More insights.
That's the whole lesson sitting inside the £2.4 million the UK government spent on influencers, and it applies well beyond Whitehall. The money isn't the problem. Political organisations everywhere are moving budget into creator work for good reasons. The problem is that the discipline applied to every other line of a media plan gets suspended the moment a creator is involved, and then everyone acts surprised when the spend can't be defended.
Here's what happened.
Between July 2024 and July 2026, UK government departments disclosed paying at least £2.4 million to social media influencers. The figures only surfaced because Conservative MPs filed written parliamentary questions. Education spent £718,971, working with 57 creators in 2025/26 alone, on childcare, teacher recruitment and careers advice. HMRC spent £591,200 across 44 creators, and that number doesn't include agency fees. Transport, Work and Pensions, Justice and Energy Security each spent between £176,000 and £303,000.
Departments defended it on reach. These channels get to people who don't read government websites and don't watch the evening news. That's true, and it's a legitimate argument.
It's also the only argument anyone made. Go through the disclosures and the coverage and you won't find a cost per acquisition, a conversion rate, or a comparison against an audience that didn't see the campaign. Teacher applications, on-time tax filings, benefit transitions completed. All of that is measurable. None of it was measured in public.
It gets worse. The numbers don't agree with each other. A separate FOI investigation by LBC put the total at £2,119,916 since the start of 2024/25, with the Cabinet Office as the biggest spender, a department that barely features in the parliamentary answers. LBC has Transport at £387,873.83; Parliament was told £302,873.83. An earlier FOI exercise by the PR firm Tangerine, reported in December 2025, found just over £500,000 since the start of 2024 across 215 creators. Different windows, different definitions, and departments declining to answer. The Ministry of Justice refused LBC outright on commercial confidentiality grounds, and most departments gave Tangerine the same answer. When nobody agrees what was spent, nobody is going to establish what it achieved.
So the spending isn't indefensible. It's just undefended. Which, politically, is the same thing.
What we do differently
We treat creator work as paid media with a named outcome attached, and the sequence matters more than any individual tactic.
The outcome gets defined before the brief goes out. Not after the invoice. We pick the behaviour the campaign exists to change, write the metric down, and agree it with the client while the budget is still theoretical. Registrations. Attendance at an event. Petition signatures. Applications submitted. If the only number available at the end is views, then what got commissioned was reporting, not results, and the campaign was set up to fail a scrutiny it was always going to face.
We hold back a control audience. This is the cheapest thing on the list and the one most often skipped. Suppress the creator campaign in a comparable region or segment, then compare movement. Without it, every claim of impact is really a claim about correlation, and any competent opponent will say so. Commercial advertisers have done this for decades. Political and public sector work mostly hasn't caught up.
We price against outcomes, not follower counts. A creator with 40,000 followers inside a real community routinely beats a 400,000-follower generalist on anything requiring the audience to actually do something rather than watch. Follower count prices the person. It doesn't price the result. First campaign sets the baseline, second campaign gets renegotiated against what the first one delivered.
And we fix disclosure at contract stage, not at post stage. The ASA's 2024 monitoring report, which scanned over 50,000 posts across Instagram and TikTok, found that 43% of influencer ads failed disclosure requirements, with 34% carrying no disclosure at all and a further 9% using labels too vague to make the commercial relationship clear. Since April 2025 the CMA has been able to fine directly, up to £300,000 or 10% of global annual turnover, whichever is higher. The rules reach anyone involved in promoting to consumers, so the paying organisation and the managing agency are exposed alongside the creator. We specify the disclosure wording in the contract, brief it explicitly, and check live posts. Labels like "gifted", "PR trip" and "affiliate" don't meet the standard, and the ASA's own advice is to lead with "Ad" or use the platform's paid-partnership tool.
None of this is exotic. It's the same rigour anyone would apply to a television buy. The reason it feels unusual in creator work is that creator work arrived through the culture side of organisations rather than the media-buying side, and it inherited the wrong habits on the way in.
Why this keeps happening
Digital in political campaigns has matured almost everywhere except measurement. Targeting is sharp. Production is fast and cheap. Distribution is nearly free. The reporting layer, in a lot of organisations, is still a dashboard screenshot showing impressions.
That's why creator budgets are the first thing cut when the weather turns. Not because they don't work. Because the people spending the money usually can't prove they do, and an unprovable budget in a political environment is a liability waiting for a journalist.
The £2.4 million isn't the story. The measurement gap is.
That's the whole lesson sitting inside the £2.4 million the UK government spent on influencers, and it applies well beyond Whitehall. The money isn't the problem. Political organisations everywhere are moving budget into creator work for good reasons. The problem is that the discipline applied to every other line of a media plan gets suspended the moment a creator is involved, and then everyone acts surprised when the spend can't be defended.
Here's what happened.
Between July 2024 and July 2026, UK government departments disclosed paying at least £2.4 million to social media influencers. The figures only surfaced because Conservative MPs filed written parliamentary questions. Education spent £718,971, working with 57 creators in 2025/26 alone, on childcare, teacher recruitment and careers advice. HMRC spent £591,200 across 44 creators, and that number doesn't include agency fees. Transport, Work and Pensions, Justice and Energy Security each spent between £176,000 and £303,000.
Departments defended it on reach. These channels get to people who don't read government websites and don't watch the evening news. That's true, and it's a legitimate argument.
It's also the only argument anyone made. Go through the disclosures and the coverage and you won't find a cost per acquisition, a conversion rate, or a comparison against an audience that didn't see the campaign. Teacher applications, on-time tax filings, benefit transitions completed. All of that is measurable. None of it was measured in public.
It gets worse. The numbers don't agree with each other. A separate FOI investigation by LBC put the total at £2,119,916 since the start of 2024/25, with the Cabinet Office as the biggest spender, a department that barely features in the parliamentary answers. LBC has Transport at £387,873.83; Parliament was told £302,873.83. An earlier FOI exercise by the PR firm Tangerine, reported in December 2025, found just over £500,000 since the start of 2024 across 215 creators. Different windows, different definitions, and departments declining to answer. The Ministry of Justice refused LBC outright on commercial confidentiality grounds, and most departments gave Tangerine the same answer. When nobody agrees what was spent, nobody is going to establish what it achieved.
So the spending isn't indefensible. It's just undefended. Which, politically, is the same thing.
What we do differently
We treat creator work as paid media with a named outcome attached, and the sequence matters more than any individual tactic.
The outcome gets defined before the brief goes out. Not after the invoice. We pick the behaviour the campaign exists to change, write the metric down, and agree it with the client while the budget is still theoretical. Registrations. Attendance at an event. Petition signatures. Applications submitted. If the only number available at the end is views, then what got commissioned was reporting, not results, and the campaign was set up to fail a scrutiny it was always going to face.
We hold back a control audience. This is the cheapest thing on the list and the one most often skipped. Suppress the creator campaign in a comparable region or segment, then compare movement. Without it, every claim of impact is really a claim about correlation, and any competent opponent will say so. Commercial advertisers have done this for decades. Political and public sector work mostly hasn't caught up.
We price against outcomes, not follower counts. A creator with 40,000 followers inside a real community routinely beats a 400,000-follower generalist on anything requiring the audience to actually do something rather than watch. Follower count prices the person. It doesn't price the result. First campaign sets the baseline, second campaign gets renegotiated against what the first one delivered.
And we fix disclosure at contract stage, not at post stage. The ASA's 2024 monitoring report, which scanned over 50,000 posts across Instagram and TikTok, found that 43% of influencer ads failed disclosure requirements, with 34% carrying no disclosure at all and a further 9% using labels too vague to make the commercial relationship clear. Since April 2025 the CMA has been able to fine directly, up to £300,000 or 10% of global annual turnover, whichever is higher. The rules reach anyone involved in promoting to consumers, so the paying organisation and the managing agency are exposed alongside the creator. We specify the disclosure wording in the contract, brief it explicitly, and check live posts. Labels like "gifted", "PR trip" and "affiliate" don't meet the standard, and the ASA's own advice is to lead with "Ad" or use the platform's paid-partnership tool.
None of this is exotic. It's the same rigour anyone would apply to a television buy. The reason it feels unusual in creator work is that creator work arrived through the culture side of organisations rather than the media-buying side, and it inherited the wrong habits on the way in.
Why this keeps happening
Digital in political campaigns has matured almost everywhere except measurement. Targeting is sharp. Production is fast and cheap. Distribution is nearly free. The reporting layer, in a lot of organisations, is still a dashboard screenshot showing impressions.
That's why creator budgets are the first thing cut when the weather turns. Not because they don't work. Because the people spending the money usually can't prove they do, and an unprovable budget in a political environment is a liability waiting for a journalist.
The £2.4 million isn't the story. The measurement gap is.