Broadband News

Ofcom calls for potential broadband USO firms to step forward

The broadband Universal Service Obligation is a welcome safety net of last resort that Ofcom is working towards delivering and Ofcom has set out more detail on the broadband obligation and is calling for prospective USO providers to step forward to be counted and importantly for operators compensation is discussed (deadline is 20th August 2018).

Correction 2:40pm: Ofcom has pointed out that the definition using sync speed is part of the The Electronic Communications (Universal Service) (Broadband) Order 2018 and forms the technical specification of the USO. So it is not a case of Ofcom watering down the broadband USO as we originally suggested. While it is possible our fragile memory forgot this, it is also highly likely that the public will have forgotten or not even been aware, therefore we have retained our paragraph highlighting why sync speed measures can be problematic.

We all talk of the broadband USO being a service of at least 10 Mbps download and 1 Mbps upload speeds, but crucially today the Ofcom publication is talking of sync speeds, which is interesting as this brings ADSL2+ back into the game where we have previously discounted it because while 1-1.1 Mbps upload sync speeds are possible once overheads are taken into account the public will only see speeds on the devices of below 1 Mbps. The overheads vary across ADSL/ADSL2+/VDSL2 but usually in best case scenarios you need to allow for 4% to 10% of sync speed being lost to overheads (Other technologies may have different levels of overheads involved).

Technical specification of the USO

The Order states that: “affordable broadband connections and services must be provided throughout the United Kingdom with all the following characteristics
  1. download sync speed of at least 10 megabits per second;
  2. an upload sync speed of at least 1 megabit per second;
  3. a contention ratio of no higher than 50:1;
  4. latency which is capable of allowing the end-user to make and receive voice calls over the connection effectively;
  5. the capability to allow data usage of at least 100 gigabytes per month.

Eligibility for the USO

The legislation states that premises will be eligible to request a USO connection in the following circumstances:

  1. where a broadband connection that meets the USO specification is not already available;
  2. where such a connection will not be provided by a publicly-funded rollout scheme in the next year (see Annex 4 for further information on existing rollout schemes); and
  3. where the cost of connection is less than £3,400. Where the costs are higher than this amount, consumers and businesses will be able to receive the service if they pay any costs over £3,400.
USO Specification - Ofcom

The footnotes give a definition of latency not exceeding 200ms, so this does rule the standard geo-stationary satellite broadband services out of the equation (unless a hybrid solution was used, e.g. low data rate ADSL2+ with low latency for voice, and high bandwidth traffic pushed over the high latency satellite solution). The 50:1 contention ratio is we believe a very theoretical thing as for standard broadband services this has not existed as a specification since 2004 - the more common thing to do in 2018 is state what the commited bit rate at peak times is.

The eligibility criteria is interesting as this seems to be an olive branch for the numerous fixed wireless operators across the UK, since if their service is available then a USO provider cannot use the scheme to overbuild them, though given the concerns we have had voiced to us over some wireless providers over stating coverage there are likely to be a number of provider says no, but USO scheme says yes then can. The keystone appears to be the Ofcom Connected Nations Report which should have quarterly updates going forward we believe, but as we have seen it is not always 100% accurate, invariably when comparing with our data sets we do agree but there are various edge cases and sometimes what appear to be over optimistic data sets provided by operators e.g. full fibre ticked as available but people still not able to actually order it months after the release of the dataset.

The differences in how the issue around the speeds 10/1 speeds are handled likely explains the difference between our figures and that of Ofcom who are talking of around 900,000 premises needing USO interventation in Jan 2018 down from 1.1 million in April 2018. If we ignore the upload criteria our latest figures from 18th June indicate 631,000 premises don't meet a 10 Mbps download speed criteria and when we exclude all ADSL/ADSL2+ services too this rises to 1,023,000 premises; therefore if we were to assume that reasonable speed ADSL2+ e.g. 13 or 14 Mbps were able to hit a just over 1 Mbps sync speed we would very likely agree with the Ofcom premises figures.

The document does make it clear that the broadband USO is not a roll-out programme, but one the public will need to request and that delivery is not overnight but infrastructure may need to be built out to people, how long will of course depend on what technology is used and how close it already gets to a location.

USO providers can be regional rather than UK wide with the smallest area that will be considered is a local authority area, which may prove challenging for some rural operators in terms of connecting people in the local urban town.

An important point is that operators will be required to offer uniform pricing across the UK and the USO service cannot be more expensive per month than the same product offered in commercial areas and under the proposals when you enquire about a USO service the operator will have 20 working days to respond as to whether it is possible and if delivery is within the £3,400 limit and if not give the option for people to pay the difference.

The £3,400 limit is interesting and the Ofcom document does go into some depth that the cost should not be bourne by the first person to order but that the ability for others to access the new infrastructure should be taken into account.

On the topic of who funds the work involved in delivery for that £3,400 limit how the USO will be funded is best explained by the following flow chart.

Flow chart of how providers will get paid for USO work
Compensation flow chart of broadband USO providers

To end there is some good news Ofcom appears hopeful that they may have things in place that around Summer 2019 that the processes may be in place and ready, though whether providers will be ready is another story.


@thinkbroadband That contention ratio section is bizarre. No-one has cared about contention ratios for ages, and with good reason.

  • @ultrafastcarl
  • comment via twitter
  • over 2 years ago

Their detailed spreadsheet is very confusing at first sight.

They have 1268180 premises below 10Mb but only 925071 covered by the USO. I assume that the difference is sites that are planned to be covered within the next 12 months.

Also some councils have more covered by the USO than below 10Mb some have fewer which confuses the above totals. eg Southampton Below 10Mb 1298, USO 2403. St Edmundsbury below 10Mb 2830, USO 1793.

  • jumpmum
  • over 2 years ago

The upload is the key to understanding their figures I suspect.

e.g. on Southampton we have
544 sub 10 Mbps
3175 sub 10 Mbps sub 1 Mbps (i.e. bigger because we include ADSL2+ only in this)

On St Edmundsbury its different
1516 sub 10 Mbps
1613 sub 10 Mbps sub 1 Mbps

The Ofcom data is of course only as good as what providers supply and with the corrections we get done for people we know its not 100% accurate and also changing every week.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

According to the spreadsheet there are 5.8% of premises in the Cotswold area that are below the USO - I find that surprising since the Fastershire rollout there is supposed to be complete, and the Gigaclear FTTP contract was for 100% coverage in that lot area.
It doesn't bode well as in my area (Forest of Dean) Fastershire contracts that run into 2020 are only targeting 65% of sub-superfast properties, and there are currently 12.8% of properties below the USO.

  • sheephouse
  • over 2 years ago

Ponder this....

If qualification for USO is based on reports from providers i.e. sync speeds that go into Connected Nation Report, then all those border line people who want to benefit but due to optimising line and routers (e.g. force low margin) need to actually put back their ring wires and remove tweaks so that when eligibility checks are done the line will be slow.

Or just as simple as disable ADSL2+ mode and use ADSL2 mode to pull speeds down just enough.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

"where such a connection will not be provided by a publicly-funded rollout scheme in the next year"
So what happens if it is in the plan but that it doesn't actually get rolled out? We know there are any number of connections that were expected within a time frame but kept getting delayed so people could end up waiting for years but always have a prospect of connection "in the next year".

  • ian72
  • over 2 years ago

I guess at that point the property falls within scope, though how soon for this to filter back down to Ofcom is hard to know since the public funded roll-out is not necessarily the same provider as will hold the USO.

Also reasons why an area fell out of public project could also mean high chance of failing on cost criteria.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

Hi Broadband Watchers.
There are many Post Codes that are showing TBB orange (10) but are able to support higher results also there are ( all in Cabs ) being inserted in the cable runs. There are also BT lines that can achieve 10 Orange at a range of 1.6 miles from the CAB when free of interference.

  • Blackmamba
  • over 2 years ago

Orange indicates our speed estimate rather than user speed test results.

So tell us where these MANY postcodes are and we can investigate, and I don't mean vague hand waving village names or maps where detective work is needed to figure out where it is, send the actual postcode.

For others reading we have looked at previous reports from Blackmamba and in many cases the estimate we show is in line with the low end Openreach figure, which is where our model aims to be i.e. realistic rather than overly optimistic.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

Post a comment

Login Register