The Unexamined Project

At Socrates’ trial, he apparently said the words
The unexamined life is not worth living
We suppose, now, that this is why he chose death over exile.  He neither wanted to stay silent, nor leave his home.
When looking at today’s major projects, we might rewrite this as
The unexamined project is not worth delivering
Too many projects live because they are already living.  They continue when the original aims are long forgotten, when the purpose is plainly no longer achievable and when the costs have far exceeded expectations and the benefits have long receded into the past.  Like a town that was built because the railroad came through but continues long after the Interstate is built several miles away. negating the purpose of the town.  The world has changed, but the sense of self-worth and self-importance doesn’t change.
These zombie projects abound.  The only abject failure in government, particularly, is to admit defeat and take the write-off.  Not so much Fear Of Missing Out as Fear of NAO and Fear of PAC.  Never have I seen such terror in people as when there is a debate about whether to stop and take the hit, or find some more money (because, it seems, even in times of austerity, there is always more money).
Whilst working in and around government, I’ve played a leading role in stopping projects, taking the hit (and a material write-off).  There are some, with long memories, who still keep score and ensure that I don’t get near their projects as a result. The three, for those interested, were 
(1) the original Government Gateway – I inherited it when the Inland Revenue stepped in with some funding.  It didn’t work out. Contract terminated.
(2) a programme called True North.  Pretty sure I can say that this was mine from start to finish – from vision and approach through to, literal, execution.  It aimed to give government ownership of a couple of data centres with one or more suppliers managing what was in them, the idea being that if government fell out with the supplier or wanted new ideas and capability, new suppliers could be brought in without needing to move the hardware (this saw new life with the Crown Hosting Service relatively recently).  It, too, didn’t work out.  Contract terminated.
(3) the e-Borders programme.  I joined that about halfway through.  It definitely didn’t work out.  Contract terminated. Much was written about it.
Let’s call them character building.  The scrutiny is tremendous. The flash to bang time, that is, the time from “we really need to do this” to “we’ve done it” is long and filled with lots of briefings, papers and more papers.  I learned huge amounts from each and try, with every new project, to ensure we find new problems to fall over, not the same old same old; and I try, too, to anticipate every one of those new problems and think about what it will take to avoid them before they land.
Generally, once the old project is put to bed, more progress happens than happened before.  New projects launch.  People are re-energised.  Ideas that were held back because there wasn’t capacity get explored.  As long – and this in capital letters really – the lessons are learned.  I don’t mean written down. I mean examined, understood and deeply explored, with necessary corrections made before starting again.
But today we still have plenty of zombie projects – ones that are sucking in people and time, preventing new ideas from surfacing and gaining traction and dragging people down with them.  It feels like we have made little meaningful progress in the last decade.  The lessons, despite endless independent reviews, PACs, NAO reviews etc, aren’t being learned.

Ask the obvious questions then

– Why this project, right now?  Does it still make sense?

– Are the assumptions we made at the beginning still valid?  Are we regularly testing those assumptions?

– Are we making progress versus our original brief or should we re-baseline?

– Is it right to continue?  Or should we reduce scope, deliver early and move to the next thing?

– Should we continue at all?  Are we clear what it would take to make this decision?

– What would make us change our mind?

How Big Is The Hole You’ve Dug?

One guide to project management suggests:

“Start digging a hole and make it so big, there’s no alternative to coming up with the money to fill it in”

Well, not really.  But you’d be hard pressed to say that isn’t how many projects start out – underestimated costs, dramatically overestimated benefits and impossible to deliver timetables.

Another says:

“Ultimately, blind faith is the only kind.”

Mason Cooley wasn’t, I think, a project manager … but you can’t help but wonder how many people running large projects pick this as their strategy.

This fabulous read on Mega Projects (for this purpose, think very large projects, whether IT or infrastructure etc), which includes the first of the two quotes above, by @BentFlyvbjerg, contains the following ten observations

1. Mega Projects are inherently risky due to long planning horizons and complex interfaces.

2. Often projects are led by planners and managers without deep domain experience who keep changing throughout the long project cycles that apply to Mega Projects, leaving leadership weak.

3. Decision-making, planning, and management are typically multi-actor processes involving multiple stakeholders, public and private, with conflicting interests.

4. Technology and designs are often non-standard, leading to “uniqueness bias” amongst planners and managers, who tend to see their projects as singular, which impedes learning.

5. “lock-in” or “capture,” leaving alternatives analysis weak or absent, and leading to escalated commitment in later stages. “Fail fast” does not apply; “fail slow” does .

6. Due to the large sums of money involved, principal-agent problems and rent-seeking behavior are common, as is optimism bias .

7. The project scope or ambition level will typically change significantly over time.

8. Delivery is a high-risk, stochastic activity, with overexposure to so-called “black swans,” i.e., extreme events with massively negative outcomes. Managers tend to ignore this, treating projects as if they exist largely in a deterministic Newtonian world of cause, effect, and control.

9. Statistical evidence shows that such complexity and unplanned events are often unaccounted
for, leaving budget and time contingencies inadequate.

Ending with this hummer

10. As a consequence, misinformation about costs, schedules, benefits, and risks is the norm
throughout project development and decision-making. The result is cost overruns, delays, and benefit shortfalls that undermine project viability during project implementation and operations.

Everyone with me?  Recognise any projects that you’ve worked on?

Read the whole paper.  Totally worth it.

Opinions Are Like Elbows

We all have them, we just don’t get them out and show them to people all the time.  There’s a much ruder version of this aphorism that I usually use, but perhaps best not here.

A piece by John McTernan in yesterday’s FT caught my eye (I won’t link to it – I don’t subscribe and perhaps you don’t either).  It’s actually about Dominic Cummings.  But the bit I liked is more generally relevant:

“Very many people in politics have opinions, hardly any have plans.  Politicians are surrounded by people who are long on views but short on actionable advice.”

And not just politicians or people in politics for sure.

Government Projects As Moonshots

Much has been written, especially in the last few weeks, about the USA’s manned space flights.   The original “moonshot” programme was everything that today’s projects aren’t.  It began, in the late 50s, with the Mercury programme (one rocket, initially empty, then with an animal and finally with a human in earth orbit for about a day), before progressing to Gemini (two astronauts in low earth orbit) and then on to the Apollo programme (which, again, was incremental – unscrewed, crewed, on the moon, space walks, lunar rovers etc).

Roughly in the middle of that was Kennedy’s speech. Note: roughly in the middle.  Put a man on the moon, bring him back alive, do it by the end of the decade.
There was already a sense of the possible – because of Mercury and Gemini and early work was already underway with Apollo – but it’s certain that Kennedy’s speech was the catalyst of more investment and effort, resulting in the original Space Race.  Along the way there were highs and lows – tragedies too.
Each project built on what had been achieved in the previous one.  And each mission within each project learned from what had already been accomplished and added more.  The envelope was pushed a little at a time – building on even earlier risky adventures with beyond the speed of sound flight.

Imagine the outsourced version of the Apollo programme.  The 200,000 page version of the Kennedy speech expressed as tens of thousands of individual requirements, all written at the very start of the Mercury programme.  I don’t even want to think about an outcomes based contract.
For at least the last twenty years, government projects start with the Kennedy speech before there’s any sense of viability:
– We need a new capability to replace Airwave (the network that supports, amongst others, the Police, Ambulances and Border Guards).  Let’s build it on top of the 4G network that doesn’t yet exist, using standards and technologies that are not deployed anywhere, learning the lessons from absolutely no one as there isn’t anyone else in the world trying to do this.
– We really need a single identity number for everyone in the UK, but one that isn’t tied to any other numbers that exist already and doesn’t replace them.  Let’s not consider the privacy implications or the business case, let’s just get on and do it.
– We should replace all of the various benefits offered by central and local government with a single cohesive benefit, changing who gets paid and when and introducing new elements to the claims process.
– We should replace all of our existing ways to pay farmers with a single system that digitises everything from the launch of the new CAP scheme, notwithstanding we don’t know what the policy changes are, that we have existing contracts that will run for decades and that we could put £2.5bn of annual subsidies at risk in a vital part of the rural economy.
We have done comparatively well with digital projects – in most cases, relatively small, discrete bolt-ons to existing services.  But when we tangle with the difficult stuff, things go badly wrong.

When Is A Lighthouse The Eye of Sauron?

People working in government, whether in project delivery, policy formulation or deep in operations, know that, occasionally, a bright light will be shone on whatever they’re doing.  They fervently hope that this is the beam of a lighthouse, knowing that, if they’re right, it will quickly move on and allow them to get on with their work without distraction.

The lighthouse can range from a routine update to the programme board, to a Minister seeking a briefing all the way up to a multi-day review by the Infrastructure and Projects Authority.  Such attention distracts key people in the project team and little is done except for that immediate task for a day or a week … but then work settles back to normal and progress continues to be made.

The trouble with government projects, particularly, is that everyone working on every other project is hoping that the light is shining on your project, and not theirs.  Because with every project, there’s a definitively non-zero (and often much higher) chance that it really isn’t a lighthouse.  It is, instead,  the Eye of Sauron (in its Peter Jackson incarnation), firmly locked onto your location and determined to see deep into your soul and seemingly never move on.

Projects that are in trouble get into a loop.  More scrutiny.  More reports.  More status updates.  More approvals.  More checks.  More reviews.  More drives more.  Soon, the only thing being done is looking more closely at what might be done, how it might be done or who might do it and when, not actually doing any of it.  Late projects get later.  Projects losing scope lose even more scope.  Over-spends become extraordinary over-spends.

And then a failed project spawns other failed projects.  There is little to no consequence, individually or collectively, for failure so people on one failed project move on to the next.  And because lessons are written down but never learned, nothing changes, and the next one fails again.  This is one reason why government, today, is full of discovery and alpha stages – if you never commit to a scope or date, you can’t be on the hook for anything.

The Eye of Sauron is, as you’d expect, a sign of impending doom.  But not always.  Some projects carry on – once a rock is rolling down a hill, you’d be a fool to jump in its way. The sunk cost fallacy is, for some, not a fallacy at all, but an explanation for why more needs to be spent.  If we just spend some more we can fix it; we can’t afford to waste everything we’ve spent so far – think of the increased scrutiny if we say it was all for nothing?  Think of the write-off!  The PAC! The NAO! It’s perverse logic but it’s intrinsic.

And then, of course, there are those projects that somehow slip away from Sauron; they somehow endure its gaze, slipping through the shadows.  Some projects have more than nine lives. Zombie projects, wandering the corridors of non-delivery seemingly forever.

Far off the shadows of Sauron hung; but torn by some gust of wind out of the world, or else moved by some great disquiet within, the mantling clouds swirled, and for a moment drew aside; and then he saw, rising black, blacker and darker than the vast shades amid which it stood, the cruel pinnacles and iron crown of the topmost tower of Barad-dûr. One moment only it stared out, but as from some great window immeasurably high there stabbed northward a flame of red, the flicker of a piercing Eye; and then the shadows were furled again and the terrible vision was removed.

HMRC … you were the future once

No-one should ever read your first draft. Neil Gaiman.

Nearly 20 years ago when I joined the Inland Revenue (years before it became HMRC), the Internet was new in government.  In the IR HQ, there was one PC that had access to the ‘net – via a dial up 28.8k modem if I recall correctly.  Maybe it was 56.6.  You don’t easily forget the noise that such a modem makes as it works its way to a connection.

Not long after, the IR’s email system was shut down for 3 days by a variant of the Melissa virus.  Sometime later, that led us to back some great work by Al Collier, at what became OGC, to deploy MessageLabs anti-virus capability across the whole of the GSI.  Email was never lost again, as far as I know, by anyone so protected.

The Revenue, as they called themselves, had a website.  But that was it.  Indeed, government had a website (, but that, too, was it.  In the weeks and months that followed, thanks to clear, forward thinking leadership and insightful direction from the Perm Sec (Sir Nick Montagu), the CIO (John Yard) and what would now be called the CDO (Barry Glassberg), Self Assessment went online (remember the £10 rebate to encourage you to file online? And the floppy disc with the “app” on it, replaced at the beginning of the next year with a web app built by Ezgov?), PAYE came next, then Corporation Tax and many other services.

We took hits – despite publishing the need for a maintenance window (to take Self Assessment down) for a few hours on a Friday evening, we made it to above the fold news the following day, when SA was down (when there were, maybe, at best, 10,000 users).  We took flak from the Welsh, Mac using vicar who couldn’t file his tax return (we didn’t do Welsh, didn’t support Macs, and vicars, it turns out, have special tax forms that were not in our initial release; this is not a fable, there really was a Welsh, Mac using vicar that wanted to file his tax return online in 2000).

We built and rebuilt and threw some things away that didn’t;t work.  We ran parallel projects in competition to see what would work and to try and ensure that at least one horse would cross the finish line in time,  Before there was agile, this was agile.

As we worked to put PAYE online, the foundation of it was really the GovTalk standard that the Office of the e-Envoy had already put together (I liked to describe GovTalk as the envelope that you put a letter in, along with the format for the address and the writing inside; the content was yours to figure out).  We worked with dozens of both major and minor software vendors – from Rutherford Webb to Sage through to Oracle – to agree the PAYE XML format that would flow through the Government Gateway (which was in-flight at the same time as PAYE) and into the IR’s systems.  It was detailed work, led mostly by the inimitable and irrepressible Phil Stradling, but it established two important baselines – (1) there would be a single front door into online government services, via the Gateway and (2) the format adopted for all messages through that route would be GovTalk compliant.  Phil was quietly responsible for many firsts in the world of e-government.  I suspect we’ve never thanked him sufficiently for the incredible work he did.

There’s no question that the Inland Revenue (and then HMRC), almost entirely because of John and Barry, led e-government from the front – and were,  I suspect, baffled when other departments got credit for doing a tiny fraction of what they were doing.  They took risks in a world where taking risks was frowned upon. They were the first to put real transactions online (SA, PAYE, CT etc – each of which won awards in its time). They provided the initial funding for the Government Gateway (the vision came from a mixture of IR and HMCE thinking with a very large extra dose from Mark Gladwyn at CITU).  They were the first to get meaningful take-up, from both citizens and businesses – with the Carter review, in 2007, HMRC (as they were by then) became the first department to focus on driving 100% take-up of online services (with the aim of achieving that by 2012; my guess is they hit that, or, at least, got closer than anyone else did by then)

Many years later, I find myself in front of my Mac, painfully rekeying VAT/expenses data from the carefully crafted Excel sheets that I put together a decade ago into the cloud accounting package that, to date, I have only used for sending and tracking invoices.

And, at the same time, I find myself wondering just how far we have progressed.  Or, indeed, if we have progressed at all.

As far as I can tell, the Gateway is still there (my login credentials remain the same, but there was talk that, by now, the Gateway would be replaced – indeed, the website that remained unchanged from 2004 when I left it behind until even a few months ago is now apparently hidden away replaced by a front end).

Is the Gateway a dead man walking?  or is it dead?  I hope they gave it a good send off, it served us all well.  Too many awesome people worked on Gateway to mention here; but they know what they achieved, up against the odds.

What used to be a single front end for transactions into government now looks fragmented across dozens of sites.

And what I’m sending to HMRC, from my cloud accounting package (one login), through some bridging software (another login), through the Gateway (yet another login) …

… is a CSV file with the 9 boxes required for the VAT form.

There doesn’t seem to be a GovTalk envelope.

There’s no additional data.

But there is new overhead and new cost.

And yet no obvious benefit … HMRC are getting what they got before … and countless businesses are sending what they sent before, but with more effort.

And, obviously, no Verify … yet if every single company in the UK is going to send their tax returns this way, and as many as 9 million individuals (roughly 50% used to use accountants, perhaps it’s more now) and then 30 million individuals who might want to check their PAYE status … or a few million students who will want to check their student loan (which inevitably ties to PAYE) … this way in is going to become the default, at least for all financial transactions with government (there may be a good case for why NHS has a different way in; I don’t have a particular view).

Clearly there is more underway here and a bigger picture … but it’s not obvious to me that we have advanced at all since achieving the 100% (or near to it) objective perhaps 7 years ago.

Citizen focused?  Joined up?

I’m not sure.  Doesn’t look like it.

Dear Michael … An Open Letter to the New Minister for the Cabinet Office

Dear Michael

There will be a lot of people giving you advice in the coming days, weeks and months.  forgive me for adding to your inbox.

– Commit to holding this portfolio for as long as this government is intact.  I appreciate that this could mean weeks or months, maybe years.  But, politics aside, what this role needs is the stability and consistency it had from 2010-2015.  I appreciate Brexit will take up a lot of time … but the rest of the show needs to go on.

– Be Francis Maude.  Not him exactly, but like him.  Treat this as the last job you will have before you become Lord Gove.  Use it to reshape the whole of government and how it does business.  Be the enforcer. Move fast, stop things, confront the ostriches (of which there are many).

– Seize hold of GDS and shred it.  It’s served its purpose.  Most of it can go.

– Not all of it though.  Find the few people who are get shit done and keep them doing that shit.  Inject new people from the rest of government alongside them and have them teach those people how to get shit done.  Have them re-work and notify, the two clear success stories, and take them to the next level.   Call them the GSD team.  GSD?  Get Shit Done.

– Scatter most of the rest of the people across government.  Don’t let them take any post it notes back with them and cancel any framework that allows their purchase.

– Stop Verify.  No, I mean it.  Really stop it.  Don’t pretend you can give it to the nebulous private sector.  Take the hit.

– Find another few people who can see round corners and know how to plan for what’s coming next.  Make them the nucleus of a new team – Policy for Digital Government.   Not Government Digital Policy obviously.   Have these folks work out a plan for where everyone needs to get to and work out how the rest of government can intercept that future by making small and regular changes to what they are doing now to line up with the new course.

– Be clear that the job isn’t about websites and tinkering with Discovery and Alpha (the endless iteration of never getting to live).  It’s about making a difference with the big things. It’s about breaking down the walls of government.  The big things are scary, transactional services that, often, are still on systems built 20, 30 or even 40+ years ago.  Those are what we need to confront.  In, as we used to say many years ago, a joined up, citizen focused way.

Because Digital Government isn’t about Government.  It’s about the citizen.  And it’s time we got back there.