Every network is pushing 4G at the moment, even the MVNOs – those networks that use the likes of O2, Vodafone, Three and EE as the back-end infrastructure but supply their own service on top.
giffgaff is probably one of the most well-known, mainly due to their community support system, low prices and easy-to-understand offering. Although you can now buy phones now, they still offer the Goodybag plans. Grab a SIM, select what monthly plan you’d like and off you go. You can chop and change as you want and it’ll roll over monthly if you don’t want to faff around with top-ups.
There’s two sets of Goodybags. The first set is 3G only, costing £7.50, £12 and £18. You’ll get unlimited texts on all and 200, 500 or 2000 minutes talk-time respectively. Likewise, the internet allowance increases, from 250MB, to 3GB, to unlimited on the top Goodybag.
However, you’re on 3G only for the first set of Goodybags. Now, with O2 (the parent network that giffgaff uses) rolling out 4G, you can switch to a set of 4G plans.
These go from £12, £15 or £18 per month. Again, there’s unlimited texts on all of them. The £12 and £15 4G Goodybags get you 500 minutes of talk time, with the top £18 4G plan getting you 1000 minutes. As for internet, you get 1GB on the 4G £12 plan, 3GB on the 4G £15 month offering and 5GB of 4G data on the £18 plan.
I’ve used the £12 giffgaff 3G plan for a while now, but I recently tested out a 4G giffgaff SIM alongside it. Trouble is, the O2 4G roll-out doesn’t come anywhere near my home, my work, or anywhere in between. My only way to really check this out would be to drive 10 miles south.
So, due to a rather unforeseen hospital visit, I ended up 10 miles south. It’s a shame that O2 is still (as I type) in the early stages of the 4G roll-out as I simply can’t use it in all the places I am regularly. That said, if you’re lucky enough to have 4G (check this map) then you can give this a spin.
As a rough rule of thumb, standard 4G should be about 5 times faster than standard 3G. Trouble is, there’s so many factors that can alter the speed you receive. It depends on the time of day, your location, how many people are using your local cell, what LTE Cat your phone is capable of, the upstream link and much more besides. As an example, if you do a speed test at 6AM on a Sunday you’re going to get some blistering speeds, but do it at 8.30AM on a Monday and it’ll drop significantly.
There’s something else to remember. Despite what you’re about to read, it’s not all about speed. Speed tests work by downloading and uploading data, and a pretty big chunk too. This uses up your data allowance and I quickly burned through my monthly limit. If you’re going to use the Ookla Speed test apps then expect your monthly data package to vanish very quickly indeed. What we see with 4G, in general, is an uptick in responsiveness and less chance of it all slowing to a crawl.
To test this I had two phones. Both with giffgaff SIM cards – one on a 3G-only plan and the other on 4G. The 4G SIM sits in the standard LG G3 Cat 4 model. I did the same speed tests at the same time on the same speed testing servers from the same location to see how they compared. Here’s the setup in action. You can already see the difference just on this quick photo..
My initial tests were pretty disappointing if I’m honest, but I’ll explain why.
Speed tests are always a little “iffy” and should have a certain amount of salt sprinkled over them. A speed testing server, when too many people use it, isn’t going to be as reliable as when only a couple of people use it. One of the slowest speeds I got in my 4G testing was 2.48Mbps download. This could’ve been a slow speed testing server (there’s lots to choose from), a bad location (I was near to a window for most of the testing, but this one was inside a building) or just capacity issues and a lot of people using the network / cell at the time (9.28 on a weekday morning).
I’d done another test just two minutes earlier (at 9.26) just to see how fast that was and received a 5.95Mbps download speed, whilst another at around the same time showed a 3.14Mbps speed.
Not what you were expecting? Wanted more? Well hang on a minute, because let’s just have a look at how the 3G speeds compare. Here’s the 9.26 test on 3G..
0.42Mbps. That’s basically around the speed you’d expect from the first ever type of home broadband that came to market (remember the 512Kbps or “half meg” ADSL connections?). On a phone that’s still going to give you pretty rapid web pages etc, but it’s a constantly evolving and self-perpetuating technology. A few years ago we’d be browsing “mobile compatible” websites over 3G which were horribly hacked-down versions of the full-fat desktop ones. Now you’re seeing “responsive” sites, which basically automatically adapt to mobile screens without losing a great deal of functionality. With 4G these sites are as nippy and responsive as the older “mobile versions” used to be on 3G.
So, basically put, at peak time, from inside a building, I was seeing 4G speeds between 2.48Mbps and 5.95Mbps whilst on 3G I was seeing between 0.42Mbps and 0.89Mbps. When I did tests head-to-head they were always more than 5 times faster on 4G. As an example, I got 0.80Mbps on 3G and, at the same time, 5.12Mbps on 4G.
Best speeds? Well, as I checked throughout the day I could see a familiar pattern emerging. Sit next to a window and get a full signal and your speeds will improve. Speeds tended to dip a little at the expected times. As an example, the classic “rush hour” was a peak time. When people were on busses or walking to work (I was in a large town testing this) you’d see the speeds dip. Likewise, speeds were down at lunchtime and in the evenings. The best speeds for me were in the morning at the weekends.
On 4G my best speed was 24.21Mbps at 08.51 on a Saturday morning. When I did the same test at the same time and place on 3G I received a 6.53Mbps download speed.
Don’t forget, I got a 24.21Mbps download speed on 4G in exactly the same location as the 2.48Mbps 4G result. There’s definitely a large degree of “flex” in the results here, but compare that with my 3G results, which topped out at 6.53Mbps but went down as far as 0.42Mbps.
Let’s imagine that you’ve got a bunch of emails that you’d like to receive. Print them off and stick them in a suitcase, then stick that suitcase into a car. With 4G you’re basically driving around in a really big and powerful Jaguar. With 3G you might have yourself a Fiat Punto. Both vehicles will get your emails to you and there’s going to be no difference in the size of the suitcase, but one suitcase is going to get there quicker than the other. Trouble is, they both have to travel down the same roads sometimes, and those roads can get a bit congested at peak times.
Overall then, yes – 4G is an improvement and a great boost over 3G. However, don’t be expecting the top-end speeds at peak times.
At this point I’d normally end an article like this. However, my wife kinda put a spanner in the whole thing. At one point in the testing I was getting between 3-4Mbps download on my 4G giffgaff SIM whilst testing. My wife overheard me mumbling about it and tried a speedtest on her 4G Vodafone handset to the same server at the same time. She got 27.72Mbps, which was faster than any of my tests.
On the flip side, her upload speed wasn’t great compared to mine, but it highlights another huge batch of variables. See, I’m not going to pretend that I know everything about how these 4G masts are put together or how they’re connected. Is bandwidth managed? Are 3G connections slowed down a little while 4G connections get a more premium share of the service? I’m not even going to guess as to whether traffic on 4G goes up a separate pipe to the 3G traffic. Perhaps it does, perhaps it doesn’t. Who knows? Does each mast get an upgrade to the backhaul pipe when it gets 4G? It seems like an obvious “yes” but how do we know? We’ve only got speed tests to compare networks and the results aren’t always incredibly accurate.
However, what I can say is that infrastructure is hugely important. You can stick a 4G transmitter up tomorrow, but if it’s connected to the data backbone with a wet piece of string then there’s really no point to it. It’s like to speedometer on my car. Sure, it SAYS that it goes to 160mph but the engine inside won’t do it. Likewise, simply showing a 4G symbol doesn’t mean that you’re going to get a set minimum, an expected maximum or a constant average. It can be all over the place depending on load, infrastructure, location, bandwidth management and more. What I can say is that at all points during my testing, it was more than 5 times quicker than 3G.
If you want to learn more about 4G and what giffgaff are getting up to, read this giffgaff blog post where all will be revealed.