Army's £113M recruitment website 'was 52 months late'

It was for a recruitment contract in totality for a decade (total £622m) including all efforts to recruit for the UK army, including everything that involves... of which a small part was the website but it was supposed to be integrated into everything and included running the old legacy system in parallel for a long while..

The details are in the report https://www.nao.org.uk/wp-content/uploads/2018/12/Investigat...

But the headline is not the story.

Is this the website?

The HTML is full of comments like:

  <!-- This was a temporary Single Sign on fix - I think it's no longer needed?  SVL -->

  
  <!-- This is being done since bundling is currently breaking css files!!!-->

https://motherfuckingwebsite.com/ is a website. Facebook is also a website. Spending £113m to build Motherfucking Website is insane; spending £113m to build Facebook is a bargain.

As other commenters here noted, we can easily imagine extremely complicated requirements. The backend requirements for a "recruitment website" for the military might include all sorts of background checks, including pulling criminal records and automatically trawling social media for content posted by candidates and doing NLP analysis on it to detect warning signs of extremism. There might be some sophisticated proprietary analytics portal for monitoring the effectiveness of marketing campaigns. And while there might be a good writeup elsewhere on the web on what exactly this project consisted of, and I'd welcome relevant links, based purely on the content of the article we just don't know.

Calling it a "website" without giving any hint of the requirements makes it sound like the army paid £113m for the digital equivalent of a recruitment brochure, yet I seriously doubt that's the real story here. I'd expect HN commenters to be more skeptical of these kinds of portrayals, rather than uncritically accepting the media narrative.

First and foremost the lesson should be that the army should have its own dedicated IT department.

Also, any company peddling 'digital transformation' should be a red flag to start with.

There's your problem, you gave it to Capita.

The whole PFI model relies on the pipedream of being able to just hand a piece of paper over from the Minister's desk detailing what they want, as if they were sending the intern out to buy milk, and magically getting the right thing back. In practice it requires all sorts of complexity and control; the project management is actually harder than just straightforwardly doing it inhouse, agility is impossible, and the opportunities for cost inflation are huge.

There is no real penalty for failure, and no real competitive market since the tendering process is so complex and expensive, and in any case the relevant small number of firms already have the Minister on speed-dial. Besides, the outsourcing inevitably ends up "too big to fail", in that when a bankruptcy does happen like Carillion the state has to take over the cost of sorting it out anyway.

> "Capita has consistently missed the Army's recruitment targets, with the total shortfall ranging from 21% to 45%, the NAO said."

Not to get off topic but...Did Capita miss the targets, or have the proles simply become wise to the ways of the powerful ruling elites? A pitch of "serve your country" certainly doesn't have the truth / appeal that it once had.

> "The Commons Defence Committee was told in October that the Army currently has 77,000 fully trained troops, compared with a target of 82,500."

Now drilling down a bit...5,500 short? That's not even 10%. The raw number looks large-ish, but as a percentage (while short) doesn't feel unreasonable (in a real world sense).

Classic :D

The Army initially attempted to develop the IT infrastructure

The Department approved the Programme on the basis that the Army would use its existing IT infrastructure to host the online system. In 2012, the Department was therefore responsible for developing the IT infrastructure to achieve this. Capita was responsible for developing the new online system. As we reported in 2014, the Department encountered problems with the delivery of the IT infrastructure, and deadlines were repeatedly missed.

The main problems were that the Department: • Had to manage a relationship between two providers who had no contractual relationship with each other; • Underestimated the complexity of what it was trying to achieve and, consequently, the resources required to manage the risks; and • Had an Army project management team that was inexperienced and under-resourced.

The Army originally intended to introduce the new online recruitment system by July 2013. In December 2013, it accepted that its approach had not worked and passed responsibility for implementing the whole approach – the hosting infrastructure and the online recruitment system – to Capita. The Department’s Accounting Officer accepted that the failure to manage the dependencies between the Atlas and Capita contracts was ‘unacceptable’.

Capita took responsibility for developing the online system.

Capita initially intended to use an ‘off-the-shelf’ commercial system but underestimated the complexity of the Army’s requirements. It found the level of customisation needed to support the three Armed Services’ recruitment processes meant that an off-the-shelf system was not a viable technical solution. In February 2015, it therefore began to develop a new bespoke system.

Capita then encountered problems developing the new application and missed launch deadlines in July 2015, April 2016 and April 2017. The main problems involved the scale of the three Armed Services’ processes and requirements, and the number of interfaces with other IT systems. For example, the Armed Forces have over 250 job roles, each with its own eligibility criteria and rules, which means a new Army candidate can have 27 possible pathways through the recruitment process. Capita told us that these requirements made system development more complex than standard online recruitment systems.

The next year was spent building a bespoke app using a RAD (Rapid Application Development) tool. Failed and abandoned after 12mths. How it was ever chosen defies belief.

The next 18mths was spent building a bespoke app with a different RAD tool (K2) supposedly that would enable catching up on the previous lost 2yrs by cutting lots of corners and doing 3 yrs work in 9mths. It was the wrong tool for the job and was never going to work. Management were told many times, and why, but they would just ignore it and tell us to carry on. They hired every K2 developer available on the market in the UK, but they seemed to be only used to building small apps in one or two person teams, whereas many of us were used to working in large teams building large complex systems. The K2 devs were considered the 'gurus' and everytime we raised the issues management would go and ask the K2 guys, and of course, they would dismiss our issues (often quite arrogantly) to protect their product. After 18mths it became clear, we were proven right and K2 was abandoned.

In all this time, the entire management was completely replaced 3 times. None seemed to have any experience in how to build a software application or website. They would make the most astounding statements that concerned many of us that they had no idea how to build not just a complex system like this, but any software application at all. We were told we werent allowed to talk to the project 'architects', and we later found out they were told not to talk to us (an instant recipe for failure by itself). When we did get to talk to them (unofficially) we soon realised why, they were also clueless. Next they tried bringing in external management, but they were just as clueless and didnt seem to know what to do either.

Of course the application was going to be large and complex, but was quite achievable if people who knew what they were doing were in charge. In my opinion the root problem is the same as in many other companies that develop software systems (and often the root cause of the many data hacks we see on the news): management who have no technical background choose the wrong tools for the job, hire incompetent numpties who dont understand software technology to do the work, and constantly try and cut corners you shouldn't cut to save time and/or money.

Even when a project is behind you should be able to say how much you have actually have completed, 10%, 20%, something (especially with 400+ people on it), but it always seemed nothing was ever considered 'complete'. We often said 3 x 10 person teams who knew what they were doing, with the correct architecture and development tools, could have built a working system in 12-18mths.

Reading the report, I find it repulsive the financial penalties have been reduced to keep Capita 'motivated'. Our armed forces deserve better. With all the cost cutting they endure, they need to recruit quickly and efficiently and fill 100% of positions they have to maintain an effective defence of our country.