Ash Mann is a strategist, advisor and facilitator who has spent two decades working in and with cultural organisations on digital, strategy and change. His new report, The Hidden Cost, draws on a survey of 187 digital practitioners in museums, galleries, theatres, archives and cultural support bodies across 15 countries, just over half of them in the UK. It sets out to measure the work that keeps digital running but never appears in a job description.

You can read the full report at hiddencost.ashmann.co

The Hidden Cost, my research project exploring hidden digital labour in the cultural sector, focuses on the essential work that underpins digital activity. That activity is often central to how cultural organisations reach and engage audiences, yet the work behind it so often goes unseen, unsupported or misunderstood.

This tallies with one of the findings of AMA’s Together We Act research, which highlighted the problem for marketing professionals of showing the value and impact of what they do. Like digital work (and indeed much of it is digital work), marketing is key for cultural organisations to understand, reach and meet audiences, yet it frequently goes unseen or unappreciated.

The Hidden Cost found that only 12% of respondents agreed that their wider organisation understands what their role involves. This lack of understanding (cited by 70% of responses as one of the main reasons why work goes hidden or unrecognised) was one of the underlying structural conditions that contribute to hidden work. The others were:

• The undefined remit of many digital roles
• The project to operations gap
• The always-more dynamic of digital practice
• Problems around capturing and sharing knowledge
• How digital careers are structured

There is also a cultural problem around how organisations perceive and respond to cost, which exacerbates much of the above.

These structural conditions result in six distinct forms of hidden work:

• Reactive firefighting (urgent fixes and addressing ad hoc requests).
• Maintenance (keeping things running, including workarounds).
• Translation and coordination (working across teams, explaining what's involved, spotting dependencies).
• Institutional memory (remembering why things are the way they are).
• Capability absorption (taking on new responsibilities and teaching yourself new skills).
• Legitimisation (explaining and justifying digital work to those who commission it).

The largest categories of work by volume were not technical; rather they were the relational labour required to work horizontally across organisations that are structured vertically. It’s a pattern I’m sure many marketing professionals will recognise.

The time that this hidden labour takes up is not inconsequential. The most commonly reported hidden workload was between four and six hours per week (reported by 38% of respondents), although 27% of responses reported a higher hidden workload than this (with 5% of responses reporting a hidden workload of more than 15 hours a week).

The consequences of this workload

The impact of this work is felt acutely by the people who have to do it.

While 72% of the people I heard from said they find the work meaningful, 72% also said they feel worn out at the end of the working day; 58% said they feel emotionally drained; and 56% struggle to recover their energy between days.

Perhaps most damningly, 62% of the survey respondents had considered leaving their role in the last 12 months, and that intent to leave was highest amongst those with the largest hidden workloads.

The impact isn’t just on individuals, with 65% of responses noting that work outside their planned responsibilities had caused project delays in the last six months and 67% saying it had led to planned work being dropped or reduced.

Doing digital anyway

While this work primarily focused on digital practitioners, about 12% of the responses said that their role does not formally include digital responsibilities. And yet here they are, in a survey about hidden digital work, because they are doing it anyway.

Of all the groups in this research, they report the highest mean hidden hours per week. They are the most likely to say the volume of that work has increased over the past year, and nearly three quarters of them said it spikes beyond manageable most weeks.

When asked what would happen if they stopped, the majority said important work would be delayed or services would get noticeably worse. They are not doing optional work; it is necessary and consequential. It just isn't in anyone's plan, or their contract, or their job description, because it was never supposed to be their work to do.

We also heard about a small number of organisations where there are no formal digital roles at all. Every respondent from those organisations said their workload felt unsustainable, and they had the lowest adequacy scores in the entire dataset, which means that almost nobody in that group felt they had enough time for the work they were actually doing. This suggests that digital labour doesn't disappear just because nobody has been formally asked or hired to own it. It gets picked up by whoever is willing, capable, or simply nearby.

In other words, digital work is unavoidable, even for organisations that choose not to prioritise it, and the only real question is whether it is done by people with the resource, the remit, and the recognition to do it well, or by people who have none of those things and are doing it anyway.

The impact of AI

The findings of Together We Act suggest that many marketing professionals are already paying a hidden cost as AI tools arrive in organisations without shared policies or guidance. The Together We Act report describes a sector where use is already happening but is uneven: some colleagues are using tools to save time, others are opposed on ethical grounds, and in many organisations there is no agreed stance to guide usage.

The impact of this ad hoc adoption is already being felt. One respondent in Together We Act described the heavy copyediting responsibilities that have emerged around content that other teams produced with AI, and having to play detective about whether an image has been generated or touched up with AI. No one formally assigned that work and it appears in no job description or agreed workflow, yet it is what protects quality and tone of voice. Other respondents said that they pay for premium accounts themselves because their organisation does not, so personal money becomes the price of organisational dysfunction. Managers described having to both rein in enthusiastic users while also holding back their own use because of colleagues' ethical objections.

Implementing an AI tool is not a neutral act when it comes to workload. It generates its own coordination and translation costs because someone has to evaluate what the tool actually does, bridge the gap between its outputs and the organisation’s real needs, manage colleagues’ expectations about what it can and cannot do, and absorb the consequences when it doesn’t perform as promised.

In organisations with well-understood digital infrastructure and realistic pictures of their own capacity, those costs can be planned for. In organisations where significant digital labour is already invisible and unresourced, they cannot.

The risk is that AI adoption decisions (as with many digital decisions) are typically made at a distance from where those coordination costs land. A senior leader with a delivery-oriented view of digital might commission a tool expecting it to reduce workload. The translation labour that the adoption of the tool generates (the bridging, the adjustment, the expectation management) will fall to the people already carrying a hidden load. Decisions like this and their consequences are separated by exactly the understanding gap that The Hidden Cost documents.

The most useful way to think about AI in this context is as an amplifier. Introduced into an organisation with well-understood digital work, stable foundations, and the capacity to iterate, it can scale those conditions productively. Introduced into one where the work is already poorly understood and the people doing it are already absorbing more than they can sustain, it will scale those conditions too, producing no efficiency gains but a faster, larger version of the existing problem.

What all of this means for you

Not everyone reading this works in a digital role, and I'd guess many of you don't. However, I think the findings still apply to marketing in at least three ways, depending on where you sit.

If you are doing digital anyway. If you run the website, the email platform, the CRM or the social platforms alongside a marketing job, you may be part of the group described above, which reported the heaviest hidden workload in the research.  I hope the language, evidence and frameworks in the report give you things to point at and use when trying to make your actual workload more visible.

If you are the one asking. A good deal of hidden work comes from requests that look small from the outside, such as a launch date that gets set before the website has been scoped, a "quick" change to the homepage, or an urgent fix before an email goes out. In the language of the research, these feed a steady supply of reactive firefighting. Most requests of this kind seem reasonable when they are made: a deadline may be set by a funder or a season, and a change can feel small in the marketing team's terms. The person asking often has no view of what is actually involved in responding, and the person receiving it may have little room to say so. If you have digital colleagues, this research shows that it’s very likely that they are dealing with a hidden workload that is having a real impact on them. I hope the report gives you some sense of what that workload involves and why it exists, and there may be small changes to how you ask that make a real difference.

If you rely on it. Every campaign sits on top of maintenance, workarounds and institutional memory that someone is absorbing. You may never see it, because that is how it looks when things are going well. I don’t think it’s realistic to expect the understanding gap I described earlier to be closed entirely, but I do think it’s reasonable to be more thoughtful about what is required to sustain the infrastructure on which your work relies.

What needs to change

Much of what I describe in the final section of the report about what needs to change has been recommended in some form to at least some parts of the sector before. The second One by One report, published in 2019 and focused on digital work in the UK museums sector, made several of the same points: that responses had to be led by individual need rather than technology, that job descriptions and performance processes should be updated to reflect digital work, that change depends on the organisational conditions leaders create, and that practitioners need the standing and language to make the case upwards.

Those recommendations were made clearly and carefully to the sector seven years ago and, the evidence from my work suggests, mostly not acted on. That is not a criticism of that work; rather it points to the fact that the barrier is not (and perhaps has never been) a shortage of good recommendations, but the decision and ability to act on them.

The patterns that this research documents in 2026 are consistent with all of this earlier work. They have not been resolved in the intervening years, but that is the predictable outcome of a system that has no structural incentive to resolve them, and a workforce that has absorbed the cost quietly enough that the urgency to address them has never quite been registered where the decisions are made.

I think the time has come for that to change, and The Hidden Cost is an attempt to provide the evidence to support that shift.

If you do this work, I hope the report gives you something solid and credible to point at. The exhaustion you feel at the end of the week, the sense that your job has somehow become three jobs, the suspicion that you are the only thing standing between a system and its failure – these are not signs that you are managing badly.

They are what these conditions produce, in almost everyone, almost everywhere. You did not create them, and you cannot be expected to fix them on your own.

And if you hold some authority over how digital work gets resourced, planned, or understood, my hope is different. Much of what this work describes is probably happening somewhere in your organisation, and a good deal of it is likely holding up something that you rely on without your being able to see it.

None of that is a failing on anyone’s part; it’s hard to resource or protect work you cannot see, which is why so much of the report is an attempt to help make that work visible. What happens next isn’t really a digital question. It is a matter of whether the work, once seen, is treated as something that warrants action.

Ash Mann

Email Ash 

Website