The night Le Monde broke two records
Nick Peart
How Le Monde monitored their biggest night ever with Tsuga, and why AWS wanted to tell the story too.
A few days after the first round of the French municipal elections, Paul Laleu, CTO and Comex member at Le Monde, posted on LinkedIn to reflect on what his team had just been through. The numbers he shared were striking: 21 million visits and 70 million page views in a single day, with an average page response time of 58.98 milliseconds. For a media organisation, election night is the ultimate stress test of the entire technical stack, and Le Monde had just set records on traffic and response time in the same evening.
The post included a screenshot of the observability dashboard his team had been watching throughout the night. It was our dashboard. He did not mention Tsuga by name, and he did not need to. He was writing for his own network, celebrating his own team, and the dashboard happened to be part of the picture he chose to share. That kind of moment is not something you can manufacture. We liked the post, offered a quiet digital pat on the back, and left it at that.
Read Paul’s original post (in French) here: Paul Laleu on LinkedIn. It’s in French, but the context is universal: an unprompted CTO, a record-breaking night, and a dashboard he was proud enough to put in front of his entire network.
What three months looked like at Le Monde
That election night moment was not a one-off. Le Monde had deployed Tsuga several months earlier under Paul’s leadership, and by the time the votes were being counted the platform had already changed how the engineering organisation worked. Within three months of going live, the results were measurable across every dimension that matters:
50% increase in observability adoption across development teams
3x expansion in the number of monitored services and applications
30% reduction in mean time to detect incidents
50% reduction in mean time to resolve incidents
The adoption figure is the one we keep coming back to. A 50% increase in the number of developers actively using observability tooling is not a performance metric in the conventional sense; it is evidence of a cultural shift inside the engineering organisation, and it is exactly the outcome we set out to produce.
And then AWS wanted to tell the story too
The results were strong enough that AWS approached Tsuga to document the Le Monde deployment as a formal case study. We are a Paris-based team of fewer than 50 people, founded in 2024, and that kind of recognition from a hyperscaler is not something we take for granted. The case study covers the architecture, the data sovereignty model, and the business outcomes in full, and it is now live on the AWS website.
Read the full case study: Tsuga takes BYOC observability to the next level, published on aws.amazon.com.
Taken together, these two pieces of public evidence tell a story we could not have scripted. A CTO sharing his team’s proudest moment with his own network, a dashboard visible in that moment, and a hyperscaler independently deciding the deployment was worth documenting. We do not have the marketing budget of the incumbents in this space, but we have an architecture that works, a reference customer that is one of the most demanding media organisations in Europe, and proof that came from somewhere other than us.
Links: Paul Laleu’s LinkedIn post | AWS Case Study