'Cost models' for BSF

I should declare that I am strictly speaking neither edu nor geek - I am a school governor - apologies. I have been tangentially involved in the elaborate and expensive distraction that is the BSF Output Specification process.

One thing that is puzzling me is the 'cost model' produced by the LEP's insanely expensive consultants. How can (at least) two radically different approaches to delivering managed services be usefully represented by a single chart? Surely thin client/Open source approach has a different cost model to thick client/Microsoft based solution? Not only would I expect the total costs to be different, but I would also expect the proportion of costs to be different. The models could be compared, but not usefully merged. It also seems maddeningly abstract to talk of Total Cost of Ownership when fundamental technological choices have not been made. Maybe someone on this forum can explain...

My other sceptical thought is that assuming that this cost modelling was done in good faith for all authorities in the phases subsequent to ours how come they have previously got it so wrong? Our LEP is demanding schools sign-up to at least £200 per student (for a term yet to be decided but at least 5 years) when a neighbouring authority pays £60 a head and others range between £120-150. Simply put the cost model is useless in so many ways - it's use is to give a semblance of control where there is none.

Hmm, if the total costs are the same it makes me wonder if they've done the costing in reverse.

Company wants x profit, tupe staff will cost y, therefore we can spend z on kit.

A TCO of each cost could be put together (and should be) to allow comparison of approaches. A good TCO will give a good outline of how well though through the approaches are - how often equipment it to be replaced, staffing cost, software licensing, etc.

With a good TCO you should be able to see where the money is going and how long the life cycle is whether 5 or years more - my expectation is that most BSF IT hardware will be leased so you need to check if you terminate a contract they don't want everything back.

Best of luck trying to make sense of it!

I will my question here the charge for the service per pupil does that come out of capitation.

Russ

Quote:

---

Originally Posted by TerryTibbs

I should declare that I am strictly speaking neither edu nor geek - I am a school governor - apologies. I have been tangentially involved in the elaborate and expensive distraction that is the BSF Output Specification process.

One thing that is puzzling me is the 'cost model' produced by the LEP's insanely expensive consultants. How can (at least) two radically different approaches to delivering managed services be usefully represented by a single chart? Surely thin client/Open source approach has a different cost model to thick client/Microsoft based solution? Not only would I expect the total costs to be different, but I would also expect the proportion of costs to be different. The models could be compared, but not usefully merged. It also seems maddeningly abstract to talk of Total Cost of Ownership when fundamental technological choices have not been made. Maybe someone on this forum can explain...

My other sceptical thought is that assuming that this cost modelling was done in good faith for all authorities in the phases subsequent to ours how come they have previously got it so wrong? Our LEP is demanding schools sign-up to at least £200 per student (for a term yet to be decided but at least 5 years) when a neighbouring authority pays £60 a head and others range between £120-150. Simply put the cost model is useless in so many ways - it's use is to give a semblance of control where there is none.

---

This is one of most people's key worries about the BSF process. You seem to be indicating that you have the choice between Thin Client/Open Source and Thick Client/Microsoft. For a start the choice is nowhere near as simple as that, you can have Thick Client Open Source and Thin Client Microsoft. It is also not true to say that one choice will fit all uses within a single school, let alone an authority.

The "best" option is to retain the flexibility to have all the options you could need. This allows you to use Thin Client where this is appropriate and Thick Clients where this is appropriate. It allows you to use Open Source where appropriate and Microsoft/Adobe etc where appropriate. Unfortuately this is the most difficult to manage and cost up which is why I suspect the consultants are trying to force you down one route or other.

The TCO is in important factor to consider when trying to choose between the different technologies, but one has to consider that TCO is not just about desktops. For example, thin clients may have a lower desktop TCO than thick clients but the network and server infrastructure needs to be scaled up to pick up the load that would be done from the thin clients.

You also indicate that the school will be paying at least £200 per student per year for 5 years. The frustration that most of us techies have is that with that kind of budget we could easily introduce the revolutionary changes in education that are being sought, whilst retaining the flexibility of having it managed locally. Currently our budget for the last 3-4 years has been between £10 and £20 per student. Ok this does not include staff costs, but even so the dramatic improvements that are being promoted as part of BSF are being purchased by a dramatic increase in costs.

I hope this helps clarify some of the questions you have posed - was there anything more specific you were after?

Jonathan

Politely, the charge per pupil of £200 comes out of the consultants' arse. Since no bidder has yet been foramlly contacted, nor any service specified, it really is a grotesque form of wasting time and money. It is part of the BSF process to go through an extensive, sapping 'Output Specification' phase - which 'specifies' 'outputs' - ie nebulous competing lists of vague desires none of which are designed to be quoted against. Afaik its real role is to let potential bidders know there are rich pickings at our/any authority. (Apparently we're in a beauty contest - we need to assuage the fears of bidders who may have been burnt by wrongly guessing how much it costs to run schools - notably Morse, whoever 'won' Sheffield amongst others.)

My problem with the 'affordability model' that underlies the arse-conjured figure is that it is entirely abstract. It does not make even broad technical assumptions and is therefore dramatically worthless. Essentially it's more confusing guff to bamboozle the ill-prepared Heads and collected other staff called upon to endure the process.

Jonathan - I didn't mean to indicate that we had been presented with a choice. Imo, no matter how flawed, that would be a step up from the abstraction we currently struggle with. The giveaway about 'Total Cost of Ownership' is in the title - it's the total cost but these costs should be concrete and comparable - what we are presented with is a single cost for something entirely unspecified.

Tbh my assumption is that this is a deliberate strategy to exhaust participants and cloud costs so that no one can make value for money judgements. The august National Audit Office agrees that 'bench-marking' is sadly lacking from BSF procurement. (The Building Schools for the Future Programme: Renewing the secondary school estate) How did this ever happen? Bench-marking is a cornerstone of normal procurement.

Thanks for your replies - it's very helpful to discuss these things.

A managed service is basically out-sourcing. Why not use the SOCITM KPI's for benchmarking that appear to be widely used in the pubilc sector, the benchmarking has been in place since 1996 and has been amended since then. The KPI's are:

User Satisfaction
Resolution of reported incidents
Project Management
Aquisition cost of workstations
Support costs per workstation
Workstations supported per specialist
ICT comptetence of employees
Service availability
Investment in ICT by ICT user
Voice and Data communications
TCO for workstations
Flexible working
Management of printing
Elecronic service delivery
Investment in ICT per head of population.

Some of these could be used as is, some would not be required and some would need modifications, but using these as benchmarking would give a better picture of where ICT is and where it could/couldnt be in an outsourced managed service.

Interestingly everyone from Becta, P4s, goverment and local government talk about TCO but when did best value ever mean lowest cost? It doesnt. I was at a meeting on Friday organised by Becta and involving P4s, representatives from local authorities and some representatives from schools and it was interesting to see the divide that was there on opinion on a managed service.

Quote:

---

Originally Posted by TerryTibbs

I should declare that I am strictly speaking neither edu nor geek - I am a school governor - apologies. I have been tangentially involved in the elaborate and expensive distraction that is the BSF Output Specification process.

One thing that is puzzling me is the 'cost model' produced by the LEP's insanely expensive consultants. How can (at least) two radically different approaches to delivering managed services be usefully represented by a single chart? Surely thin client/Open source approach has a different cost model to thick client/Microsoft based solution? Not only would I expect the total costs to be different, but I would also expect the proportion of costs to be different. The models could be compared, but not usefully merged. It also seems maddeningly abstract to talk of Total Cost of Ownership when fundamental technological choices have not been made. Maybe someone on this forum can explain...

My other sceptical thought is that assuming that this cost modelling was done in good faith for all authorities in the phases subsequent to ours how come they have previously got it so wrong? Our LEP is demanding schools sign-up to at least £200 per student (for a term yet to be decided but at least 5 years) when a neighbouring authority pays £60 a head and others range between £120-150. Simply put the cost model is useless in so many ways - it's use is to give a semblance of control where there is none.

---

yes terry, the total costs should be different, as are the proportion or allocation of costs. For instance a thin client solution is necessarily very much backend heavy, so much of the costs will come in terms of server hardware and server software investments. Realistically a traditional fat client solution can be developed which doesn't have as significant an investment in the backend infrastructure. But even that is very simplistic and that perception soon changes as you properly investigate the various thin client solutions. And trust me, there a huge amount of vendors in this space.

The TCO is always wheeled out for thin client vs fat client analysis, irrespective of what the solution actually is.

And this is daft for a number of reasons, mainly becuase there is no such thing as A thin client solution, there are many different types and depending on how far you want to replace rather than compliment a thick client solution the cost of ownerships can be vastly different -

for instance if i go with an xpe thin client VDI solution where the end devices consume say 25w of power those devices need updating and patching just like a regular windows pc, the VDI image needs to be maintained, whereas a zero client solution may consume 5w of power at the client and requires no patching and same issues with VDI image mgmt. Explain the tco differences in such a scenario, or the tco if i choose to go with a hybrid solution. Some thin and hybrid solutions have a dizzying array of options and add-ons, what is the tco of going with a thin client solution with no options vs going all-in and purchasing all licensing options. Also, there are hybrid solutions where thin client comes in the form of a software service runing on fat client hardware, such a solution should atleast be offered as an alternative by any provider.

If the provider wants to say, this is thin client as a strict definition, we are not considering zero client or hybrid, than does that mean no-one wants to consider other viable options ? or a mixed approach ?
this is the kind of stuff you can find out in meetings once you get down to the technical brass tacks.

I always look to cut through the gumph when i go to meetings, it's important to get into technical issues quickly and ask pertinent questions rather than just accept the wishy washy 1000ft overview that is often presented. If you can't get product specific responses to your question then such exercises in 1000ft overview analysis of thin client vs thick client and perceived TCO is utterly meaningless in my view. simply because the options are so vast that a generic thin client vs thick client analysis can't be had. At primary school class maybe, but not in the real world.

Quote:

---

How can (at least) two radically different approaches to delivering managed services be usefully represented by a single chart? Surely thin client/Open source approach has a different cost model to thick client/Microsoft based solution? Not only would I expect the total costs to be different, but I would also expect the proportion of costs to be different.

---

You're absolutely right to question this. Two completely different models should have two different costs associated with their setup and running expenses. If there was no difference, you have to ask what the incentive for thin clients would be.
Given the choice I would choose fat clients over thin clients if the costs were identical, in a theoretical choice situation, but in my mind something is very wrong as they shouldn't be the same at all.

£200 per pupil does seem excessive when you have a comparative with other authorities which are much lower. I find the whole BSF idea a huge mistake. The Government calling for outstanding learning establishments and managed solutions, but with such volatile differences in opinion/consultancy and costs, again you're absolutely right to question it.

When managed services for BSF were first proposed more than 4 years ago, the 'cost' per student was envisaged to be £60-£80 per student per year, a figure plucked out of 'thin air' & designed to convince HTs & Governors that managed services would cost schools less than their in-house services.

As things have progressed, the costs have gone up & the schools are getting less :(

At a school I know of, they were told they should have an aspirational goal of 1:1 computers : pupil ratio. It is now looking to be closer to 1:3, far worse than their current ratio of 1:2.......