Alex Holmes, Deputy Director and Chief of Staff in the Office of the CTO, tells us why government has moved away from the “Tower Model” for running IT services.
When we published the Technology Code of Practice and resources for chief technology officers, they set out a vision for what good IT looks like.
A fundamental part of our guidance was about taking accountability for decisions about technology and digital services back into government. For large parts of the Civil Service that had so completely outsourced their IT, this meant a massive shift in approach, which takes time and can be scary.
This fear of change meant some organisations clung onto the concept of outsourcing, which they understood, but they also wanted to comply with the new policy of multi-sourcing IT provision - something that is recognised as best practice across the industry.
Tower Model
Unfortunately, the combination of these two forces created a hybrid model unique to government. The model is usually referred to as the Tower Model. It combines outsourcing with multi-sourcing but loses the benefits of either. The model has arisen because organisation have used a procurement-led solution in response to legacy outsourcing contracts ending. Rather than changing their approach and emphasis, they have ended up outsourcing their IT again, but in pieces.
It was still all about us, not about the needs of our users.
Organisations have adopted the Tower Model, believing they are following government policy and using best practice, but they are doing neither. I am now writing this post to be clear that the Tower Model is not condoned and not in line with Government policy. Government should use the best of what is already out there - not develop its own model.
Multi-sourcing
As the programme that has replaced IT services in Cabinet Office, DCMS and the Crown Commercial Service has shown, multisourcing can deliver over 40% savings, delivering services that transform how people work, quickly.
Multi-sourcing works because at its core is an understanding of user needs. The team responsible for the IT service knows what services are needed to currently deliver those needs. The team, which is largely made up of civil servants, owns the accountability and architecture. Departments should make sure they have the capability to do this.
An important point about multi-sourcing is that different things are bought in different ways: there is no “one size fits all” methodology. Commodity products like hosting will still likely be outsourced to utility suppliers, but novel or unique things close to the user may be built in-house. And components can - and will - be changed often.
The Tower Model doesn’t work because it doesn’t fully consider what services are needed, or how they fit together and it uses a “one size fits all” methodology. It relies on procurement requirements to bundle together vertically-integrated outsourcing contracts called things like ‘network’ or ‘desktop’. It also usually outsources the service accountability, architecture and management to a third party.
There can be a role for this Service Integration and Management (SIAM) layer, but placing too much responsibility with it increases risk for both the department and the supplier by confusing roles. The SIAM provider should not replace good in-house IT capability.
The Tower Model can create a situation where the customer buys a number of incompatible parts and then asks a SIAM provider to put them together and make it work.
Own the solution
What’s much more effective is to design and own the overall solution so that you know it works and can put it together and run it yourself. Or you own the solution, put it together yourself and you use a SIAM provider to run it.
Both of these options require a number of steps well before looking at procurement options. Tom Read’s blog post explains the steps you might take in selecting solutions.
Generally speaking, you should understand user needs, bring in the right capability and skills, analyse existing applications, architect a disaggregated desktop using cloud infrastructure and consider platform options before procuring and commissioning what’s needed.
Follow the Cabinet Office Technology blog for more detail on what worked there.
You can follow Alex on Twitter and don’t forget to subscribe to the Government Technology blog.
17 comments
Comment by Tax Payer posted on
If Government marshaled its resources to provide the financial latitude to secure the best knowledge and IT skills in the market place on a sustainable basis, this model might work. But in reality government has been unsuccessful in securing and maintaining the best talent and has therefore had no choice but to work with systems integrators to secure the expertise at scale required to meet demand.
Also this further disaggregation of services will add yet more complexity (read RISK) to the service management overhead of ensuring this multi-source fabric can be governed to provide high quality services.
There is no doubt that platforms and utility providers will become more ubiquitous. Leveraging these commodities 'building blocks' will be an essential part of providing value for money and effective services for the UK tax payer. But please please please, don't use this strategy as a catalyst to commission cottage industries in government that produce ill-considered and expensive bespoke systems in the name of user / citizen centricity.
Yours a concerned UK tax payer.
Comment by Kevin Holland posted on
Thanks Alex, this provides a refreshing and welcomed clarity that should prevent disaster down the line. My forthcoming AXELOS whitepaper on an example SIAM model will provide specific examples of the challenges of the Tower model.
Comment by Loz posted on
Government is the issue, not what model they deploy to manage their IT business.
Comment by tax payer posted on
Cabinet Office approve all ICT spend & monitor delivery across government departments? So if the majority of the departments have misinterpreted the strategy why have Cabinet Office allowed this to happen? It would be interesting to run a cross government analysis to record the amount of time and effort wasted as a direct result of these continual cabinet office led initiatives. My practical experience of dealing with cabinet office is that their strategies are ill thought out, poorly documented and not born from practical experience. As for their approvals process its a minimum of 3 months from start to finish to navigate their paper based governance process and that's if you are lucky enough to pass through first time round.
There are many areas of government that have a proven track record of successfully delivering ICT services for their business. Why don't we leverage and learn from these "real life" experiences of best practice rather than bringing in groups of people who have an academic view of the world with little to no hard core delivery experience.
Comment by John posted on
I am afraid that this article does something that government is rather good at – finding something to blame to divert attention or to draw attention to something else. In this case they are blaming the ‘Tower Model’ without defining what they mean by this phrase – it becomes a useful football to kick when trying to make a point. The underlying point about multi-sourcing is a good one, but setting something up to shout ‘yah, boo, sucks’ at it to allow one to make a noise to draw attention to something is lazy.
The authors claim that: "... the Tower Model doesn’t work because it doesn’t fully consider what services are needed, or how they fit together and it uses a “one size fits all” methodology. It relies on procurement requirements to bundle together vertically-integrated outsourcing contracts called things like ‘network’ or ‘desktop’. It also usually outsources the service accountability, architecture and management to a third party... "
My experience is that failures of the Tower Model are not an inevitable result of using the tower model per se, but merely the result of using any approach without the skills or understanding to use it properly.
My view is that one man’s tower is another man’s service in a multi-sourced environment – the scope or constituent parts of a tower are defined by business need, just as in a service. Having defined your towers or services to meet business needs, you go ahead and decide where best to source it (in or out), and then procure it. So, the tower model and multi-sourcing are not, as the author seems to imply, alternatives, but live very happily together if used properly. In fact – create a Sourcing Strategy that defines business requirements, groups them into coherent services (or towers), and then decides, based on their criticality to the business and the ability of the market place, where to place their delivery.
Finally, the author trots out what I feel is a very simplistic view – ‘… you should understand user needs …’. It seems to me that much more important is business (or departmental) needs. User needs are usually incompatible with each other (every user wants something different) and also with the commercial, political and legal obligations of the department for whom they work. By all means listen to user needs, but they should not take priority over what the department needs to do to meet its obligations.
So – lots of noise to make a point that is like to curates egg – good in parts, but ultimately still only an egg.
Comment by Kris posted on
What is a "disaggregated desktop"?
Comment by Alan T posted on
A hindsight-helped statement of the obvious from a guy who’s never been on the end of this misguided, but well-meaning, road-map to a psychiatric ward.
Comment by The Northern Taxpayer posted on
Two points.
In the real world of complex businesses delivering highly complex integrated / non-integrated services from a mixed economy of in sourced and outsourced services, often servicing tens of thousands of system users and usually millions of citizens, having highly competent Service Integration and Management capability is not old hat - its current and vital. SIAM emerged as a result of multi sourced IT infrastructure provision not providing an end to end service (my computer isn't working - is it the desktop, the network, the hosting or the application layer?). Is the author really proposing that we really regress to the unjoined up it services of the past?
Government embarked on the wholesale outsourcing of It in the late 90's for many reasons, but one of those was the lack of ability to attract and retain skilled resources. From my very recent experience the major Departments of state are really struggling to attract IT / Digital talent as employees. Most Departments simply do not have the skills needed in house to define requirements let alone implement, manage them and subsequently innovate them and I am not seeing in any part of government the changes needed to attract this talent, such as doing away with the ridiculous competency based approach to recruitment, which just gets people turned off at the prospect of writing 150 words on six competencies when all they want to do is press send on their iPhone during their commute to apply with their their cv. To disagree with Tax Payer, shockingly government pays reasonably well (to £65k for middle management), it's the culture and environment that puts talent off applying.
There is a very long way to go....
The Northern Taxpayer
Comment by Jonathan posted on
Thanks Alex. Does this mean that central government departments currently implementing Tower and SIAM models will be required to stop (e.g. the Ministry of Justice?). Interesting that other public bodies like the Met Police are currently implementing this approach as well. It would be interesting to provide more details about the Cabinet Office approach so that other government departments are won over to different ways of approaching this (if it's not too late given these are mega-programmes over multi-years). Looks like different views prevail on what works!
Comment by Andrew Buxton posted on
Government / Public sector organisations have traditionally gone to the Large Service Integrators and outsourced required services. However recent Government purchasing guidelines require organisations to multisource (preferably to SME’s) rather than follow the path of traditional monolithic Outsourcing deal.
To answer the challenge , the larger SI companies have responded by breaking down outsourcing requirements into deliverables and re-invented themselves as Service Integration and Management providers (SIAM). They created a new multisourcing / outsourcing hybrid, the Tower Model to deliver services back into government. Basically a bastardised version of multisourcing.
A SIAM - Tower Model is easy to understand and provides a single point of contact , but essentially it is ‘the emperor’s new clothes ‘ - the same old thing with a new structure and label, Outsourcing dressed up as a tower model with a SIAM layer included.
In your article you are recommending bringing the Service Integration and Management back in house …which beggars the question “is there the appetite to upskill your supplier management teams to be effective Multisourcing governance professionals?”
Alternatively what about considering a true Multisourcing Service Integration provider? Who would deliver overarching governance, retain independence and impartiality from both vendor and service providers/ suppliers, and work within a predefined multisourcing Governance model. Most importantly the MSI does not deliver any of the services, only the governance and so sits squarely CLIENT –SIDE.
Government procurement process is very prescribed and dogmatic, so the added benefit of the intervention of a Multisourcing Service Integrator at the service definition and design stages would ensure that the client organisations, end user and business addressed, rather than retro fitted to a one size fits all tower service.
What’s not to like?
Comment by 'We're all in this together' posted on
'm not sure who suggested the tower model was official government policy. However, in the absence of a clear government technology strategy it does provide a way of working with a range of large and smaller suppliers who have breadth of experience in a particular field. Given no government department has delivered a tower model yet it is a little early to write-off.
I absolutely get that GDS has been set up to challenge the status quo of supplier government procurement and technology. You do a good job in that respect. The financial controls have saved money by stopping projects and your exemplars (when given a blank canvas to work on) are starting to deliver some well put together smaller systems. What is less transparent is how much the Digital teams cost or how they provide value for money, but it is a start.
What GDS does not have is experience delivering larger infrastructure or complex line of business applications. In this respect I mean large rather than Cabinet Office and department of media and Sport size. What would go a long way to help would be to engage constructively with government departments and suppliers (even the large ones which we demonise so enthusiastically). A little humility in achievements might also help. If the tower model is not government policy, what is? Your blog does not make that clear. When will it be published and who will deliver it?
Comment by Baz posted on
This is helpful, but no revelation.
We should be wary of fashions and fads (adherence of fools in technology design and development) and focus instead in what approaches best fit our needs and, more instinctively these days, the depth of our pockets. Look forward to GDS apology for costly dogmatism over the last two years, that almost crippled a major programme (am sure there are others). For now, we will look forward.
Comment by Max posted on
The other thing the author completely neglects is that most departments are under significant pressure to maintain or reduce their headcount, particularly in corporate services. The Cabinet Office, and GDS in particular, do not appear to have a similar constraint. How many staff in GDS now? 700? And still hiring..
Comment by Nicky Stewart posted on
It has been a little over three weeks Alex blogged on the demise of the tower model. Reverberations continue in the trade press even now. In the furore it is easy to lose sight of the simple, but important, message: “It is time for government to put user need first and take back control of its IT”.
There has been outrage at the way Holme’s made his points: a blog! Its content hadn’t even been discussed with industry. How dare GDS announce policy changes in such a gung-ho way.
The collective amnesia is staggering. The tower model is not, and never was formal government policy. It was never articulated, even implicitly, in any published government ICT strategy, or in the GDS Technology Code of Practice. The tower model was only ever an attempt to break supplier monopolies, introduce competition, remove duplication of multiple systems operating across multiple agencies within the same department, and enable departments to get a better grip on their ICT spend.
The tower model served its purpose but should be seen for what it is – the first step on a long journey towards government wresting control of its IT, and casting off the shackles caused by decades of wholesale outsourcing. It was never the end game.
Curiously, in the noise there has been a deafening silence. The vendors with the most vested interests in the tower model have been unusually quiet. They have a lot to defend. Holmes’s blog strikes at the heart of SIAM – precisely where the incumbent vendors have been re-positioning to ride out the changes of the last five years: digital by default, G-Cloud, and the SME agenda. SIAM is fundamental to any multi-sourcing model, but has become an artefact crafted by industry, for industry. Little wonder then, that their trade associations and analysts are working very hard for their subscriptions by calling into question government’s ability and capability to take back the control of its own IT, and are even suggesting that a move from the tower model could dissuade investment in the public sector from companies of all size.
The “good in-house IT capability” that Holmes writes about is as potentially huge threat to the SI as it is a huge opportunity for government. Holmes’s blog sets out government’s strategic intent. Government doesn’t need to have lengthy discussion with the industry to agree policy and set strategy. It has had enough experience (mostly painful) of wholesale outsourced IT, to have a thorough understanding of the issues, where it needs to go, and what it needs to do to get there.
Government as a Platform is the next stage of government’s digital journey. It is high time that the industry put the needs of its customers, their users, and the taxpayer first, and got behind Holmes, GDS and government to make it happen.
Comment by Ed posted on
Hi Nicky. Are you the same Nicky Stewart, Commercial Director of Skyscape, a company who has done very well (£9m) since the introduction of G-Cloud? Not that I begrudge you your success, but when complaining about vested interest it iscustomary to declare your own. If not, what a coincidence.
But to Alex's post. If there was no specific policy for the public sector to adopt a Towers model, then the Cabinet Office were doing a very good job of suggesting that was the right approach. The Government Procurement Service spent most of 2012 working on a SIAM framework and engaged with a wide variety of departments to get their input. And GDS similarly engaged with departments to endorse their procurements for disaggregation, defining logical service boundaries and attempting to make the Towers compelling enough to engage the SME market. Sure, there was never a blog post to announce the policy, but it was pretty clear as to the direction of travel.
But what I think is most significant about Alex's post, is that it IS a major policy announcement with far reaching implications that you would normally expect to be announced by a Minister. Whilst not explicit, it is clear that this blog post is announcing the in-sourcing of ICT provision in the public sector which is significant news for job creation and normally accompanied by a..."Bong...the Secretary for Work and Pensions announced a significant shift in Civil Service resourcing strategy today with the creation of thousands of new jobs as the public sector begins to insource ICT provision".
It's really rather good news.
Comment by Nicky Stewart posted on
Hi Ed,
Interesting post, and of course I would like to see your final point a reality. Great that we agree that. Like you we support government doing the right thing. But to answer your points:
1) Yes, I am Skyscape's Commercial Director. As an individual I am entitled to make individual comment. Serious issues need to be debated openly and not hidden behind the anonymity of the web. So I ask the same of you. Who are you and who do you represent?
2) Why would Skyscape have a vested interest in SIAM? We do IaaS and PaaS - very far removed from SIAM and G-Cloud Lot 4. We succeed in what we do but have no ambition to extend further. Commodity pure play cloud is the polar opposite of SIAM. Appreciate your acknowledgment of Skyscape's success, but it's not really the point here,
3) Ref CCS SIAM framework.- it was never tendered. DSF debates?
Like you, I believe policy is announced by Ministers, and amended in that same process. There has been no policy announced by that process here at any point. Hope this clarifies.
Comment by Ed posted on
I apologise if I was unclear, English is not my first language. What I was trying to say was that you, as the Commercial director of Skyscape was representing a company that was financially benefiting from the policy that Alex was blogging about and you wrote a comment to support. So, you have a vested interested in that policy. Do you see? And you're quite right, you are more than entitled to comment as an individual, though it is somewhat hypocritical to complain of vested interests and yet not declare your own. Hopefully I have explained that, and I apologise again if I was not clear. For my part, I have no vested interest and I represent myself.
I don't know what DSF means. But if such a critical department as the Cabinet Office spends a year investing time and effort into developing a framework to support a particular operating model I would hope that they did so as a policy. Not for fun.
I must admit to being similarly perplexed by your last point. Alex (not the Minister for Work and Pensions) seems to be saying that it is now Government policy that ICT provision should be in-sourced, but there has been no corresponding policy announcement to underwrite the change in resourcing model, allocation of budget and the creation of thousands of new Civil Service posts. You said "How dare GDS announce policy changes in such a gung-ho way" so I interpreted that as you saying that GDS was announcing a policy change. The policy change from SIAM (that wasn't policy) to in-house (also not policy?). So no, it doesn't clarify anything for me.