Bring lesser-known places into view
Guide aims to help visitors move beyond overcrowded routes and discover villages, small towns, rural heritage and the paths less travelled.
Explore without overcrowding
Sustainable tourism is not only about travelling less. It also means spreading visits more evenly, choosing low-impact mobility and designing digital tools that do only what is needed.
See our commitmentsOur direction
A good route does more than move people. It can redistribute attention between places, make time outdoors more meaningful and avoid some repeated digital requests.
Four commitments, one direction
Guide aims to help visitors move beyond overcrowded routes and discover villages, small towns, rural heritage and the paths less travelled.
Our routes are designed first for walking and cycling. Travel time becomes discovery time, without a fixed pace or unnecessary motorized journeys.
We favour lightweight pages and statically served content where possible, avoiding repeated computation and unnecessary data transfers.
We publish orders of magnitude with their assumptions and limitations. The next step is to measure a real Guide session, with and without caching, to refine the model.
Our impact study
What is the carbon impact of a tourist looking for cultural information about a place? Our model compares five representative three-minute smartphone sessions using shared assumptions for networks, data centres and French electricity.
These are modelled orders of magnitude, not direct measurements or complete life-cycle assessments.
What we compare
Each scenario addresses the same general need, but does not necessarily provide the same amount or type of information. The five results are alternatives and must not be added together.
Three minutes on a smartphone to obtain cultural information about a place.
The service, data centres, network transfer, phone electricity and charging losses.
Phone manufacturing, content creation, development, Guide installation and travel.
The Guide scenario assumes 0.6 MB transferred, optimised images and a cached map. A real-world measurement will replace this model.
Methodological transparency
The figures are generated by reproducible calculations using low, central and high scenarios. Open each section to review the formulas, factors, breakdown and references used.
For Guide and text-based journeys, the calculation adds the service-specific impact, network transfer, data centres and phone electricity.
impact = service + (données × réseau) + (données × centre de données) + téléphonePhone = power × duration ÷ charging efficiency × carbon intensity. Video uses an hourly Arcom–Arcep–ADEME allocation instead of a simple per-gigabyte factor.
| Journey | Service and data centres | Network | Phone | Total |
|---|---|---|---|---|
| Guide | 0,0163 g | 0,0300 g | 0,0020 g | 0,0483 gCO₂e |
| Direct website | 0,0490 g | 0,0900 g | 0,0016 g | 0,1406 gCO₂e |
| AI assistant | 0,1514 g | 0,0025 g | 0,0016 g | 0,1555 gCO₂e |
| Google + page | 0,2126 g | 0,1150 g | 0,0016 g | 0,3292 gCO₂e |
| YouTube 720p | 0,0830 g | 1,4000 g | 0,0020 g | 1,4850 gCO₂e |
For AI and Google, the first column includes a query-specific computation factor. For Guide and the direct website, it is estimated from transferred data.
The simulation fixes 1,000 tourists, five places and therefore 5,000 views. It multiplies the per-view impact from the previous model; actual usage proportions are not applied.
impact ville = impact par consultation × 5 lieux × 1 000 touristes0,0483 gCO₂e × 5 × 1 000 = 241,5 gCO₂e = 0,2415 kgCO₂e
Epoch AI · énergie d’une requête textuelleEstimation centrale de l’énergie d’une réponse IA standard.
02Google · impact d’une requête GeminiBorne de comparaison pour l’ordre de grandeur d’une requête IA.
03Google · énergie d’une rechercheFacteur de calcul d’une recherche, publié en 2009 et associé à une forte incertitude.
04HTTP Archive · Web Almanac 2025Poids bas, central et haut d’une page web mobile.
05ADEME · transmission de donnéesFacteur central de 50 gCO₂e/Go pour le réseau mobile.
06Sustainable Web Design · modèle v4Facteur de 27,2 gCO₂e/Go pour les centres de données associés au transfert.
07UC San Diego · puissance des smartphonesOrdre de grandeur de la puissance du téléphone, notamment pour la vidéo 720p.
08RTE · Bilan électrique 2025Intensité carbone centrale de 29 gCO₂e/kWh pour l’électricité française.
09Arcom, Arcep et ADEME · usages audiovisuelsAllocation horaire du réseau et des centres de données pour la vidéo en ligne.
Google caveat The search factor is based on an estimate published by Google in 2009. It is retained as an order of magnitude with a wide uncertainty range, not as a current measurement of the service.
At the scale of one day
Simulation: each tourist views five places for three minutes and uses the same solution exclusively. This scenario compares journeys at equal volume; it does not represent actual usage patterns.
Central scenario · 15 minutes of digital use per tourist
Key takeawayIn the central scenario, the usage carbon impact of one Guide view is estimated at 0.0483 gCO₂e, the lowest of the five journeys studied. This result depends on the assumptions used and does not prove that Guide is carbon-neutral or inherently ‘green’.
A progressive approach
Define a realistic measurement scope.
Measure key routes, content and uses.
Publish results, limitations and sources.
Improve the product based on what we learn.
Build with destinations
Let’s share lessons, useful data and ideas that work in the real world.
Talk with Guide