Our former federal government (under Harper’s Conservatives) decided to migrate some 1500 government Web sites to a single content management (WCM) system. They chose Adobe’s AEM product and it looks like the project collapsed in failure. As a Canadian citizen, I’m glad the project has not worked out. I can see how such a project could have merit but the software choice was a bad decision in principle and apparently the project planning and management were not undertaken properly. Too bad it didn’t fail sooner to prevent wasting so much money. CBC’s article about the project is here and I’ll point out a few thoughts about why I think the choice of WCM system was bad in principle and raise some questions on project management.
In 2015, I was in a discussion with a group that was bringing consultants on board to work on the migration. Having had previous experience in this domain and with a large AEM migration, the project interested me but I chose not to get involved and so I do not know any details of what happened other than what I’ve read in the CBC article. There are a few things about the project that I always thought were questionable: namely the choice of software and the management of the project. I felt there was an anything-goes sense of desperation in the recruiting, which didn’t make very feel confident about how the project would unfold.
When I read the CBC news article, the following quote raised a flag for me in terms of questioning such a large project’s management.
“A government source with first-hand knowledge of the Canada.ca project, speaking on condition of anonymity, said IT government workers have been told that none of the government’s arm’s-length agencies have been moving their material over to the new site for some time.”
If people haven’t been migrating material for some time, I would want to ask what happened to lose their participation and throw the project off track?
When you have that many sites, involving a large group of people, they’ve got to understand and feel that their concerns are heard, respected, and involved out the outset of the decision-making. There ought to be things motivating them to move toward the project’s goals. Did they participate in selecting the system? Did they receive training and have the necessary support to comprehend the vision of how the new system and associated processes would work? Did they understand and buy in to the goals? Was it going to improve their work experience and outcomes? Were reasonable deadlines established with communication to keep people aware of the project’s movement as they succeeded with their milestones? I’m sure that one could ask a lot more in a post-mortem of the project..
Of more concern to me though is the selection of the most expensive proprietary WCM platform on the market, Adobe AEM, and the choice of hosting Canadian government sites through an American service provider, Amazon. AEM has many good qualities (as well as bad) but the good do not merit its cost.
First, the platform.
Government software selections ought to, by default, consider free/libre and open source software products first and then develop or select proprietary solutions only if no sufficient F/LOSS choice exists.
Some F/OSS advantages include the following. By definition F/LOSS choices result in greater transparency (and potential security) since the code can be audited. F/LOSS can result in large cost-savings due to an absence of licensing fees (though support incurs costs just like it does with proprietary software). In the future, migrating the systems, code, and data for preservation, archival, or shifting formats can be accomplished with more flexibility if the code is open and accessible rather than closed and controlled by a foreign company (which may or may not continue to operate).
F/LOSS does not lock a government into the control of foreign entities and better, it has the potential to involve communities of domestic businesses in support or customization that might otherwise be prevented.
F/LOSS choices align well with government initiatives in that they enable us to also control our own systems, foster our own businesses and communities, and master our own security. In contrast, Adobe AEM is a proprietary product controlled by a foreign company, with no transparency, and no requirement to give back to communities involved with it. Canadians are paying for something that we cannot control and which doesn’t contribute to the Canadian digital ecosystem. It enriches a foreign company.
Second, the hosting.
Government sites hosted by Amazon—a foreign (American) company? Even if the servers sit on Canadian soil, isn’t this a questionable decision? How does that intersect with American laws and American government agencies, which could demand information from Amazon? As I lack detail about this setup, I am certainly not well-informed and I would imagine that all sorts of precautions were taken in this arrangement. Nevertheless, why introduce this kind of a security and privacy risk? Shouldn’t we be critical about putting Canadian government information (not to mention potential personal information) on servers hosted by a company of foreign origin (using foreign proprietary software) when Canadian companies could have provided the same service (or the government could operate these itself)?
Regardless of how the remaining Canadian government sites manage their content, there can be an information architecture, design strategies, and style guidelines developed, which provide usable experiences and good aesthetics for people seeking information from these Web sites. The back-end system can certainly help improve the management, updating, and preservation of the content for the people developing or administering it but it does not necessarily dictate how the user interacts with the public web front. So even with today’s state of affairs on the WCM migration project, Canadian government Web sites can continue to serve their purposes and improve.
Image already added