Why are your journey times sometimes different to Google's?
Although we regularly benchmark our models against other providers including Google to ensure that the results are comparable, there are a number of important reasons why they differ.
When it comes to public transport data, in benchmarking exercises we often find that we have data from more public transport agencies - therefore Google may return a slower public transport route as it lacks the data of a faster alternative
- Parameters such as maximum walking time, or maximum number of changes, will likely differ between providers (although these are completely configurable when using the TravelTime API)
- If there is no route available at the arrival/departure time selected, Google will try to return an alternative (such as a journey that is actually the night before), whereas we will simply return an unreachable result
- Our driving model does not take into account live traffic conditions, but instead is modelled around expected congestion at a particular time of the day
- When a driving journey starts away from the road network, we apply a time penalty based on how long it would take to walk to the nearest road, based on an average walking speed - Google will often snap to the network without a time penalty