HoC 85mm(Green).tif


Digital, Culture, Media and Sport Committee 

Oral evidence: Influencer culture, HC 258

Thursday 27 January 2022

Ordered by the House of Commons to be published on 27 January 2022.

Watch the meeting 

Members present: Julian Knight (Chair); Kevin Brennan; Steve Brine; Julie Elliott; Simon Jupp; Giles Watling.

Questions 404 - 452


I: George Lusty, Senior Director for Consumer Protection, Competition and Markets Authority; and Guy Parker, Chief Executive, Advertising Standards Authority.

Examination of witnesses

Witnesses: George Lusty and Guy Parker.

Q404       Chair: This is the Digital, Culture, Media and Sport Select Committee and our latest hearing into social media influencers. We are joined today by George Lusty, senior director for consumer protection at the Competition and Markets Authority, and Guy Parker, chief executive of the Advertising Standards Authority.

Before I welcome George and Guy, does anyone have any interests they wish to declare? I am the chair of the APPG on New and Advanced Technologies.

George and Guy, thank you very much for joining us this morning. Can you give us an overview of why you think the compliance rate for advertisement disclosure among influencers is so low at present?

Guy Parker: It is to do with the scale and pace of online. Previous people giving evidence to your Committee have talked about the very large number of influencers. The figure of 15 million has been bandied about globally. I don’t know whether that is right but certainly from our experience there are a lot of influencers and a lot of new entrants to the influencer market. There are a lot of brands partnering with influencers, including small businesses that are using online and seeking to do deals with influencers to promote their products and services.

It will always be challenging to enforce our codes and to make sure, in particular, that influencer ads are properly disclosed. That is despite the fact that in recent years we, working with others like the CMA, have made some progress in improving the compliance rate for disclosure, particularly in the more durable formats like straightforward Instagram posts, YouTube videos and so on. Despite that, it is still very challenging and some of the more popular influencer formats now are quite ephemeral.

Instagram stories are a very popular format for brands and influencers. Stories, as you probably know, disappear after 24 hours so that presents additional difficulties for capturing and monitoring them and requires us increasingly to use technology to do that. Our report on the influencer content in advertising of 122 influencers, which I know we have talked about in our written submission, I think back in March, to this Committee, showed specific problems in that area. Most of the 24,000-odd influencer ads that we monitored were Instagram stories and the compliance rate was disappointingly low.

That was a bit before the CMA received undertakings from Instagram and the situation has moved on since then, but it remains an ongoing challenge. We have to carry on with a multi-pronged approach where we ask brands, influencers and the talent agencies, PR agencies and so on that work with influencers to do more. We need to keep doing more and keep investing in technology.

Q405       Chair: That is very interesting and I appreciate the massive scale of it. Do you take an exemplar approach? Do you look at the ones with the most followers or those with the most attraction? Does your categorising look at the brands as well as the influencers, or where does your focus lie? Does it go towards the brands or towards the influencers?

Guy Parker: It depends on the circumstances. Both have responsibilities under our code to make sure that ads are disclosed and that many other rules in our codes are followed. If a brand has done deals with a bunch of influencers and we receive complaints about those ads, we will probably shine the brightest spotlight on the brand but we will also name the influencers—as long as there are not too many of them—in any rulings that we publish. If we have concerns, either as a result of complaints or our own monitoring about a particular influenceroften with a large reach, because they seem to have a cavalier attitude to following the code and disclosing their ads in a number of different deals they are doing with brandswe will shine the spotlight on the influencer, although we will again name the brands if there are not too many of them.

Q406       Chair: It is probably easier to look at the brands because there are far fewer of them; there are not 15 million brands.

Guy Parker: There are a lot of brands, though, if you include small businesses that are doing business online. The barriers to entry for brands as well as influencers are very low, which is a fantastic thing for the economy but of course presents scale challenges for regulators.

Q407       Chair: George, following on from the legislative gaps around marketing that you mentioned in your written evidence, can you outline key areas where improvements must be made and why you think those improvements are necessary?

George Lusty: The starting point is that you need the existing systems to work better. We think everyone in the chainthe influencer, the brands, the intermediaries and the platformshas responsibilities. We consider them all to be traders for the purposes of the consumer protection regulations and at various points we try to target different levels of the value chain through our enforcement activity.

What is clearand where our focus is right nowis that the platforms can do the most. They have the tools at their disposal to allow them to detect what is likely to be an unlabelled endorsement, to create prompts that then trigger a contact with brands. Brands can verify whether or not a payment or incentive was given and then the appropriate sanctions can flow if it turns out that a post has not been properly flagged as advertising. That is incredibly important because that can cause real harm for people. It can cause people to buy things that they would not have otherwise bought or make other decisions that they would not have otherwise taken.

We flagged in our evidence to the Joint Committee on the Draft Online Safety Billwhich I know a number of members of this Committee are part ofsome of our concerns that the proposals in that area could undercut the existing high level of consumer protection under the CPRs. We have specifically recommended that in certain areas the responsibilities of platforms for illegal content on their sites could be clarified. We believe that they exist already because platforms are traders under the consumer protection regulations when they are engaged in commercial activity targeting the UK public. We think in particular, to clarify, that is a proactive standard, and that is the view we have taken, which is reflected in the undertaking secured from Instagram. We want other platforms to do that. We think that if they take those responsibilities seriously and those systems work effectively, it will go a very long way to help to reduce the problem of undisclosed advertising on social media.

I will add one last point. The other issue concerns the existing strength of the powers under consumer protection law. It is a court-based enforcement model. When we are taking civil investigations where we have no ability to secure a fine we can, in some cases, get redress for consumers and get a direction from the court that the practice stops. We think, and we are pleased that the Government agree with us, that there needs to be a defining power for the CMA and other civil enforcers of consumer law.

Q408       Chair: It is what you inherited effectively, is it not? Guy mentioned Instagram and the fact that previously there were found to be major failings, but he intimated that you had secured some commitments from Instagram for its new way of doing things. Could you very briefly outline precisely what you have seen since and maybe quantify the areas of improvement, if there are any?

George Lusty: We are talking about the undertakings we secured from Facebook, which applied to the Instagram platform. They were back in October 2020 and most of them were required to be implemented by the summer last year. We are continuing to monitor them now and we will continue our activity with other platforms and that is going into the current period. At lot of their effectiveness depends on all of the onward prompts to brands. That is probably the area where we still need to see more done, in particular to ensure that when Instagram uses its tools to detect content that it thinks could be hidden advertising, there is direct contact with the brands, the brands follow up with the influencer and, if appropriate, a sanction is imposed. We are still seeing across all platforms a very low level of sanctions being applied in practice on influencers.

We think, though, that what we have secured from Instagram sets the benchmark and we are pleased that it already seems to have shifted the dial. In particular, we asked for significant changes to the terms and conditions relating to the platform to be much more upfront, that people will see advertising. By its very nature, it is a platform that connects people with brands, products and services and, among other things, a paid partnership tool that puts a wrapper around particular posts will be extended and has been extended to all users, whereas it was previously made available to only a limited subset.

The automated tools that have been developed are very significant and effective. That is why we think our focus for the time being should be on the platforms and looking at the ways in which they can ensure that they meet their consumer law responsibilities and are doing everything they can to create the right incentives, processes, prompts and tools to make sure that hidden advertising is hidden no more and disclosed properly.

Q409       Giles Watling: Picking up on something you just said about the automated tools used, do you have those automated tools or do you rely on the platforms to have those automated tools?

George Lusty: We have our own data science analytics, data tech team with the CMA. We are able to use our own experimental tools to run checks on platforms and to scrape large amounts of information and analyse it at speed. We can build tools ourselves quite effectively. The ASA does this too. Those tell us what we can achieve and if we know we can achieve that we know that the platforms, with much greater resource at their disposal, can achieve even more.

Q410       Giles Watling: You said earlier that they have the tools and they can deal with the consumer protection. Clearly there is reticence that comes across from the platforms to use those tools. Do you have a strong arm to get in there to enforce?

George Lusty: Ultimately, if we need to pursue these matters we have to go to the courts and argue, if the facts are appropriate, that the platform has not met its obligations under consumer protection law. We brought a case against Instagram, and we are pleased that it agreed to do the right thing and gave us undertakings that meant we didn’t have to go to court. But ultimately that is the route that we would need to follow if we were concerned that the platforms were not meeting their full obligations and to seek clarity that the obligation is a proactive obligation to take the reasonable and proportionate steps to prevent hidden advertising on its platform.

Q411       Giles Watling: Do you find the threat of legal action is enough? Do you actually have to go to court is what I am trying to ask?

George Lusty: We have had to go to court in recent cases. Our powers are not the strongest because even if we do get to courtand that takes a long timethere is no fine for historic breaches if the court concludes that there were some. There is also quite a powerful incentive to reach undertakings that if the party is willing will hopefully sort out the problem sooner for consumers and create a fix. Most often when we approach platforms and want to engage, we get good engagement and response but clearly ultimately if we had a stronger toolkit we would be more powerful and have greater leverage.

Q412       Giles Watling: That is exactly what I am coming to now. This is possibly an unfair question, but if you could wave your magic wand and get the toolkit you wanted, what would you want?

George Lusty: It is set out in the BEIS consultation on the CMA’s consumer powers. That reflects what we asked for in a submission from our previous chair, Lord Tyrie, and which the CMA has continued to push for because we think that the existing consumer toolkit is underpowered. We think that the general public expect it to be more powerful and to have a greater deterrent within it. We can do and have done a lot, particularly during the pandemic, in getting people’s money back for cancelled holidays and other matters like that, but it relies on parties agreeing to do the right thing to reach good strong settlements in lieu of court action. We do have to go to court and when we do it takes a long time.

Q413       Giles Watling: Moving on to you, Guy, could you outline what is meant by collective regulation, in simple terms?

Guy Parker: Working with partners so that we can regulate better. Those partners can be quite diverse. In the influencer area, it is obviously working with the Competition and Markets Authority but also with Ofcom, which has relatively new powers under the new video sharing platform regime, changes to the Communications Act deriving from changes to the audiovisual media services directive in 2018 that places responsibilities on platforms to play their part in making sure that influencer ads are properly disclosed.

Q414       Giles Watling: I gather that there is a constant dialogue?

Guy Parker: There is a constant dialogue, yes.

Q415       Giles Watling: You come together and come up with the answers of what you need for regulation?

Guy Parker: Correct, and we try to make sure that we are not duplicating unnecessarily, that the work we are each doing is complementary and supporting each other. We talk regularly to George and his team about the work that we are doing and the work that they are doing. The VSP regime is new and we will be having similar conversations with Ofcom. We have worked with Ofcom and the video on demand side for nearly 20 years, so we have a good relationship with them. In the context of influencer advertising, that is the best explanation of what we mean when we talk about collective regulation.

Q416       Giles Watling: What are the strengths and weaknesses of your system of collective regulation as you see it?

Guy Parker: A strength of the ASA system is our flexibility. We have been regulating influencer advertising for many years. The influencer market has changed substantially in the last 10 years or so. When we first started publishing rulings against unlabelled, undisclosed influencer ads, they were quite often on what people called vlogs and not many people use that term any more. TikTok had not even been thought about and Instagram stories did not exist as a format. Instagram did, but it was not anything like it is now.

Q417       Giles Watling: That is the big issue; you are always behind the curve.

Guy Parker: No, I don’t think we are. One of the strengths of the code that we administer and enforce is that it is media neutral, so our regulation over the last 10 years of influencer advertising has not required any change to the rules that were already in the CAP code. We did not need to take the time to put in new rules to capture influencer advertising, new powers to administer and enforce under our code. They were already there. We have been able, therefore, to respond fairly quickly with updated guidance when new platforms and new formats have become very popular with brands.

Q418       Giles Watling: I gather you are fairly reliant on industry funding. Does that compromise you in any way? Does it get in the way of your ability to enforce compliance from brands?

Guy Parker: No, it doesn’t. We are a self-co-regulatory body. We are funded by the industry and the funding is arm’s length. There is a separate company called ASBOF that collects the money, mostly through a 0.1% levy on the cost of advertising. If a brand spends £1,000 on an outdoor campaign it will be asked to provide £1 of that to ASBOF, and we then get our budget from ASBOF. ASBOF’s main reason for existing is to properly fund the ASA system. You only have to look at the rulings that we regularly publish to see that we administer and enforce the code robustly.

Q419       Giles Watling: You are truly independent?

Guy Parker: In the delivery of our regulation, the ASA is truly independent. The chair is independent of the industry and the majority of the ASA Council, which is also my board, is independent of the advertising industry. You can see from our outputs, from the rulings that we publish and the monitoring and enforcement action that we undertake, that it often covers the big guys as well the fringe elements of the advertising ecosystem.

Q420       Giles Watling: It is a question of optics, isn’t it? If you are funded by the industry you may be influenced by the industry, but you say that is not the case.

Guy Parker: No, we are in the business of making sure that ads are responsible. Honestly, I think our record speaks to that, but because we are a self-co-regulatory organisation we have to explain that to people and seek to persuade them that we are not in the industry’s pocket. That comes with the job.

Q421       Giles Watling: That is very important. Finally to you, Guy, we have heard that the Advertising Standards Authority’s reporting mechanisms are being used by internet trolls to harass influencers. Are you aware that is happening?

Guy Parker: It does happen, yes. It is not just in the area of influencer advertising. Sometimes people complain to us about ads and they have an agenda. We are very strict in making sure that we assess objectively the complaints that we receive. If they raise potential issues about the compliance of an ad or a campaign with the advertising codes, we will look to take action. We very much prefer to operate an education-first approach with influencers. A lot of the influencers are young people. They always start off small and of course they do not know the rules because they are not big advertising companies.

Q422       Giles Watling: You have to get in touch and tell them the rules?

Guy Parker: We talk to them and we provide them with advice and training. We have a wealth of advice and training resources on our website. It is very accessible and is written to speak to influencers. That is always the first thing that we do. A lot has happened since we submitted evidence to this Committee back in March. We have ratcheted up our enforcement against some influencers. Those influencers have had repeated opportunities to comply with the code and for whatever reason they have not taken them. The first of the sanctions that we have deployed has involved creating a register of non-compliant influencers on our website, which has been viewed well over 30,000 times by people. It is now being used by brands to decide who to work with. That has helped to bring some of the influencers that we have posted on that webpage into compliance.

In the last couple of weeks we have ratcheted up our sanctions again against the six worst offending influencers, most of whom are ex Love Island, and we have deployed a new sanction that we call our on-platform targeted ad sanction. That involves us using credits from Instagram, targeting our own ads at people who are likely to have an interest in following those influencers to draw their attention to the fact that these influencers are not complying with the code. Those ads have been served over a million times and we have deployed them for one or two weeks. This is very recent. All six of those influencers are now complying with the code.

Q423       Giles Watling: Thank you very much. You mentioned that what you have on your website is accessible to anybody so that they can see what the rules are and how to comply. The issue I would suggest is that we all look at websites, don’t we, “Here are the rules and you go, “Yes, I agree. You must come across that. How can you be more proactive?

Guy Parker: We do a lot that is proactive. We do webinars, we have seminars and events talking about this specific issue. We and the CMA produced jointly-branded guidance called “An Influencer’s Guide” and that has been viewed or downloaded over 35,000 times on our site.

Reality TV celebrities can be among the influencers who too often are not disclosing their ads. We have partnered with ITV and produced a cheat sheet for influencers to help them navigate the rules, the dos and don’ts. That is provided to all “Love Island” contestants by ITV in the duty of care pack that they provide to them after they have been on the programme.

We try to be creative in getting through to the influencer community. We obviously work with influencer professional bodies, which tend to be quite small. There is not one big body that represents loads of them. We work with influencer agencies, talent agencies, the PR community. We are constantly trying to get the message out.

Q424       Giles Watling: So you are proactive.

Guy Parker: I think we are very proactive, yes.

Q425       Simon Jupp: Good morning to the panel. Thanks very much for coming along. Guy, you mentioned new technology in your annual report. I imagine that new technology played a part in identifying, for example, “Love Island” contestants who are not playing by the rules and were not playing by the rules. What does it do, in layman’s terms, and has it been effective?

Guy Parker: It helps us to capture influencer ads, particularly in more ephemeral formats like Instagram stories where they disappear after 24 hours. You cannot look on someone’s timeline and find their old ads, screen-grab them and then analyse them. You have to catch them in the moment. It helps us to capture influencer ads. The machine learning models that our data science experts have built help us to filter the content, because obviously influencers are putting up a lot of content, help us to flag and predict content that is likely to be advertising and tell us automatically whether or not it is disclosed or flagged.

That leads to big efficiency gains for our compliance people because they are no longer having manually to monitor loads and loads of influencers accounts. The machine is doing a lot of the work for them and is flagging up, for human review, the cases that are likely to be problematic. That is deployed and we are using that. We will continue to invest in that sort of data science approach. It is crucial for meeting the scale challenge of regulated influencer and a lot of other advertising online.

Q426       Simon Jupp: Can you or should you have access to, for example, Instagram stories once they have disappeared after 24 hours to make your job easier? Can you do that? I know that they disappear for mere mortals like me, but can and should you be able to access those stories?

Guy Parker: It would be great if we could do that. I don’t think that the APIs that Instagram makes available cover Instagram stories yet.

Q427       Simon Jupp: Is that something you would like to see? If there is a social media platform that does 24-hour stories, 12-hour stories, whatever they might be, should you be able to view them afterwards so that you can properly review their content? Technology is great but it will not capture everything, will it?

Guy Parker: No, and we have to make decisions about where we deploy this automatic monitoring. Yes, that would be great. We are having ongoing discussions with the likes of Instagram about these sorts of issues. Access to data is critical to get the best out of your data science work.

Q428       Simon Jupp: Without wanting to traipse into private discussions, why is it ongoing? Why can’t you just demand it? Are you expecting to get the powers to see the stories that have lapsed?

Guy Parker: We do not have statutory powers because we are a self-co-regulatory set-up. We have ongoing discussions with the likes of Instagram and other platforms about these sorts of things. We have a bigger, broader initiative at the moment, which we call online platform and network standards, which is looking at bringing principles into the code that will allow us to hold platforms to account for the role that they play in making sure that the ads that they make money from are responsible and comply with our code.

Issues around data are very much a part of the conversations that we are having and will continue to have with the platforms. This is all happening, but there is certainly more progress to be made.

The point I will finish on, very quickly, is that we are able to capture influencer advertising at the moment through the technology that we have. We will be scaling that up anyway while these conversations continue. We are already able to get sufficient data to make a real difference to our enforcement.

Q429       Simon Jupp: Would you be surprised if social media platforms refrained from giving you access to the stories that have lapsed? It is an interesting point, isn’t it, because, as you say, you will not be able to capture everything, no matter how impressive the technology might be?

Guy Parker: Yes, I would be surprised if they refused to give us the data that we need.

Q430       Simon Jupp: I think I would be as well. It seems that you only began to develop some of the technology you are using, this data science capability, in 2020. Given that we know that the issues we have discussed already in this session have been around for longer than that, why has it taken you so long?

Guy Parker: We have a lot of priorities, obviously. The technology investments we have made in the last few years have been focused in a lot of other areas too. We were the first advertising regulator in the world to deploy a programme of monitoring called avatar monitoring that is focused on age-inappropriate targeting of paid ads online. Think of alcohol ads, gambling ads, ads for food time—fat, salt and sugare-cigarette ads and so on that are being targeted at children online, which is against our rules. The challenge there is that if you and I go to the same webpage at the same time, we will be served different ads, so how do we know what ads are being served and to whom?

Our response through the avatar monitoring programme was to build profiles that mimicked the browsing behaviour of different-aged people, particularly children—an avatar that mimics the browsing behaviour of a primary schoolchild, a child just transitioning to secondary school, a teenager. We then deployed those avatars to hundreds of websites and video channels. We captured the ads that were served to them and analysed those ads, and we have been able to take action on the back of that against inappropriately targeted ads. That has been a big focus of our work as well. We have limited resources, so we cannot do everything at the same time. We have to prioritise our work.

Q431       Simon Jupp: To develop a picture, can you give me an example of a couple of adverts you were discussing that breached the guidelines, that you were concerned about, that were targeting children, just to give an example for the Committee this morning?

Guy Parker: The headline from the first sweep of avatar monitoring that we did back in 2019 is six gambling operators were being slapdash in their targeting, working with their ad tech partners. They were not paying enough attention to the targeting of their ads and they were serving gambling ads on websites that are popular with children.

Q432       Simon Jupp: Thank you, Mr Parker. I might come back to you in a minute. Mr Lusty, what monitoring analysis are you conducting on the state of the influencer marketplace, vast as it is?

George Lusty: We have a slightly different set of information-gathering powers from Guy at the ASA. We have statutory information-gathering powers, which means that as well as monitoring, we can ask directly and we can follow that up with court action if a firm or individual for any reason were not to comply with our information request. That, to some extent, gets around some of the issues that exist in the ASA’s regime, but we also do our own direct monitoring.

We are a regulator for the whole UK economy, so looking at what is going on on social media platforms is just one part of what we do to look at the economy. We use social media monitoring software to identify a raft of different consumer profiles, to see if there are spikes of complaints by people about any particular product, service or trader. We use social media proactively in our wider programme of work.

Specifically in the area of influencer marketing, we use tools in a very similar way to the way that Guy has described—to capture evidence, to allow us to sift very large volumes of posts and to use technology to identify indicators of possible hidden advertising. There is always an information gap, because we do not know whether a payment or gift or other incentive has been made, and we would need to use information-gathering powers—and we do that—to find out whether or not that exists. It can give us a starting point, and it is exactly that kind of tech that in our view forms part of the solution. It is effectively the base standard that we think all platforms should be using with automated tools to create the prompts that trigger a conversation with brands, which then leads to sanctions on either them or influencers as appropriate. Those need to be effective and—

Q433       Simon Jupp: We are not there yet, though, are we? Given that this industry has blossomed in the last couple of years, when do you expect we will get to the point that you have reached your wish list of powers to solve this problem or at least solve some of the problems that occur as a result of this?

George Lusty: We are following up the Instagram undertakings with other work with platforms. That will be ongoing, and my team is working on that as we speak. We hope that will push that standard out and it will be recognised, accepted and welcomed. Indeed, from the evidence they have given to this Committee, we understand that platforms recognise the need to take those proactive steps. We hope that we will be pushing at an open door with that.

We think that what needs to follow is that everything within the value chain needs to act on its responsibility. We have, as I have said, flagged some areas where we still think there could be greater clarity introduced within the consumer protection regulations about the extent of the duty on platforms to proactively address illegal content. We think that stronger enforcement powers for us and for our trading standards counterparts and other sector regulators with concurrent consumer law powers would make a huge difference in creating the right deterrents within the enforcement framework.

Q434       Simon Jupp: Where are those discussions? You have your wish list, but where are they at the moment, and when do you expect to see further powers given to you and your counterparts?

George Lusty: BEIS has consulted on those powers and we are awaiting the publication of its response to that consultation. It is very much, therefore, a matter for BEIS about the timing of those changes being introduced. We understand that it is proposing to give us what we have asked for, and we welcome that.

Q435       Simon Jupp: At the moment, as someone who is looking at this pretty much from a layman’s perspective, if I am really honest with you, it seems like a patchwork quilt of monitoring and enforcement. Do you think that is accurate?

George Lusty: If you look at our written submission, you will see that when it comes to our action we have tried to tackle it at various different levels, going back to the days of the Office of Fair Trading, our predecessor, and work that started in 2010. We have tried collectively to address what has become an endemic, market-wide problem. As you have said, the market has blossomed, the influencer marketing market has doubled in two years and it is very effective. It is much more powerful than traditional advertising techniques, and it is particularly powerful with younger people in persuading them to buy things.

We think the best way to address thatand the way that we can have the most effectiveness and tractionis to look at the point in the value chain where we can achieve the greatest difference. At the moment, that is looking at platforms and what they can do with automation, detection, prompts, and following it up. Inevitably, what I have described has been going in and trying to sort out the problem at a number of levels, and it still persists. We think that platforms need to act on this responsibility, brands need to do their bit and influencers need to follow the very clear guidance. It is not rocket science. We have been able to articulate it jointly with our counterparts in the ASA, and we have gone out to train people. We put out prompts using user-friendly social media messaging ourselves to say what you have to do, and that just needs to be followed up.

We think that there is a need for more powerful deterrents. We welcome what the ASA is doing with ads directly on the platforms to draw attention to bad conduct, the naming and shaming list, which we think at the very minimum needs to inform what the platforms are doing and inform the sanctions that they are willing to impose. Ultimately, we think that civil fines would be an important addition to the arsenal of regulators.

Simon Jupp: That is very interesting. Thank you.

Q436       Kevin Brennan: Mr Parker, I will ask you a couple of questions first. Do you think that the influencer route is a particularly powerful—to use Mr Lusty’s words—form of advertising and particularly effective, compared with traditional forms of advertising?

Guy Parker: Yes. I think the fact that the influencer market has grown so rapidly and is forecast to continue to grow rapidly—

Q437       Kevin Brennan: I thought I read something the other day saying that young people are pretty savvy about these things and a very small percentage of them say that they are influenced by influencers when they put products in their content.

Guy Parker: We all say advertising does not affect us, though, don’t we? The brands would not be spending the money they are spending in influencer campaigns if they were not seeing a return on their investment.

Q438       Kevin Brennan: Tracking back to what you were saying earlier about ITV and “Love Island”, the duty of care provisions and the work you have done with ITV on that, are the contestants on programmes like that contractually obliged, in their contractual arrangements with the broadcasters, to follow a code of conduct around this sort of thing or to agree that they will comply with Advertising Standards Authority regulations and so on?

Guy Parker: That is a good question. I don’t know.

Q439       Kevin Brennan: Do you think it would be useful if they were contractually obliged and it formed part of their contract to go on the programme that they have to be responsible and follow the rules in any influencer-type advertising activity and product endorsement that they take part in?

Guy Parker: I am always a fan of that sort of thing, yes. It speaks to the point that George made earlier that we need to carry on with this multi-pronged approach. Everyone has to do their bit. One of the parties here is a broadcaster, ITV. I have to say that it has taken seriously its responsibility in this area—

Kevin Brennan: It is a public service broadcaster.

Guy Parker: as has the trade body for the big advertisers and the IAB, but we have to keep at this, and keep getting better. ITV has the duty of care to, for example, “Love Island” contestants and we do not. We have a responsibility to protect the public and provide a level playing field to responsible businesses.


Q440       Kevin Brennan: If it is not the case that they are contractually obliged, that might be a sensible suggestion in this area for us as a Committee.

Guy Parker: I recommend that you talk to ITV first, in case it already does that.

Kevin Brennan: Of course. I am sure it is watching, so I am sure it will let us know.

George Lusty: More broadly, if brands, in their own contracts with influencers, were routinely including terms and conditions requiring the appropriate labelling and disclosures, I think that that too would make a very big differencethe use of model terms and conditions that would put those responsibilities very clearly out there as between the businesses, the brands and the influencers that they contract with.

Q441       Kevin Brennan: Thank you. That is a very helpful suggestion. Mr Parker, a couple more questions to finish. A lot of your sanctions rely on naming and shaming. Do you believe that brands and agencies suffer reputational damage in the same way that influencers do by that kind of approach?

Guy Parker: Yes. Thinking beyond influencer advertising, the overwhelming majority of our naming and shaming applies to brands, to advertisers. We know that it is effective. There are challenges when it comes to smaller companies, new entrants to markets, businesses that are disruptors and they are trying to disrupt a market. Of course there are challenges to that, which is why we have other sanctions that we escalate if we need to. We know that naming and shaming works in this space and we see it.

There is a good example from 2019. Lord Alan Sugar posted a tweet that was an advertising tweet but did not disclose, which we investigated and published a ruling. He was quite angry about it. He said a couple of things about us, one of which was that we were jobsworths, but importantly he has since complied and has been careful about making sure that he puts #ad in his tweets, and that is what matters. Lord Sugar has a reputation. Some people do not have such an august reputation as his to protect and that is when we need to ramp up the sanctions. We know that that works. Influencers have come off the non-compliant influencer register on our site because they have started to comply consistently.

Q442       Kevin Brennan: I think we have already talked about the six non-compliant influencers you were talking about earlier. Mr Lusty, do you think that obtaining undertakings from specific influencers is an effective way to enforce influencer compliance?

George Lusty: We have obviously done that. It is going back a while to 2019 when we secured undertakings from 16 influencers whose practices had been of concern. They agreed very quickly to provide undertakings, which effectively is a form of promise and is not directly enforceable. I cannot walk into a courtroom with a credit card and say, “There is a breach”, and the court gives me an order the next day. If we want to enforce that, we still need to take it back through the court system and effectively argue our case from scratch. That is one of the things that we are saying is an inadequacy of the current system. It does not create the powerful leverage that we need as an enforcer to be able to take those things through. However, talking about the effectiveness of naming individuals who we believe are not complying with consumer law, we think it was a powerful thing to do and drove compliance.

Q443       Kevin Brennan: What legal obligations are relevant to social media platforms across the influencer advertising process, and are they being met by social media platforms?

George Lusty: There are a number of different bits of law that bear down on them, particularly in this space. The bit that we are concerned with is the consumer protection regulations that, among other things, require that platforms, where they are acting as traders—engaged in commercial activity in connection with influencer marketing and we consider that they are—need to behave in a professionally diligent way, which means meeting the standards of honest market practice. In very practical terms, it means that they need to take reasonable and proportionate steps to ensure that people are not being exposed to hidden advertising that can cause them to make decisions that they would not have otherwise made and buy products that they would not otherwise have chosen to buy.

As to whether they are meeting those standards, that was the reason that we decided we needed to begin an investigation into Instagram in the first place. That was a formal investigation, using our powers under the Enterprise Act, which they resolved through undertakings, and we very much welcome that. It is the start of a process because until every platform has tools like that and they are working effectively and being followed up, the problem has become endemic and will persist.

Q444       Kevin Brennan: Do you know that thing on the media where politicians go on representing their party or the Government to talk about one thing and then they get asked about something else? I am going to do that to you now, just for a bit of revenge. The Competition and Markets Authority this morning has issued a press release headlined, “CMA launches probe into music streaming market”. Of course, that probe has come about as a result of a report this Committee compiled—a ground-breaking report, I would say—into the economics of music streaming. You may be aware that I have a Bill that is still alive in the parliamentary process, albeit in the deep freeze at the moment.

The press release says, “While focusing on potential harm to consumers, the CMA will also assess whether any lack of competition between music companies could affect the musicians, singers and songwriters whose interests are intertwined with those of music lovers”. What will be the process going forward for that inquiry and when can this Committee expect it to compile all the evidence and come to its conclusions in that market study?

George Lusty: My lead role is in consumer protection, and that is being conducted as a market study, which is within a different part of the CMA. However, you are right, that launched today. We announced several weeks back that we were intending to launch and that has now happened. It is a formal market study, which means that the CMA has access to very robust information-gathering powers and operates within a tight statutory timeframe to conduct its investigation to gather evidence from parties using those powers to first publish its initial findings and to put those out to see views on them before concluding its final report.

There can be a number of outcomes from that. It can lead to direct recommendations to Government. It could, of course, lead to the use of the CMA’s other tools, if things are uncovered as part of it. I suggest, if that is of particular interest to the Committee, that other CMA colleagues could come back and talk to that directly with you. We do, of course, recognise the prompt that came from this Committee, the urge that we take on work in that space. I am pleased that we have been able to announce that is launching today.

Kevin Brennan: On the bombshell that Neil Young’s music is being removed from Spotify today, streaming really is in the news. Thank you.

Q445       Steve Brine: You have a heart of gold, Mr Brennan. Mr Parker, we have heard that influencers are unaware of UK advertising rules, which is a good one because, of course, ignorance of the law is no defence of the law. It could have cross-read in other areas at the moment. Can you tell the Committee about the support services that you provide to make sure that there is no ignorance of the law?

Guy Parker: Yes, I am happy to. I have talked about some of them already such as the jointly branded influencers guide that we have put together with the CMA. There is an awful lot on our website. There are video guides. We provide advice through our copy advice team on influencer labelling for different platforms. That is platform-specific advice, because some influencers are active on TikTok but not other platforms. We have put together with TikTok a series of TikToks with nine of its content creators, with the TikTok content creators talking about the importance of following our rules, not just disclosure rules but including our disclosure rules. That is a really nice partnership. We like this because it is quite creative and is good fun to do. We are constantly challenging ourselves to try to get the message out as well as we can.

For some people, ignorance is not a defence if they are relatively established influencers with talent agencies and so on. I do not think they have an excuse to say, “We didn’t know there were rules here”. With newer entrants to the influencer market, it is an excuse. They do not have professional help and are often very young people. That is why it is important that we adopt an education-first approach when we come across problems with those influencers.

Q446       Steve Brine: Yes, but if I am 17 and I have just passed my driving test, I still have to stick to the speed limits. Saying, “I didn’t realise it was 70” is not a defence to the law just because I am new to driving.
There are so many new entrants to the market, and you rightly say that you are putting all that stuff out there and you are challenging yourselves to do TikTok vids, and the mind boggles. How do you measure whether that is being received? How are you able to go behind the curtain and measure its impact? Just because it is out there does not mean it is being consumed.

Guy Parker: Correct. We can count the number of times pieces of advice are viewed or downloaded. We can look at indicators. For example, we can monitor the use of #ad and we know that has increased in the last three years by 600% or 700%, which I think is a bigger rate of growth than the influencer market. That is a global figure, but it is giving a sense. We talk to people the whole time and we talk to the influencer community the whole time. We get feedback. You will always get some influencers who say, “Well, I don’t think you’re doing enough”. Equally, I know this Committee has heard evidence from influencers and influencer agencies that says that there is plenty of guidance out there.

We need to keep at it because formats change and there is a grey area that George has alluded to, which is genuinely difficult—when is the line crossed that turns a piece of content into an ad requiring disclosure and compliance with the rest of our rules in the code? Those can be quite knotty judgmental issues, and our guidance needs to keep being updated to reflect that. The message of the general principle is that if it is advertising you should be telling people clearly and upfront—members of the public should not have to play the detective to try to deduce the status of a piece of content—is well and truly out there. I don’t think that anyone other than pretty brand-new entrants to the market should have an excuse to say that they did not know there were rules.

Q447       Steve Brine: You talked about different profiles and different ads being served to different profiles, which of course is nothing new these days. Clearly the ASA’s world was much simpler when there was just published media, before there was online media. What horizon scanning do you do on the future? The online world is now quite old hat. There are all sorts of other worlds that are being created—virtual reality worlds, artificial intelligence worlds. What is your organisation doing almost to look into the future? You have to stay ahead, don’t you?

Guy Parker: Yes. It is a great question. In a sense, we have an advantage, because we are a self-co-regulatory set-up. We have very good access to very clever people who work in the ad industry and the tech industries, who care about trust in advertising and responsible advertising and who can help us to understand these sorts of likely future developments, like the metaverse, and what the impact might be on advertising and ad regulation. We talk and think about it a lot. We take advantage of the access that we have to speak to very clever people about these sorts of issues and that is how we seek to stay on top of it. It is hard, and anyone who says they are ahead of this is lying. The pace of change is too fast.

Steve Brine: Finally, Chair, if you will indulge us, I believe that Mr Lusty wanted to come in on the first point we were discussing.

George Lusty: Yes, on your question about how you make sure that people access the guidance so we stay on top of whether they are digesting it. It is a very good question and I think the reality is that we can’t rely on that alone. I want to draw your attention to one of the features of the undertakings that we concluded with Facebook. It was required to provide prompts not only when new influencers sign up about what they need to do to properly label advertising if they are using the platform for that but to create automatic prompts when they start posting what looks like an ad.

Again, the technology allows that. You could use certain keywords that make it look very likely that you are about to promote a product or a service. That then triggers a prompt to say, “Is this an ad? Yes or no. If yes, then here are all the specific paid-for partnership tools at your disposal to properly label it.” If no, then there is also a very good record on that platform that the person has said no, and if there is in fact a commercial relationship behind the scenes, the very fact that they have said there was not would be, in our view, very material in the platform reaching a decision about appropriate sanctions.

Steve Brine: Thank you very much.

Q448       Julie Elliott: Guy, you have already said this morning that targeting ads at children is against ASA rules. Is non-compliant advertising that is targeted at children actively monitored or do you respond reactively to user complaints?

Guy Parker: It is a good question. Yes, it is proactively monitored. I talked about the avatar monitoring programme earlier. We are on about our third deployment of that. We also have another tech-based programme, which we call CCTV-style monitoring, which is seeking to achieve a similar thing, to make sure that age-inappropriate ads are not targeted at under-18s or under-16s, depending on what the restriction is. It uses technology in a slightly different way from avatar monitoring but its intention is the same.

Using technology, including data science but also working with third-party tech providers, is a thread that runs through our strategy. We are doing it in other areas too. We do not just regulate paid ads online. We also regulate advertising claims that companies make on their own websites and in their own social media channels. We have to deploy technology there too, either our own or third parties’ technology, just to keep on top of the sheer number of websites, Instagram accounts, YouTube videos, and so on. We do that. That will include web scraping of claims in hot areas so that we can identify problem claims more efficiently.

One of the issues that we tackled in the last two or three years was advertising of Botox, which is a prescription-only medicine and it is illegal to advertise it to the public. Surprise surprise, most beauty clinics and beauticians do not know that. Who would know that? Botox is pretty prevalent in our culture. They do not know that, so lots of them were promoting Botox on their organic Instagram accounts. There are thousands and thousands of beauticians and beauty clinics in the UK. We deployed technology to capture examples of these organic ads and we provided regular lists of them to Facebook, which took down the organic posts. There was a lot more to that project than that, but it was a proactive project. It is an important part of our arsenal now.

Another aspect of our strategy in the last several years has been to shift our resources away from the reactive complaint case handling, important as that still is, towards this proactive tech-assisted monitoring and enforcement. It has to be the future. How else can we meet the scale and pace challenge of regulating online?

Q449       Julie Elliott: The CAP code does not have special disclosure requirements for children, despite evidence that they have lower advertising literacy than adults. Will you update the code to make it more appropriate for children?

Guy Parker: We are always looking at the code and arguments and evidence in favour of tightening up restrictions in different areas. One area where we have enhanced disclosure requirements in recognition of the fact that children are particularly vulnerable and are not so commercially aware until they get older—they do not know that they are being sold to—is in ads for games. We have that. We are always looking to do more to protect vulnerable people and children.

I know this is not a direct answer to your question, but we very recently introduced a new restriction saying that you cannot target ads for cosmetic interventions at under-18s. That is in recognition of concerns around body image and some really bad outcomes that people have had when they have travelled to Turkey to have cosmetic surgery. We are constantly looking into these sorts of issues and trying to make sure that the codes are set at the right level.

Q450       Julie Elliott: You have not answered my question about whether you will change the code or update the code to make it more appropriate for children. Will you look into that? If you are not prepared to say you will change it, will you look into changing it?

Guy Parker: I think we are. We have a body image project at the moment that has included a call for evidence, and we have received arguments and evidence, as you would expect, around the level of protection.

Q451       Julie Elliott: It is not just about body image. Body image is very important, and I feel very strongly about that across the piece, but children are just not mature enough to know sometimes whether it is an advert or a fact or whatever, because they are not as developed as people. There needs to be more protection and I think that changing the code would be a helpful step.

Guy Parker: I can say that we are certainly open to that. As with any other request that we should make changes to the code, we would want to see the arguments and evidence that we are supporting a change or a tightening up or the introduction of new rules. We very much welcome those on that point, but there is a lot going on in a number of areas that broadly speaks to your point, if not specifically.

Q452       Julie Elliott: Finally, on children’s exposure to advertising within influencer culture, do you think the amount of advertising children are exposed to should be limited, or is there any way of limiting it?

Guy Parker: That is a very good question. I do not know whether there is any way of limiting it. I think that children’s exposure to advertising for products and services that are inappropriate for them has to be limited, and that is what we seek to do through policing the under-18 and under-16 restrictions that we have in our codes.

We have a second line of defence that there are very strict rules that apply to ads for things like alcohol, gambling, e-cigarettes, foods high in fat, salt and sugar and so on, that relate to the content of the advertising as well, to make sure that that content does not appeal either strongly or particularly to children. Recognising that no system is perfect, and even if you have scheduling or placement restrictions in place, some children will see some ads, it is important that if they do see those ads that the content does not appeal to them. That is a summary of the two-phased approach that we bring to protecting children through our rules.

Julie Elliott: Thank you.

Chair: Thank you. That concludes our session. George Lusty and Guy Parker, thank you very much for attending today.