OMG I actually bid on ahem a remarkably similar project at my old job! It was the worst tender I've ever done. We were shortlisted, and they asked for massive amount of detail in the proposed architecture as part of the bid. Early on, they stated that they'd be open to looking at a solution that used headless Drupal, which is what several shortlisted agencies were proposing including us. I spent a month on the bid, which is a lot more than I usually would have because I love the place, and had a really good architecture in mind. I was really pleased with the architecture that I produced, which had quite complex requirements with all the integrations they needed. The main site would have a Gatsby front end, and we'd be saving them tens of thousands on their hosting bills. So, a week before we're due to go up to present to them, they contact all of the bidders saying that they've decided that they won't consider headless after all, and would we mind pitching vanilla Drupal instead. A month of work down the pan. My colleagues cobbled something together over the weekend. Our bid was about 50% above the winner, and it was generally a miserable experience.

So, happy ending though: a year later we sold a site with an architecture very similar to my original proposal, to a client that doesn't mess us around, with a realistic budget and the project went really well. A few months after that I left the company, and I now work at Gatsby, so I can build open source Gatsby stuff all day and don't need to deal with fickle public sector clients ever again.

Almost 200K USD. That seems like quite a steal, almost a reverse heist. Amazing that the museum was able to work that out. I wonder about the size of the team, how tight the spec was, etc. It's been fun to search around...a very impressive collection for sure.

Near the bottom of the UK bid site is public Q&A with some interesting technical tidbits, such as integrating with a Magento web shop, etc.

Yeah, that's pretty interesting for sure:

> A previous project to replatform and rebuild the site that was aborted in early 2017, has yielded a significant amount of development work on the design, creating multiple assets including content models and front-end design that we would like to repurpose.

> Earlier this year we conducted a new discovery phase to re-establish what we were setting out to achieve and this has culminated in a new vision and objectives for the website, high level requirements, user stories and technical options and recommendations.

Phew. Reading that and trying to imagine the circumstances, I feel for those previous-platform people too.

Maybe there are some off the shelf elements to this. The search engine and back end integration is standard, so they only had to write the front end code.