Hosting Crowns

Late in 2013 there was a flurry of interest in a project called the “Crown Hosting Service” – covered, for instance, by Bryan Glick at Computer Weekly. The aim, according to the article, was to save some £500m within a few years by reducing the cost of looking after servers.  The ITT for this “explicit legacy procurement” (as Liam Maxwell accurately labelled it) was issued in July 2014.

Apparently some £1.6bn is spent by government on hosting government’s IT estate.  That figure is about half what it costs to run government’s central civil estate (buildings); and that £3bn is only 15% of the cost of running the total estate.  The total cost of running the estate is, then, something like £20bn (with an overall estate value of c£370bn).

It’s interesting, then, to see increasing instances of department’s sharing buildings – the picture below shows two agencies that you might not associate together.  The Intellectual Property Office and the Insolvency Service share a building – though I’m hoping it’s not because they share a customer base and offer a one stop shop.  The IPO and the IS are both part of BIS (which is just around the corner) so perhaps this is a like for like share.

But over the next couple of years, and maybe in the next couple of months, we are certainly going to see more sharing – DCLG will soon vacate its Victoria location and begin sharing with another central government department.  Definitely not like for like.
Such office sharing brings plenty of challenges. At the simpler end things such as standard entry passes and clearance levels.  At a more complicated level is the IT infrastructure – at present something that is usually entirely unique to each department.  A DCLG desktop will not easily plug straight into the network of another department – even a wireless network would need to be told about new devices and where they needed to be pointed at.
With increasing commoditisation of services, and increasing sharing, it’s easily possible to see – from a purely IT point of view – government buildings that function, for large numbers of HQ and, perhaps especially, field staff, as drop in centres where desks are available for whoever is passing provided that they have the right badge.  Those who want to work from home can continue to do so, but will also be able to go to a “local office” where they will have higher bandwidth, better facilities and the opportunity to interact with those in other departments and who run other services.  
In this image, the vertical silos of government departments will be broken up simply because people no longer need to go to “their” department to do their day job, but they can go wherever makes most sense.  Maybe, just maybe, the one stop shop will become a reality because staff can go where the customers are, rather than where their offices are.

G-Cloud By The Numbers (To End June 2014)

With Dan’s Tableau version of the G-Cloud spend data, interested folks need never download the csv file provided by Cabinet Office ever again.  Cabinet Office should subcontract all of their open data publication work to him.

The headlines for G-Cloud spend to the end of June 2014 are:

– No news on the split between lots.  80% of spend continues to be in Lot 4, Specialist Cloud Services

– 50% of the spend is with 10 customers, 80% is with 38 customers

– Spend in June was the lowest since February 2014.  I suspect that is still an artefact of a boost because of year end budget clearouts (and perhaps some effort to move spend out of Lot 4 onto other frameworks)

– 24 suppliers have 50% of the spend, 72 have 80%.  A relative concentration in customer spend is being spent across a wider group of suppliers.  That can only be a good thing

– 5 suppliers have invoiced less than £1,000. 34 less than £10,000

– 10 customers have spent less than £1,000. 122 less than £10,000.  How that boxes with the bullet immediately above, I’m not sure

– 524 customers (up from 489 last month) have now used the framework, commissioning 342 suppliers.  80% of the spend is from central government (unsurprising, perhaps, given the top 3 customers – HO, MoJ, CO – account for 31% of the spend)

– 36 customers have spent more than £1m.  56 suppliers have billed more than £1m (up from 51).  This time next year, Rodney, we’ll be millionaires.

– Top spending customers stay the same but there’s a change in the top 3 suppliers (BJSS, Methods stay the same and Equal Experts squeaks in above IBM to claim the 3rd spot)

One point I will venture, though not terribly well researched, is that once a customer starts spending money with G-Cloud, they are more likely to continue than not.  And one a supplier starts seeing revenue, they are more likely to continue to see it than not.  So effort on the first sale is likely to be rewarded with continued business.

Close Encounters Of The Fourth Kind

Much to my surprise, O2 sent me a text on Wednesday.  The text wasn’t the surprising bit – they often send me texts offering me something that I don’t want.  The surprise was that this time they didn’t offer, they told me that I was just going to get it.  It was 4G.  And I wanted it.

4G. For nothing.  No options. No discussion. No questions allowed.  No SIM change needed.  No conversation about impact on battery life.  Just: turn off your phone in the morning and turn it back on, and if you’re in a 4G area, it’s all yours.

The next morning I was in Camden Town – no sign of 4G there.  Clearly Camden is a bit rural to have coverage just yet.

But later, in Whitehall, it worked just fine.  And fine means a consistent 15mb/s download (versus the previous day’s 3G download speed of 2mb/s).

During 2012/13 I set up a JV owned by the four mobile operators, called at800, that had the task of managing any negative impact from interference with TV signals that might occur because the bottom of the 4G range aligns with the top of the TV range (and, until some recent work by Digital UK, overlapped).

at800 – you might have seen the ads or had a card, or maybe even a filter through your letterbox – has been a great success (last I checked, they’d been in touch with probably 45% of UK households).  That’s in part because the problem that all the TV technologists worried might affect up to two million households has actually been far less of a problem but, for the most part, it’s because we put together a great team, worked closely with the mobile operators and the broadcasters, ran pilots, tested everything we could and smoothed the way for the 4G roll out.  In truth, we were ready long before the operators were.  They were all fun/challenging/annoying/exciting to work with, but I liked O2’s approach most.

After a couple of days testing 4G, I have this to say:

– Coverage in buildings where 3G coverage was previously poor to non-existent has much improved (I can even make calls from the dead centre of buildings where previously I stared only at “No Service”)

– Download speeds are certainly faster (roughly equivalent to what I get from Satellite broadband, but without the round trip lag)

– Battery life seems unchanged (I wonder if battery usage is higher during download but because download is so much faster, there’s less overall drain)

That said, the nearest mast to my home is still some 200 miles away.  Keep rolling it out O2.

I have no idea how widespread this offer is but, if you get the same text, say “yes”. Not that you’ll have any choice.  But so far, it’s all upside.

The Trouble With Transition – DECC and BIS Go First

In a head-scratching story at the end of last week, DECC and BIS made
the front page of the Financial Times (registered users/subscribers can access the story). 
Given the front page status, you might imagine that the Smart Meter
rollout had gone catastrophically wrong, or that we had mistakenly paid
billions in grants to scientists who weren’t getting the peer reviews
that we wanted, or that we’d suddenly discovered a flaw in our model for
climate change or perhaps that the Technology Strategy Board had made
an investment that would forever banish viruses and malware.

The BBC followed the story too.

no.  Instead we have two departments having problems with their email. 
Several Whitehall wags asked me weeks ago (because, yes, this story has
been known about for a month or more) whether anyone would either
notice, or care, that there was no email coming to or from these
departments.   It is, perhaps, a good question.

Business Secretary Mr Cable and Energy and Climate Change Secretary Mr Davey were reported in the Financial Times
to be angry about slow and intermittent emails and network problems at
their departments since they started migrating to new systems in May.

The real question, though, is what actually is the story here?

It appeared to be a barely-veiled attack on the current policy of
giving more business to SMEs (insider says “in effect they are not
necessarily the best fit for this sort of task” … “an idealistic Tory
policy to shake up Whitehall”)

– Or was it about
splitting up contracts and of taking more responsibility for IT delivery
within departments (Mr Cable seemingly fears the combination of
cost-cutting and small firms could backfire)?

–  Was the story leaked by Fujitsu who are perhaps sore at losing their £19m per annum, 15 year (yes, 15. 15!) contract?

– Was
it really triggered by Ed Davey and Vince Cable complaining to the PM
that their email was running slow (“Prime Minister, we need to stop
everything – don’t make a single decision on IT until we have resolved
the problems with our email”)?

– Is it even vaguely possible that it is some party political spat where the Liberal Democrats, languishing in the polls, have decided that a key area of differentiation is in how they would manage IT contracts in the future?  And that they would go back to big suppliers and single prime contracts?

– Was it the technology people
in the department themselves who wish that they could go back to the
glory days of managing IT with only one supplier when SLAs were always
met and customers radiated delight at the services they were given?

#unacceptable as Chris Chant would have said.

Richard Holway added his view:

In our view, the pendulum has swung too far. The Cabinet Office refers
to legacy ICT contracts as expensive, inflexible and outdated; but
moving away from this style of contract does not necessarily mean moving
away from the large SIs.

And it appears that it is beginning to dawn on
some in UK Government that you can’t do big IT without the big SIs. A
mixed economy approach – involving large and small suppliers – is what’s

By pendulum, he means that equilibrium sat
with less than a dozen suppliers taking more than 75% of the
government’s £16bn annual spend on IT.  And that this government, by
pushing for SMEs to receive at least 25% of total spend, has somehow
swung us all out of kilter, causing or potentially causing chaos.  Of
course, 25% of spend is just that – a quarter – it doesn’t mean (based
on the procurements carried out so far by the MoJ, the Met Police, DCLG
and other departments) that SIs are not welcome.

Transitions, especially, in IT are always challenging – see my last blog on the topic
(and many before).  DECC and BIS are pretty much first with a change
from the old model (one or two very large prime contracts) to the new
model (several – maybe ten – suppliers with the bulk of the integration
responsibility resting with the customer, even when, as in this case,
another supplier is nominally given integration responsibility).  Others
will be following soon – including departments with 20-30x more users
than DECC and BIS.

Upcoming procurements will be
fiercely competed, by big and small suppliers alike.  What is different
this time is that there won’t be:

–  15 year deals that
leave departments sitting with Windows XP, Office 2002, IE 6 and dozens
of enterprise applications and hardware that is beyond support.


15 year deals that leave departments paying for laptops and desktops
that are three generations behind, that can’t access wireless networks,
that can’t be used from different government sites and that take 40
minutes to boot.


– 15 year deals
that mean that only now, 7 years after iPhone and 4 years after iPad,
are departments starting to take advantage of truly mobile devices and

With shorter contracts, more competition,
access to a wider range of services (through frameworks like G-Cloud),
only good things can happen.   Costs will fall, the rate of change will
increase and users in departments will increasingly see the kind of IT
that they have at home (and maybe they’ll even get to use some of the
same kind of tools, devices and services).

To the
specific problem at BIS and DECC then.  I know little about what the
actual problem is or was, so this is just speculation:

We know that, one day, the old email/network/whatever service was
switched off and a new one, provided by several new suppliers, was
turned on.  We don’t know how many suppliers – my guess is a couple, at
least one of which is an internal trading fund of government. But most
likely not 5 or 10 suppliers.

– We also know that
transitions are rarely carried out as big bang moves.  It’s not a
sensible way to do it – and goodness knows government has learned the
perils of big bang enough times over the last 15 years (coincidentally
the duration of the Fujitsu contract).

– But what
triggered the transition?  Of course a new contract had been signed, but
why transition at the time they did?  Had the old contract expired? 
Was there a drive to reduce costs, something that could only be
triggered by the transition?   

– Who carried the
responsibility for testing?  What was tested?  Was it properly tested? 
Who said “that’s it, we’ve done enough testing, let’s go”?  There is,
usually, only one entity that can say that – and that’s the government
department.  All the more so in this time of increased accountability
falling to the customer.

– When someone said “let’s
go”, was there an understanding that things would be bumpy?  Was there a
risk register entry, flashing at least amber and maybe red, that said
“testing has been insufficient”?

In this golden age of
transparency, it would be good if DECC and BIS declared – at least to
their peer departments – what had gone wrong so that the lessons can be
learned.  But my feeling is that the lessons will be all too clear:

– Accountability lies with the customer.  Make decisions knowing that the comeback will be to you.

– Transition will be bumpy.  Practice it, do dry runs, migrate small numbers of users before migrating many.

Prepare your users for problems, over-communicate about what is
happening.  Step up your support processes around the transition

– Bring all of your supply chain together
and step through how key processes and scenarios will work including
when it all goes wrong.

– Have backout processes that you have tested and know the criteria you will use to put them into action

don’t come along very often.  The last one DECC and BIS did seems to
have been 15 years ago (recognising that DECC was within Defra and even
MAFF back then).  They take practice.  Even moving from big firm A to
big firm B.  Even moving from Exchange version x to Exchange version y.

this story isn’t, in any way, is a signal that there is something wrong
with the current policy of disaggregating contracts, of bringing in new
players (small and large) and of reducing the cost of IT).

The challenge ahead is definitely high on the ambition scale – many large scale IT contracts were signed at roughly the same time, a decade or more ago, and are expiring over the next 8 months.  Government departments will find that they are, as one, procuring, transitioning and going live with multiple new providers.  They will be competing for talent in a market where, with the economy growing, there is already plenty of competition.  Suppliers will be evaluating which contracts to bid for and where they, too, can find the people they need – and will be looking for much the same talent as the government departments are.  There are interesting times ahead.

There will be more stories about transition, and how hard it is, from
here on in.  What angle the reporting takes in the future will be quite fascinating.

More On G-Cloud Numbers (May 2014 data)

The latest data show increasing spend via G-Cloud – this month tantalisingly close to the £200m arbitrary round but important number level at £191.6m.  The news after that is not terribly interesting:

Cloud spending may, it turns out, be seasonal.  Spend last month dropped to £12m, the lowest seen since October 2013 – all the more noticeable after the bollard budget blockbuster that was March spending.  Start of the new financial year and everyone is, it seems, planning rather than doing.

Lot 4 continues to dominate with 79% of the spend (Lots 1 to 3 are 6%, 1% and 13% respectively).

Much of the rest – top spending customers, top earnings suppliers etc – stays the same.

But there are some anomalies.  Last month I reported that the lowest spending customer had spent only £63.50.    This month they’ve moved higher with £85.90.  Thirteen customers have, though, still spent less than £1,000.

We do, though, have nearly 500 customers (489) which is, in my view, more important than the growth in spend – it shows either (a) that more people are looking at what the cloud can do for them, which would be good all round or (b) that more people have found that G-Cloud as a framework, GCaaS, can help them which is still good because it’s transparent and we can see whether they spend more in the coming months.

51 suppliers have seen revenues of more than £1m. Some of those are brand name, paid up, members of the Oligopoly.  Others look new to the public sector and certainly new to having access to quite so many customers.

There are some other anomalies too – I assume the result of data capture errors.  One supplier has a revenue line showing £1,599,849.80 which is listed as “blank” – there are 9 other such lines, though the other numbers are far, far smaller.  It would be nice to know where to allocate that money.  It may be that it is correctly allocated by Lot (so shows up in the graph below where there are no “blank” entries) but not correctly tagged with a product description.  Still be nice to know.

A couple of other points to wonder about:

– The Crown Commercial Service are a bigger user of Skyscape than any other purchaser (£1.4m – double HMRC spend, neary triple Cabinet Office spend, nearly 5 times Home Office).  Is that all hosting of the G-Cloud store and other framework services?

– There are only 125 instances of the word “Cloud” in the line items of what has been purchased (which run to over 2,000 separate lines)

To repeat the last paragraph in my last entry on this topic, for the avoidance of doubt:

Still, there is no other framework in government that gives access to
such a wide variety of suppliers (many new to the public sector) and no
framework that publishes its information at such a transparent level. 
For those two reasons alone, G-Cloud still deserves applause – and, as
it grows month on month, I hope that it will only get louder. 

Digital Government 2002 – Doing Something Magical

Now here’s a blast from the past!  Here’s a “talking head” video recorded, I think, in early 2002 all about e-government (I am, of course, the talking head).  Some months later, much to my surprise, the video popped up at a conference I was attending – I remember looking up to see my head on a dozen 6′ tall screens around the auditorium.

It’s easily dated by me talking about increasing use of PDAs (you’ll even see me using one) and the rollout of 3G, not to mention the logo flashing up in the opening frames and e-government, as opposed to Digital By Default.

But the underpinning points of making the move from government to online government, e-goverment or a Digital by Default approach are much the same now as then:

“The citizen gets the services they need, when they need them, where they need then, how they need them … without having to worry about … the barriers and burdens of dealing with government”

“You’ve changed government so fundamentally … people are spending less time interacting and are getting real benefit”

Lessons learned: get a haircut before being taped, learn your  lines, even when in America don’t wear a t-shirt under your shirt (my excuse is that it was winter).

G-Cloud By The Numbers (April 2014 data, released mid-May 2014)

I haven’t looked at the G-Cloud spend data for a few months (the last review was in December) – something changed with the data format earlier in the year and it screwed up all my nicely laid out spreadsheets; I’ve only just got round to reworking them.

– After 25 months of use, total spend via the framework is £175.5m

– Spend in all of 2013 was £85m.  Spend in the first 4 months of 2014 is £81m, about 46% of the total spend so far

– The run rate for 2014, if that spend rate continues, is perhaps more than £240m.  I suspect we could see much higher than that given the expiry of many central government IT contracts in 2015 and 2016 (and so an increase in experimentation, preparation for transition and even actual transition ahead of expiry)

– The split between the lots in December 2013 was Lot 1: 4%, Lot 2: 1%, Lot 3: 16%, Lot 4: 78%

– As of now, the split is similar: 6%, 1%, 14%, 79%
– The 2014 year to date split is little different: 8%, 1%, 12%, 80%

Conclusion:  The vast bulk of the spending is still via Lot 4 – people
and people as a service.  I’d expected that to start changing now, with the Digital Services Framework fully live.  That said, Lot 4’s spend per month has changed little since November 2013, except for a peak of £23m in March (roughly double the average spend over the last 6 months) which you can easily see in the graph above

Conclusion: Infrastructure as a Service (from Lot 1) is gradually increasing – it’s gone from c£800k/month to c£1.5m a month in the last 6 months.  Again, there was a peak in March, of £2m. 

Conclusion: It’s an old cliche but plainly there was a bit of a budget clear out in March with departments rushing to spend money.  March 2014 spend was £30m – roughly double any other month either side.

– In December 2013, BJSS was the largest supplier, followed by IBM.  Today, BJSS are still number 1, but Methods have moved to number 2, with IBM at 3.

– The Home Office is still the highest spending customer, at £24.7m (nearly double their spend as of December).  MoJ are second at £16.7m with Cabinet Office third at £12.5m

– The top 10 customers account for 50% of the spend on the framework. 
The top 20 make up 67%. That’s exactly how it was in December.  More than 100 new customers have been added since December, though, with over 470 customers now listed.

– Some 310 suppliers have won business.   The top 10 have 32% of the market, the top 20 have 47% (that’s a better spread than the customer equivalent metrics)

– Last time, the lowest spending customer was the “Wales Office”, with £375.   We are at a new low now, with “Circle Anglia Limited” spending £63.50 (I wonder if the cost of processing that order was far greater?).

–  Thirteen customers have spent less than £1,000.  Thirty one have spent more than £1m

Conclusion:  Much the same as in December – Adoption of the framework is still spotty, but it is definitely improving.  A greater spread of customers, spending higher amounts of money – though mostly concentrated in Lot 4. 
A few more suppliers have likely seen their business utterly transformed with
this new access to public sector customers.

Overall Conclusion: G-Cloud needs, still, to break away from its reliance on Lot 4 sales.  Scanning through the sales by line item, there are far too many descriptions that say simply “project manager”, “tester”, “IT project manager” etc.  There are even line items (not in Lot 4) that say “expenses – 4gb memory stick” – a whole new meaning to the phrase “cloud storage” perhaps.

Still, there is no other framework in government that gives access to such a wide variety of suppliers (many new to the public sector) and no framework that publishes its information at such a transparent level.  For those two reasons alone, G-Cloud still deserves applause – and, as it grows month on month, I hope that it will only get louder.