You
can’t push an elephant through a needle
hole,
irrespective of how hard you try it…
If
you don’t know what you want,
You
will get what you asked for…
A few days ago, a Dutch parliamentary committee
presented its conclusions, with respect to the investigation that the committee
performed, into governmental ICT projects gone awry during the last decade. As
some of you might know, this topic has my undivided attention, as it is
connected with my daily work as a very experienced ICT testing consultant.
The conclusions of the committee, chaired by a
liberal-conservative MP from the VVD party,Ton Elias, never transcended from
being a list of open doors and advices, which were often simply ‘too obvious’.
The report seemed to fight the symptoms, instead of the disease.
Besides that, the committee has been too eager, at
occasions, to point the finger at the ICT suppliers, instead of looking at the
government itself as a roused and impatient principal, which offers unclear
requirements, desires and demands as a starting point for large, open-ended ICT
projects.
The report is yet another testimony of generally clueless
politicians, who are genuinely shocked by the number of governmental
ICT-projects gone awry and the financial damage that these failed projects
caused. And now these clueless politicians are desperately looking for easy
solutions, which they won’t find as they seemingly fail to look at the root
cause of most problems.
Does this sound too harsh?
Well, imagine that your teenage son approaches you and
says: “Dad, yesterday I went to a
discotheque with my girlfriend and I spent there between €100 and €500. Yet, I
don’t know how much money I spent exactly, because I didn’t count it, when I
took your wallet with me”.
Would you be a happy camper, when your son would do
that?!
Since
no comprehensive report on the national public finances has been drawn up after
1995, nobody knows how much money the Dutch public sector is really spending on
ICT. Nor how much is being wasted on failed projects. A conservative estimate
based on information from a variety of experts suggests that the figure may be
anywhere between €1 billion and €5 billion euros each year. Whatever the true
amount, the Committee believes it to be unacceptably high.
I cannot understand why the committee got away with
this amount of sloppiness, in its predictions regarding the bandwidth of the
ICT losses. Besides that, the main recommendation of the report is filled with
an incomprehensible self-confidence, clearly not based on substantive prowess
of the committee with the matter on hand:
Nevertheless,
the committee feels that a few robust organizational measures – provided they
are implemented consistently and coherently – will be sufficient to prevent a
repeat of a large proportion of the problems identified. The Committee’s
recommendations are closely interrelated and should be viewed as a total
package of measures for the Cabinet to adopt.
Much
will be gained simply by involving not only ICT specialists in government
projects of this kind, but also users and those responsible for monitoring
government spending. On this point the Committee is strongly critical of the
House of Representatives itself, as up until now it has not made sufficient
efforts to scrutinize public expenditure in this area.
Do the creators of the report really think that these
measures will help them to prevent the central and local governments from
suffering big losses on future ICT projects?! Really??!!
The suppliers in the ICT industry get taunted by the
following, slightly villainous statement:
On
the rare occasions when a supplier does point out avoidable problems, these
warnings are all too often not taken seriously.
I have spent 16 years, working as an ICT consultant and
specialist at functional testing.
I dare to say that most ICT suppliers – small as well
as large ones – are honest, hardworking and respectable firms, which try very
hard to find the best solutions for their customers.
Consequently, most companies are indeed willing to
point out possible problems ahead when they see and grasp them, as these might
and probably will backfire at them as suppliers, when a government project
fails and the ‘blame game’ starts.
Nevertheless, these ICT companies are in it for the
money. So simple is that… They need their staff to make sufficient billable
hours and thus earn money for the company; this is their raison d’etre and you
can’t blame them for doing so, mostly with the best intentions for the customer.
So – in the end – when their customers/principals (i.e.
the central and local governments and/or governmental bodies) want to execute large
projects, in spite of many foreseeable problems ahead and an uncertain outcome,
most companies do start developing the project!
When their customers demand certain services through a
tight time-schedule, in spite of warnings regarding impossible deadlines and
opaque specifications and requirements, they try to deliver such services. When
these companies would not do that, their competition will and consequently they
will lose the assignment to their competitors.
“You
demand. We deliver…! No questions asked”.
You could blame these companies for doing such, but
that would not be fair, in my opinion.
Some conclusions of the report, however, are spot on:
On the one hand there is unbridled enthusiasm for ICT, with
proponents
viewing it as the solution to every problem. On the other hand the
House of Representatives regularly demands policy measures, without realising
that ICT is almost always needed in order to implement them.
The
minister in question promises delivery, without first checking whether the
measures
required are technically possible in ICT terms.
Even
when they know that the promises being made to Parliament cannot be fulfilled,
officials do not challenge the political leadership enough. When they do voice
their concerns, the necessary information does not reach the top political
level. This results in ministries issuing tenders for, as they were called
during the hearings, “cars without steering wheels”: something, which, by its
very nature, cannot work.
The first red and bold paragraph shows
exactly why so many ICT projects at the government are doomed to fail.
Unfounded enthusiasm for ICT and lacking insights in the possibilities and impossibilities
of it, in combination with unclear and politicized requirements and
unachievable deadlines, predict havoc on any ICT project.
Here is the summary, containing the ‘most important’
conclusions:
1. Establish a temporary ICT authority to act as a project
gatekeeper: the BIT (Bureau for ICT assessments).
3.
The House should increase its ICT awareness, for example by including ICT in
the induction programme for new MPs and maintaining regular contact with the
BIT.
5.
From now on the Cabinet should explicitly consider ICT in its decision-making
processes, in a structural manner, weighing up the possible consequences and
risks of its decisions from that perspective.
6.
The government should introduce more central management of its ICT policy, among
other things by appointing a single minister responsible for policy concerning
ICT project
management.
7.
The national government CIO should be given more authority, including decisive
powers over the implementation of general ICT policy.
8.
The cost savings and societal benefits of ICT policy in general must be made
visible.
10. Continue the centralization of ICT procurement and
government-wide ICT facilities.
11.
Clearly define roles and responsibilities within all government ICT projects,
including those at executive agencies. A single minister should always be
responsible for any ICT project in which there is a major public interest.
17.
Introduce a compulsory initial test for projects worth more than €5 million
with a significant ICT component.
21.
Make ICT a permanent component in the internal training for all civil servants.
22.
The ministerial CIO should ensure that roles and responsibilities are clearly
understood.
24.
Those implementing the project and every layer of management should provide
their senior officials and managers with realistic information concerning its
progress.
28.
From now on, a supplier’s past performance should be taken into account when
evaluating tenders.
32. The government should avoid additional work and the use of
hourly rates, turning any perverse incentives into positive incentives.
This (reduced) list of obligatory, open door-statements
actually contained the most significant and important conclusions of the report
– in my humble, but nevertheless experienced opinion.
This means that the other conclusions were even more
trivial and neglectable. I will focus, however, on the red and bold conclusions, as these
stood out even more negatively than the others and miss the point by
lightyears.
Conclusion 1.
This BIT bureau, as an official, governmental
gatekeeper for ICT projects, will probably develop into a Soviet style ‘Polit
Bureau’, as it gets much power. Too much power, according to me.
Their ability to endorse one project or halt another
can lead to stagnation at one hand (i.e. a bureau “saying nyet to any project”)
and corruption at the other hand, especially when such a bureau turns into a ‘private
kingdom’, without sufficient political control and transparency. Their final verdict
could decide on the sheer future of governmental ICT projects and that opens
all options for bribery, corruption and nepotism.
Besides that, it ignores one of the most important
conclusions from the report, that almost any government decision requires a
change in the ICT infrastructure. Just saying ‘nyet’ to a project does not
solve the political needs or fulfills the requirements.
Conclusion 10.
Central procurement
is something that has created havoc in the ICT industry and it bears the same
problems as the BIT bureau.
First, it makes it virtually impossible for small,
‘lean and mean’ ICT companies, with eager employees and smart solutions, to
make a tender offer on government projects in any kind. The only companies that
could offer to such government tenders are the same ‘big ten’ companies that
“messed up” so many other government projects.
Second, such an extremely powerful procurement
organisation has the power to make and break ICT suppliers with a simple strike
of a pen. This will have a downward effect on the price of ICT services and
hourly fees. Services and fees, which already received blow after blow in the
past six years.
And – to the contrary – when the demand for ICT
services would increase dramatically in the future, there are only few ICT
suppliers that are able to meet the demands of this procurement bureau. Hence,
the costs of government projects would soar again, as the procurement agency
has only a few suppliers to choose from: suppliers that could also deploy their
employees at other companies.
Conclusion 32
This statement is so naive.
As principals, central and local governments and
semi-governmental organizations are notoriously capricious and restless and – consequently
–unreliable.
Many central and local politicians have – next to an
extremely short attention span – an undying tendency to respond to any event
with more and different legislation. Too often, short term political gains and
short-sighted responses to crisis situations are favored at the expense of developing
a consistent long-term vision on political and governmental subjects.
Mixed with inconsiderate policies and the eternal and
ubiquitous desire for additional austerity measures at the central government,
this makes that tax legislation, healthcare regulations, social security arrangements
and other ICT-intensive governmental areas virtually change with the speed of
light.
The whole ICT-infrastructure at the local and central
governments and the governmental bodies should change accordingly to keep up
with all the political changes.
However, this is often impossible, as a consequence of
the often very old and non-scalable hardware and the obsolete and static structure
of the software, which was often developed by people who are currently retired
or have even died from old age(!).
Much governmental software exists from a (litterally) decades-old
foundation, extended with numerous layers, caused by dozens of changes. This
turned the whole software complex in an incomprehensible and unstable entity.
This particular circumstance makes it virtually
impossible for ICT suppliers to offer fixed price projects, without running a
substantial risk at enduring financial disasters, due to drawbacks and
unforeseen circumstances .
In the past, too many ICT suppliers saw fixed contract
projects turn into a financial cataclysm for their company, due to reality
biting at one time and allegedly clear requirements being suddenly unclear,
flawed or obsolete after all, due to a changing environment. Penalty
clauses, and other financial sticks in
hands of the principals did the rest to scare these companies sh*tless.
The only way in which a fixed price contract can be safely
executed by suppliers of software and hardware services, is by having extensively
elaborated, ironclad contracts of 150 pages, which tell to a T what the
supplier will and will not do during the project. No exceptions and no
flexibility possible!
On top of that, they charge large (read: exuberant)
premiums for unforeseen circumstances to the customers, amalgamated in the
project price, which.
I know by heart that the results of such ironclad
contracts are always disappointing and almost never bring the desired software
and hardware infrastructure. It is simply much too hard to catch reality in a
contract.
This is the main reason that most ICT suppliers only
offer labour at an hourly rate and at ‘best effort’ basis: an obligation to deliver certain
efforts, instead of an obligation to deliver fixed results (!). Many ICT
companies can simply not offer more than their best effort to their government
customers.
Because that is the simple elephant in the room: ICT
projects often fail due to unclear and changing requirements, in combination
with an obsolete existing infrastructure.
And the stricter the deadline is for such projects, the
bigger is the chance that a project fails totally or – at best – yields only
half of the desired functionality, against twice or triple the original contract
price. Very little in the report of the Elias Committee will prevent that from
happening in the future.
The government itself should focus on developing a long-term
ICT strategy and it should reserve vast budgets for replacing obsolete software
and hardware at governmental bodies with modern, better scalable solutions.
This replacement should happen, without combining it with numerous software
changes.
And most important: the leading politicians in The
Netherlands should abolish their restless and short-winded vision on things,
leading to legislation-for-the-sake-of-legislation.
Instead, they should offer the central and local
government and governmental bodies a chance to reattach to the reality as-is
with improved software and hardware, instead of forcing these parties to pursue
their own tail with endless rounds of new legislation and governmental
arrangements.
That lesson is so dearly missed in the report of the
Elias Committee.