One building, more than one operating logic
When the National Portrait Gallery reopened after its 2023 refurbishment, the changes were not confined to exhibition space. The Gallery had expanded and refurbished its hospitality areas, introduced new operators, improved their street presence and extended evening opening for parts of the offer. The National Audit Office says onsite hospitality revenue doubled to £5.4 million, on which the Gallery receives a concession.
The figure matters less than the operating decision behind it. The same building can provide free access to a national collection, host exhibitions, give visitors somewhere to eat and drink, attract destination diners and produce income for the institution. Those functions do not automatically conflict. Neither are they interchangeable.
For the Gallery, hospitality is one part of a wider financial model. Membership is another: its latest annual report describes membership as a strategic driver of both engagement and income. Exhibitions, fundraising and trading contribute through different mechanisms again.
The National Portrait Gallery is useful as an illustration, not a template. England’s 15 museums and galleries sponsored directly by the Department for Culture, Media & Sport have very different collections, buildings, audiences and opportunities to generate income. What they increasingly share is the importance of being able to generate some of their resources themselves.
The question is not whether a museum should have a restaurant. It is what happens to the operating logic of a public cultural institution when the ability to generate income becomes more consequential to keeping it financially viable.
This is not a museum-versus-business story
There is an easy version of this story in which museums, squeezed by public funding, have recently discovered commerce. The history does not support it.
More than two decades ago, a 2004 National Audit Office report was already examining how DCMS-sponsored museums and galleries generated income through retail, catering, venue hire, licensing, special exhibitions and fundraising. It argued for stronger income-generation skills and better understanding of profitability and risk.
So the appearance of shops, restaurants, corporate events or paid exhibitions is not evidence that museums have suddenly become businesses. Mixed funding and entrepreneurial activity have been part of the institutional model for years.
What is different after the pandemic is the weight placed on those capabilities.
Novelty is not the same as consequentiality. A capability can exist for decades and still become more important when the conditions around it change. A museum does not need to invent membership, licensing or hospitality for its dependence on the income they produce to increase.
What changed after the pandemic
The NAO’s 2026 review provides the clearest picture for the 15 DCMS-sponsored museums and galleries in England.
Their self-generated income, excluding donated assets, rose from £368 million in 2021-22 to £563 million in 2024-25 in real terms. Thirteen of the 15 increased it. Yet £563 million was essentially a recovery: only 0.1% above the annual pre-pandemic average in real terms. The significant finding is therefore not record income. It is the NAO’s conclusion that the institutions have become more reliant on self-generated sources.
That reliance sits inside a model in which government support remains substantial. Grant-in-aid averaged 46% of the institutions’ total income in 2024-25. Total grant-in-aid was above its pre-pandemic real-terms average, partly because capital funding had increased. But revenue grant-in-aid—the funding available for day-to-day operations—had fallen to £330 million and was 11% below its pre-pandemic real-terms average. Capital and revenue support solve different problems; combining them into a claim that public funding simply rose or fell obscures the change.
Visitor recovery was incomplete too. The 15 institutions recorded about 42 million visits in 2024-25, 13% below pre-pandemic levels. Costs had risen since reopening. Some institutions required additional short-term government assistance, and in August 2025 more than half told the NAO that their financial position was worse than three years earlier.
Self-generated income recovered. Financial resilience did not automatically follow.
A portfolio rather than a category
“Self-generated income” can sound like a euphemism for commercial revenue. It is not.
In the NAO’s analysis it includes donations and legacies, charitable activities, other trading activities, investment income and other income. Trading was the largest single category in 2024-25, but donations and charitable activities together represented a substantial share of the total.
Those sources demand different capabilities and create different relationships.
The V&A, for example, describes commercial activity explicitly as part of building financial resilience. Its current approach connects exhibition ticketing, differentiated pricing, membership, retail, visitor data and customer communications. The Royal Armouries has developed image licensing around its intellectual property. The Science Museum Group has created paid visitor experiences. The National Portrait Gallery has invested in hospitality.
These are different operating systems, not variations on one generic commercial strategy.
Fundraising adds another logic. Money from a donor may be substantial yet restricted to an acquisition, programme or capital project. Membership can combine financial contribution with an ongoing audience relationship. Licensing depends on the characteristics and recognition of a collection. Venue hire depends heavily on a building and its location. Paid exhibitions require suitable spaces and an audience willing to buy tickets.
The ability to use any of these mechanisms is uneven. The NAO notes that income-generating capacity varies with location, collections and facilities.
Greater reliance on self-generated income therefore does not produce one new museum business model. It makes the quality of each institution’s particular portfolio more consequential.
What the portfolio hides
That portfolio becomes harder to judge when “income” is treated as the outcome.
Consider two audited trading subsidiaries in 2025-26. The Natural History Museum Trading Company, whose activities include retail, catering, venue hire, touring exhibitions, licensing and other commercial operations, recorded turnover of £34.3 million and profit before tax of £11.9 million, with the taxable profit due to be distributed to the museum.
The National Portrait Gallery Company recorded turnover of £4.64 million and a net contribution of £476,000.
Those numbers should not be used to rank the institutions. Their activities, scale and accounting boundaries differ. Their usefulness is conceptual: turnover is not contribution.
The same distinction applies elsewhere. Gross admissions income is not trading profit. A restricted donation is not freely available operating cash. Money passing through an activity says less about resilience than what remains available after its costs and restrictions are understood.
A separate 2026 study commissioned by the Association of Independent Museums makes this management problem visible in a different cohort. Looking at 30 accredited museums, it describes organisations operating portfolios of small activities while sometimes having limited understanding of their real financial return beyond the income generated. It also stresses that there is no single viable model: buildings, collections, funding arrangements and local context shape what each organisation can do.
Those findings cannot simply be transferred to the 15 national institutions. They do, however, clarify why revenue visibility and contribution visibility are different managerial capabilities.
Nor does contribution have to mean profit alone. An exhibition, education programme or membership scheme may create public or institutional value that is not reducible to a margin. The point is not that every activity should pay for itself. It is that a museum increasingly dependent on multiple sources of income needs to know which activities generate usable financial resources, which consume them, which carry restrictions and which justify their economics through other parts of the institution’s purpose.
The strongest objection is also an important correction
There is a serious counterargument to treating any of this as a significant change in institutional design.
Museums have operated mixed funding models for decades. The 2004 NAO report already identified many of the same activities and management questions. Aggregate self-generated income in 2024-25 was not historically exceptional. Even other trading activity, at £210.8 million in real terms, remained below its 2019-20 level of £225.2 million.
Public funding has not become incidental either. Grant-in-aid remains structural to the 15 institutions. DCMS has provided additional support where pressures became acute, including short-term revenue funding in 2023-24 and 2024-25 and higher baseline support from 2025-26 for institutions judged to be at greater risk.
And there is no single commercial opportunity waiting to be exploited across the cohort. A large London institution with international visitor traffic, extensive hospitality space and a globally recognisable collection has options that a smaller specialist museum does not. More entrepreneurial capability cannot erase geography, estate constraints or the nature of the collection.
Most importantly, the available evidence does not show that greater reliance on self-generated income has changed curatorial judgement, weakened free access, improved inclusion or reduced institutional autonomy. Nor is there comparable information showing the full net contribution of each major revenue stream across all 15 institutions.
Those limits rule out a more dramatic story about museums becoming commercial institutions.
They do not rule out a change in the importance of capabilities museums already possess.
The relevant shift is not that hospitality, licensing, fundraising or paid exhibitions exist. It is that when institutions are more reliant on the resources they generate themselves, the performance and governance of those established activities matter more to the durability of the organisation.
That is a narrower claim. It is also the more consequential one.
Resilience is not the same as diversification
A broader revenue portfolio can look like evidence of resilience. The NAO uses a tougher definition: financial resilience means being able to prevent, adapt and respond to disruption and absorb shocks without widespread damage to long-term finances, service delivery or the achievement of objectives.
By that standard, own-income growth is not enough.
Different sources can diversify dependence, but they also expose an institution to different conditions. Visitor spending can move with tourism and the wider economy. Exchange rates and travel costs affect overseas visitors. Membership income can suffer from churn. Sponsorship and philanthropy depend on donors and partners. Major paid exhibitions can produce meaningful income but are also volatile.
That does not mean diversified income necessarily makes an institution more volatile. Reducing reliance on a single source may itself be valuable. The available evidence does not establish which effect dominates across the 15.
It establishes something more useful: generating more income changes the map of dependencies; it does not eliminate dependency.
An institution can remain dependent on government while becoming more exposed to visitors, members, donors, sponsors and commercial partners at the same time. Resilience therefore depends on the shape and governability of the portfolio, not merely on the percentage labelled self-generated.
The governance question
Once the issue is framed this way, “public versus commercial” becomes a poor guide to decision-making.
A more useful set of questions starts inside each mechanism.
What does the activity actually contribute after its direct costs? Is the money unrestricted enough to support core operations? What fixed or management costs sit behind it? How exposed is it to visitor behaviour, tourism, donors or partners? What capabilities does the institution need to operate it well? And what part of the institution’s public purpose does it help sustain or deliver?
Those questions do not require the museum shop to justify a collection, or an education programme to maximise profit. They require financial contribution and public contribution to remain legible rather than being collapsed into a single income target.
That distinction matters because an income-generating capability can become more consequential without its wider cultural effects being known. If exhibition ticketing, membership, hospitality, fundraising or licensing becomes financially more important, boards and executives need to understand what those activities optimise and what dependencies they introduce. That is a governance requirement, not evidence that the resulting incentives have already altered curatorial decisions.
Institutional variation strengthens rather than weakens that principle. The right portfolio for the Natural History Museum need not resemble the right portfolio for the National Portrait Gallery—or for any of the other 13 institutions. What can remain consistent is the standard by which the portfolio is understood.
Govern contribution, not income growth
GOVERN THE PORTFOLIO BY CONTRIBUTION TO RESILIENCE AND PUBLIC PURPOSE, NOT BY INCOME GROWTH ALONE.
As self-generated income becomes more consequential, the useful question is not simply how much income an institution can generate, but what each funding and income mechanism contributes after costs and restrictions, what dependencies and volatility it introduces, and whether the resulting portfolio improves the institution’s capacity to sustain its public purpose.
That is different from becoming “more commercial.” It is also different from assuming public funding can make operating questions disappear.
A museum can earn more and remain financially fragile. A trading activity can generate substantial turnover but only part of that becomes usable contribution. A donation can be valuable yet restricted. An activity that produces little profit can still deserve resources because of what it contributes to access, education, research or the collection.
The harder task is to see those differences clearly enough to govern them.
For England’s 15 DCMS-sponsored museums and galleries, the increasing importance of self-generated income does not point towards one ideal revenue mix. Their assets, audiences and obligations are too different for that. It points instead towards a more demanding definition of financial competence: knowing not only how to generate income, but what each source makes possible, what it costs, what dependence comes with it and whether the institution is better able to continue doing what it exists to do.
Earning more is an activity. Resilience is a condition. The work between them is governance.