Broadband News

Yougov survey and difficult of judging whether broadband is better or worse

The newspaper headlines scream broadband was worse for a third of households during lockdown, but judging what is better or worse is very difficult. Why?

Well for one as the survey suggests people were online a lot more (73% of respondents) and that means the possibility fo spotting random drop-outs is much higher. Additionally lots of people have probably brought a laptop home from work to work from and many recent ones do not have Ethernet and so they may have traded a reliable corporate wi-fi network for unreliable home wi-fi with congestion in the airwaves as neighbours networks overlap. Other issues people might be attributing to the internet/broadband when answering these surveys is problems with online shopping e.g. supermarket sites being inaccessible or site load causing a shopping basket to fail.

The minimum figure for people saying it was worse during lockdown should be 22%, because this is the proportion of the online households using Virgin Media who suffered the blackouts on 28th April that went on for periods of 5 to 15 minutes over the course of a few hours. This was an exceptional event and would have made headlines even without the pandemic, but there has been no indication that the issue was load related i.e. one of those fortunately rare but major incidents that happen. Our own watching for dips in performance has been heightened, i.e. whereas we would wait for two or three alerts over a number of hours we have tended to highlight possible slowdowns on the first internal alert.

One and it is a sample of one, but no doubt others will know someone similar is a poster on our forums who thought based on their previous usage they would be OK during lockdown with their ADSL/ADSL2+ and its speeds around 4 to 5 Mbps (multiple people in household). Safe to say their broadband experience during lockdown will not have been good if they all tried to work or zoom video call at the same time, the irony VDSL2 was previously available though at only around 15 Mbps speeds and there is a also a full fibre option.

We have hunted for a link to the YouGov survey but cannot see an article on the YouGov website yet.

ISPA sent across a statement on Thursday i.e. they had seen an embargoed copy of the survey.

As indicated by the survey results there has been a significant increase in broadband usage in households during the lockdown with the UK's workforce logging on from home and friends and family keeping in touch only virtually. However, despite this large increase in usage, the network has proved to be resilient and continues to provide vital support for the public through this difficult period.

The performance of the network itself has a vital but not exclusive impact on the user experience – in-home devices, the number of users, corporate IT set-ups and capacity constraints within online services (e.g. video-apps) can play an equally significant role. However, there are several steps that users can take to improve performance - Ofcom, the telecoms regulator, has published a useful guide on this.

Till Sommer, Head of Policy of the Internet Services Providers’ Association

No doubt there was people with genuine problems, but all the data suggests a possible 2 to 3% decline overall i.e. nothing significant compared to the disruption that has gone in other parts of our lives. Obviously there will be individuals with problems e.g. broadband has totally broken but that happens anyway, of course without the usual home worker backup option of the local coffee shop this becomes a bigger issue and with the problems of getting through to support the level of annoyance will have quickly grown.

If you are reading this at while on a break from work, best advice - plug everything into your broadband via Ethernet and if on an important zoom call turning off the Wi-Fi will help stop someone else in the house causing a blip and a drop of your call. The scope for others activity impacting your own decreases as your broadband speeds increase, but download services such as Steam can cause problems even if you have a connection of a few hundred Mega bits per second and let if download at full speed.

Also now that installs are getting back to approaching normal with additional social distancing measures for safety the next few months are the time to check our what better broadband options you have are and our package search will show the majority of options. We work hard to make the filtering as good as we can and one of those things means we don't offer packages that you cannot buy.


Surely the headline ".. broadband was worse for a third of households during lockdown" only applies to those who completeed the survey. That is not the same as saying broadband was worse for a third of all internet users.
Is it the case that there is more inclination for someone to complete the survey if thay have a complaint to make?
This is true of reviews left on sites such as Trustpilot which tends to attract people who've had a problem and not those who had a good service.
I'm retired so my internet usage has not altered during lockdown but my Plusnet fibre has been no different

  • zhango
  • about 1 month ago

Agree with Zhango, and suggest that 2,300 is far too small a sample from millions of BB users. I'd rather go by TBB's monitoring. For what it's worth I and friends online daily notice no difference.

  • mike41
  • about 1 month ago

Bear in mind also, that streaming providers like Netflix voluntarily throttled their bandwidth for the duration of the lockdown. What will be more interesting, will be what the performance after lockdown is when these streaming services begin to allow full bandwidth streaming, and a lot of employees are still working from home. I suspect that the resilience, especially in more rural areas, will struggle to perform.

  • mollcons
  • about 1 month ago

Services in rural areas are not provided with less capacity per user for fixed line services generally. For Openreach which will be the great bulk of rural areas their data travels over fibre from the cabinet or exchange to one of 1,200 main urban exchanges

Resilience can mean many things, but based on your wording it seems to be a reference to capacity.

  • andrew
  • thinkbroadband staff
  • about 1 month ago

My experience is that for all conference calls / home streamed video is to ensure that your cloud storage is paused / stopped on all devices, as one unused but turned on device can suddenly swamp your upload by starting a backup just when you don't want it to.

Many self employed service type people have started to provide 'classes' etc online and don't realise that stopping others using devices while they broadcast does not stop the device using the internet, ( unless turned off), and can in fact cause it to decide that it will backup during the period it is not being used.

  • jumpmum
  • about 1 month ago

zhango and mike41
These surveys usually need to be managed by contacting a random but representative sample of broadband users, else yes a risk of self-selection bias.

1000 people can be sufficient for a statistical sample of a single variable, the problem comes if you need multiple questions or are comparing answers across different demographics.

Once you have grouped people according to multiple factors when looking for correlations, the number of people so banded can be magnitudes less than 1000 and become unrepresentative.

In that case a much larger overall sample is desirable.

  • prlzx
  • about 1 month ago

re: jumpmum

Good QoS would mitigate that, at least on the upload direction, but only a minority of users would have the tools or knowledge to apply that.

It would rely on the conference software marking its real-time media traffic (e.g. using DSCP) and the router reserving/prioritising accordingly.

Harder enforce on the download side but it could still use WRED or even tail-drop to signal back to the sending side to slow down.

A firewall/router OS like pfSense has these features built-in, but would be great to see basic implementations on more consumer routers as has happened with CoDel.

  • prlzx
  • about 1 month ago


I was talking about the thousands of ordinary users using basic consumer devices to stream their service and the hundreds of thousands ( maybe millions) watching and interacting with them. Far easier cure for them is to turn off the cloud services on the other devices in the house. BT Cloud, Google docs, One drive etc that are all loaded by default, some people don't even realise they have multiple services and are using 2 or more!
The improvement when these are turned off can be dramatic!

  • jumpmum
  • about 1 month ago

Post a comment

Login Register