates to 37% of the total number of switchers! That makes me think that, as soon as one of the cloud insurgents make data migration so simple that people stop fearing switching, the transformation of the market could be dramatic. I've just blogged on the subject, including a Tableau visualisation which you can use to analyse the underlying data: Bring more data: The School Information System market is changing faster than you think Thanks matt40k for liberating the data from the DfE - I had great fun looking at it!