When you punch in the final destination into a sat nav device, most will then attempt to find different routes (fastest time, shortest distance, avoiding motorways, avoiding tolls) and then give you an estimate travel time and arrival time. It does this by looking at the average speeds between A to B as well as using traffic forecasts, incident reports and other bits of information it can find.
This is when your sat nav will tell you the time it thinks you'll get there and how long the journey should be.
Different devices get their data from different sources, this is one of the biggest contributors to wildly different performance levels.
In the olden days of satellite navigation, many providers received their data via the traffic message channel. It's a bit old school now as most sat navs these days use crowdsourcing as a way to receive traffic information to be able to estimate journeys. TMC use floating vehicle data from over 160,000 fleet vehicles which is a large number overall but most of these will unlikely be on the exact journey you want to take at any given time. This data is then merged with incident data from journalists.
Because TMC can have a delay and don't overly use crowdsourcing it could often be slow and inaccurate but still serviceable.
One of the biggest contributors to satellite navigation data in the UK is the Ordnance Survey. They don't just provide traffic information but also the maps themselves. The OS does not manufacture any sat navs and make their data available to anyone who wants it.
The most common method of receiving traffic and map data is through crowdsourcing. This means that your device is receiving live data from other users up and down the country giving you real-time and accurate information on what's ahead on a journey. As more and more people are using free to use map systems such as Google Maps and Apple Maps it means more data which can be used for accurate estimated travel times.
A sat nav will have a couple of different methods to predict what speed you'll be going and this has a big impact on the estimated travel time of your journey.
Based on how your sat nav gets its data, your estimated speed will be based on those factors.
Road closures and diversions are the same stories as for how sat navs measure speed. It all depends on where the data comes from. If you've got an old sat nav, especially those built in ones that need to be manually updated, then you could be in a world of pain as it struggles to keep up. If you have one that hasn't had its maps updated for a while then you might be in even more trouble as roads which are no longer there will be in the maps and newly built roads will be completely absent. This has happened to me personally as my sat nav still hasn't got the new Runcorn bridge and keeps trying to get me to drive over the river!
You could have probably guessed by now that the answer to this question is "it depends on the sat nav".
An old sat nav that doesn't use crowdsourcing and has outdated maps will have a very inaccurate estimated travel time and will likely disappoint you. I myself have an inbuilt sat nav on both a Ford Fiesta and Volkswagen Golf and they don't seem to be that accurate and often don't predict congestion very well.
A modern sat nav or maps app such as Google Maps or Apple Maps won't be exact as real life can never quite be predicted but it should be incredibly accurate as you will be making use of data from those who are on the same journey or nearby. Whenever I have used one of these apps in the past I have been very impressed with their accuracy.
We put this to the test with two journeys one from a WA10 postcode and M28 postcode to our offices at WA16 and then the same tests going back home and we used both the on-board sat nav and Google maps in both.
Google maps estimate: 46 mins
On board sat nav estimate (Ford Fiesta): 40 mins
Actual travel time: 45 Mins
Google maps estimate: 55 mins
On board sat nav estimate (Ford Fiesta) - 45 mins
Actual travel time: 55 mins
Google maps estimate: 50 mins
On board sat nav estimate (Volkswagen Golf): 45 mins
Actual travel time: 1 hour
Google estimate: 45 mins
On board sat nav (Volkswagen Golf): 40 mins
Actual travel time: 45 mins
By no means is this is a proper scientific study or a rigorous test. But, it's still quite interesting to see that Google maps were the most accurate in all 4 tests. Both tests, going both ways, encounters rush hour traffic and we feel it's that which causes the on-board sat nav to struggle a bit. Unless it's a crystal clear day with no traffic whatsoever most estimates are going to get it wrong and at the end of the day, you should always give yourself plenty of journey time and never rely on technology for your punctuality!