SamKnows is now part of Cisco | Learn more

Spotlight: Broadband nutrition labels

The US FCC is mandating ‘nutrition labels’ on broadband providers, in a bid to give consumers more accurate information about the level of service they will receive and any hidden costs. Will this lead to more honest broadband marketing and better deals for consumers? Or is the information on the labels too vague to make a meaningful difference? We examine the FCC’s proposals and how it compares to other regulators around the world.

Download this issue

Broadband performance transparency

What are the true broadband speeds consumers should expect? How high is the latency on the line? What are the hidden costs?

These are questions that broadband consumers have faced for years, but the answers haven’t always been forthcoming. That looks set to change in the US, however, with the FCC once again throwing its weight behind a ‘nutrition label’ scheme, which would “require broadband providers to display easy-to- understand labels to allow consumers to comparison shop for broadband services”.

In the same way a pot of yogurt will display key information such as fat and sugar content, so broadband providers will be told to disclose details such as average speeds, data allowances and network management practices. The goal is to arm consumers with more information about the broadband products they’re buying, to give them a true sense of the performance and the costs they can expect once a line is activated.

Labels aim to ease consumer frustration

What impact will the nutrition labels have on broadband buyers and the providers themselves? And are there any lessons the US can learn from countries who’ve pursued similar initiatives before? We’ve spoken to experienced regulators and SamKnows’ own expert to find out.

Is speed still the killer metric?

Packet loss and latency becoming critical for modern apps

It’s fair to say consumers, broadband providers and regulators have been largely focused on raw download speed since broadband performance first started being measured independently. However, the increased reliance on videoconferencing seen during the pandemic and the advent of cloud gaming and ‘metaverse-like’ applications has thrown the focus onto other performance-critical metrics.

For example, Walter Johnson highlights the importance of including packet loss data on the new labels. “Don’t forget, the last administration cut back and took out packet loss, as if it didn’t matter,” he said.

“The rule of thumb is if you have more than 1% average packet loss in your network, it’s pretty bad. If you have more than 2% packet loss, it’s pretty unusable for a lot of applications.”

Johnson says he would like to see the new labels encompass a metric for carrier consistency, shifting the focus away from speed and giving a greater weight to factors such as packet loss and latency, even if many consumers don’t understand those terms.

“The information can be coalesced into nuggets that people understand,” Johnson said. “You can turn that into a metric that’s one form of reliability and compare that to an industry standard.”

What's on the labels

The FCC is consulting with broadband providers and other interested parties on what exactly should appear on its labels, but it’s not starting from scratch. In 2016, the FCC launched a voluntary labelling scheme for broadband providers, but it’s fair to say it didn’t garner widespread support. We’ll come back to why it didn’t take off later.

In the meantime, however, the FCC is using the 2016 label template as a starting point for the new, mandatory scheme – and nobody could accuse the FCC of not trying to be comprehensive. There are no fewer than 18 different pieces of information that broadband providers will be told to disclose to consumers at the point of sale. These can be broadly broken down into three different categories: charges, performance and network policies.

When it comes to charges, broadband providers will be required to disclose the monthly cost of the connection, both if the customer is on a month-to-month deal or a two-year contract. Data caps, excess data charges, the cost of renting modem/gateways and any other fees must be disclosed before purchase.

As for broadband performance, broadband providers will be forced to declare typical download and upload speeds, latency and packet loss. On the sample templates, typical is defined as “typical peak usage for this tier of service, consistent with the Open Internet Orders and FCC guidance”.

On network management, providers will be asked to disclose whether they manage specific applications (such as limiting the amount of bandwidth available for game downloads or video streaming, for example), as well as providing details of their privacy policy and customer complaints line numbers.

It’s not only fixed-line providers who will be subject to the labelling system, either. There’s a broadly similar label for the mobile broadband providers too, this time with a little more detail on data charges for specific allowances (ie. 20GB, 40GB and 100GB plans).

There are no fewer than 18 pieces of information that broadband providers must disclose

How labels could look

How might the nutrition labels convey information about broadband performance? We asked SamKnows’ head of creative to give us his ideas, shown below.

Will labels stick?

As previously mentioned, this isn’t the FCC’s first swing at nutrition labels. The first attempt to implement them in 2016 fell flat. However, FCC chairwoman Jessica Rosenworcel said from the moment she took office in January 2021 that “we have work to do to put consumers first” and she seems intent on doing just that. The key change this time around is that there will be no wriggle room for ISPs. “The FCC’s last effort to put labels like this in place never got that far,” said Rosenworcel in a recent statement. “It was also just voluntary. But this is different. The broadband nutrition labels we are proposing today are mandatory.”

She wants the broadband labels to be as easy to understand as the health information on a pack of cornflakes. “If you walk into any grocery store and pull boxes of cereal from the shelves, you can easily compare calories and carbohydrates,” said Rosenworcel. “That’s because they have a common nutrition label. It’s black and white, simple to read, and easy to understand. It helps consumers make good choices.”

“I think the FCC needs to o the same with broadband,” she added. “That’s because it is now an essential service – for everyone, everywhere. So we want to make it easier for consumers to compare their options and understand just what they’re signing up for. We want to end efforts to bury facts in the fine print and we want to stop unexpected costs and fees.”

Allowing consumers to accurately compare the performance of different broadband providers is key, agrees Walter Johnson, a former member of the FCC’s Office of Engineering & Technology and a chief architect of the Measuring Broadband America programme.

The FCC’s Jessica Rosenworcel

“There’s no marketer in the world – whether it’s internet services, cars or potatoes – that really wants to be transparent,” he said. “They don’t want to say ‘well, I’m selling potatoes at $10 a pound, but if you look down the street you can buy at $9 a pound’. It is against the grain of a marketer to provide any information which makes their product comparable to some other product.”

The only exception to that rule, Johnson said, is “if you can definitively prove you’re better than somebody else”.

As was witnessed with Measuring Broadband America, publishing data prompted the broadband providers to raise their game around reporting periods – and the improvements that were made were lasting.

We want to end efforts to bury facts in the fine print and we want to stop unexpected costs and fees.

How labels could look

Here’s an alternative set of label designs from SamKnows’ head of creative, this time focusing on color-coding to help consumers rate performance at-a-glance.

Making meaningful data

Walter Johnson believes the success of the 2022 labels will also hinge on the reliability of the information they provide. That will require strict standardization of the performance metrics being quoted to prevent providers from naturally trying to make their products look as good as possible.

“I can show you an average number and say ‘well, here’s my average performance,’ but maybe it’s taken when the sun is out and the moon is full,” said Johnson. “There’s different ways to calculate these [averages] and it’s not the number itself that’s important, whether you’re getting 100 megabits under one calculation or 90 megabits under another calculation. It’s how you compare them.”

Johnson gives the example of one well-known internet speed test, which quotes a latency measurement as the speed of the fastest packet it was able to get through the network during the test. “The important thing is not ‘here’s the magic way to calculate it’,” said Johnson, “but everybody’s got to do it the same way.”

SamKnows’ director of government projects, Roxanne Robinson, points out that the FCC already has access to much of this data measured in a standardized form via the Measuring Broadband America programme.

Stats can easily mask true performance

“It’s an enormous data set that is available to them, particularly on this kind of quality-of-service metric and the industry is very familiar with it,” she said.

She suggests that average speeds expressed as a range, rather than a single figure, might give a more accurate picture of the typical performance that consumers can expect to see.

I can show you an average number.... but maybe it’s taken when the sun is out and the moon is full.

In the current draft of the label, there are only four performance metrics stated: upload and download speed, packet loss and latency. Over time, however, Walter Johnson believes it’s critical to improve the quality of data being presented to consumers at the point of sale, and to focus on specific applications – not least video streaming.

“We need to bring in better statistics on consistency or reliability of the network, because the averages don’t capture the total experience,” said Johnson.

“If you’re reading a gas meter, you really don’t care if it takes five seconds longer. If you’re watching Game of Thrones or if you’re doing an operation on somebody remotely, and we’re heading into an environment where these things matter, then understanding consistency is important.”

“Don’t ever interfere with Game of Thrones,” Johnson warned, only half-jokingly, “at least in the United States.”

Is there a danger that consumers could be bamboozled with too much information at the point of sale? Johnson thinks simple measures could help. “One thing I would like to see on it [the label] is what’s the industry average, because then I know whether the carrier I’m dealing with meets the industry average, exceeds it or is below it,” he said.

The average, Johnson argues, would help ease confusion over technical terms such as latency and packet loss. “Every parent who sends a child to school [in the US] might not appreciate what a 93 or 94 means on a score, but they do know if their child is performing above expectations, at expectations or below expectations.”

Every parent who sends a child to school may not know what a 93 means, but they know if their child is performing above expectations.

Lesson from abroad

The US isn’t the first country to try and give consumers better information when it comes to buying broadband services. In the UK, for instance, regulator Ofcom has long had a Voluntary Codes of Practice on Broadband Speed, which is adopted by most – if not all – of the UK’s fixed-line providers.

The codes were last revised in 2019 and included measures to improve the relevancy of line speed estimates, by ensuring those estimates reflected peak- time performance. They also required signatories to provide a minimum guaranteed speed at the point of sale.

The end result is that broadband speed advertising has become more realistic in the UK. Take for example, BT’s Fibre 2 product, which was previously advertised as an “up to 80Mbits/ sec” or “up to 76Mbits/sec” line. It’s now promoted on its website as having a “67-73Mbits/sec download speed range”. Most other providers quote similar figures, even those who don’t participate in the voluntary code.

“There was a disconnect between the speed that was being advertised to a customer and what you actually might receive in your home,” said Roxanne Robinson.

“For example, you might have had an ISP offering a product at 76Mbits/sec download, but there are so many things that can affect that top speed – anything from the number of subscribers using the network, the number of people accessing a particular service at a particular time, or even just the type of line that you have.

Ofcom operates a voluntary code

I think that Ofcom wanted to make sure that you had information that was more about what you were most likely to get, rather than the absolute best speed you could possibly get in the most perfect conditions.”

“It’s a much more realistic way to sell and that’s really what the voluntary code has helped to achieve,” Robinson added. “It’s much clearer, more realistic information at the point of sale.”

New Zealand is currently creating a marketing code of its own, which too looks set to focus on peak-time speeds, showing that regulators are learning from the experience of one another.

There was a disconnect between the speed that was being advertised and what you might receive in your home.

Bringing the industry with you

The key to any of these regulatory schemes is involving the industry and bringing the broadband providers with you. As we discussed earlier, the 2016 US label scheme partly failed because it became a political football. But giving the broadband providers
a clear benefit for participating, even in voluntary codes, is vital for encouraging participation, according to Roxanne Robinson.

Reflecting on her experience with Ofcom’s scheme in the UK, Roxanne said: “Ofcom and SamKnows have worked together for over 10 years. When Ofcom published its High Level Testing Principles for its Voluntary Codes of Practice, this was based on the SamKnows testing methodology.”

That led to closer cooperation with the ISPs. “Using the SamKnows methodology as a blueprint for how to collect accurate internet performance data meant that we could work with the ISPs in the UK to embed our SamKnows Router SDK into customers’ CPE. This allowed ISPs to conduct testing at an enormous scale, that wasn’t previously possible with just our Whiteboxes.”

Metrics can reveal in-home issues

The end result is that both the regulator and ISPs get access to vastly more quality of service (QoS) and quality of experience (QoE) data than they would have previously, not to mention the benefits of delivering greater transparency to consumers.

“We’ve seen ISPs who want to join projects such as Measuring Broadband America or Measuring Broadband UK because they see the benefit of being involved and being able to talk about it. It’s something that resonates with their customers and it’s something that helps them attract people to their brand.”

The SamKnows methodology allowed ISPs to conduct testing at an enormous scale that wasn’t previously possible.

Listen now

Walter Johnson, former member of the FCC’s Office of Engineering & Technology gives us his frank views of why the labelling system failed in the past and what should be included in the labels this time around. Roxanne Robinson, SamKnows’ director of government projects explains what lessons can the US learn from other regulators when it comes to giving consumers more information?

Nutrition labels Pt 1 - Walter Johnson

Nutrition labels Pt 2 - Roxanne Robinson