John McIntyre
Founder, HotPMO
Every month I trawl through the research papers, reports, and podcasts so you don’t have to. Here’s what caught my eye this month.
One of the project owners in this month’s steering committee study described how he prepared for the job: he reread the red PRINCE2 manual he had kept from a training twenty years earlier. There was no internal guidance, it was assumed he would know what to do.
These kind of assumptions are challenged across multiple dimensions in the PMO news this month. The papers and insights I’ve focused on here show examples of how the assumptions we make about how our people are working can differ from reality, how non-professional project managers need more than a template library, and what we do (and don’t) trust AI to do for us.
Happy reading!
Trust in AI falls away at the point of decision
Expleo runs a monthly tracker of what UK business leaders will and will not let AI do. The August figures are worth a look because they break the question down task-by-task.
Around three quarters of respondents were comfortable letting it summarise emails, documents and meeting notes. Roughly the same proportion let it organise calendars and book meetings. Then it drops away. Just over half will let AI make decisions within predefined rules. Only just over four in ten will let it approve a financial or budget transaction.
Where the comfort stops. Chart: HotPMO, drawn from the figures published in the Expleo AI Pulse, UK, August 2026.
What separates the tasks people are relaxed about from the ones they are not is how easily a mistake can be undone. A bad summary is spotted and rewritten in seconds. A payment approved in error has already gone. The dividing factor seems to be perceived reversibility.
Lindsay Scott puts it in PMO terms: Most of us are quite relaxed about AI drafting the highlight report. But how many would let it change a RAG status, or approve a change request?
A note of caution here: While you may want the comfort of knowing a human is making the final decision, the data and knowledge thread is also important. If the human decisions are based on the documentation and CRs that were compiled and proposed by AI, then large language models are almost certainly influencing your decision making already. This is not necessarily bad (or good), but it is something to be mindful of.
The practical step is to write your own line down. List your PMO’s regular decisions, from RAG changes and risk escalations to change approvals and gate recommendations, and mark each one AI-drafts, AI-recommends-human-signs, or human-only. Then encourage frank disclosure about the use of AI in drafting and decision making, so decision makers understand where the options they are approving came from. It takes twenty minutes, and it turns a vague policy about being careful with AI into something actionable, specific and consistent that your team can follow.
Half of micro-shifters have not told their manager
Micro-shifting is the habit of breaking the working day into blocks instead of working it straight through: an early start, a gap for the school run or an appointment, then the work picked up again later. It grew out of home working, where nobody was watching the clock in the same way, and it has outlasted the conditions that created it.
Monster surveyed 876 workers about it in April and May. Just over half say they micro-shift. Of those, nine in ten do it at least weekly and three in ten do it every day. Roughly eight in ten say it makes them more productive than a fixed schedule, and the reason they give most often is wanting more control over the shape of the day. It is a US jobs board surveying US workers on a self-selecting panel, so treat it as a direction of travel rather than a measurement of your own organisation.
The finding that should interest a PMO is a different one. Just over half of the people who micro-shift say they have done it without their manager knowing. Asked about the downsides, around a quarter said collaboration gets harder, and roughly a fifth each reported slower communication and difficulty scheduling. Lindsay asked on the stream whether project work can absorb that kind of flexibility, given how much of it is collaborative. Monster’s own respondents highlight the things that suffer: coordination challenges and a lack of workflow visibility for managers.
While some of us settle into that, the Financial Times reported this month that consulting firms are considering pulling junior staff back into the office more often. The argument is about interpersonal skills. EY’s UK head of consulting, Sayeh Ghanbari, told the FT that AI has made those skills more valuable, and that firms let training in empathy, storytelling and leadership slide during the remote-working boom while they chased technical ones. Consulting runs on an apprenticeship model, where juniors learn by watching how a senior colleague handles a client and then talking it through afterwards. If AI absorbs the analytical work, what a junior consultant offers that a strategy tool cannot is the human part, and the firms have concluded that the human part is learned in a room with other people.
I agree with that. Interpersonal skill is built in person, and I have never seen a remote substitute for watching someone handle a difficult meeting and unpicking it with them afterwards. The skills required used to be referred to as ‘soft skills’, but were redefined by PMI as Power Skills in 2023
It will meet resistance, though, and the resistance will be reasonable. Anyone with a long commute, or who has finally built a working pattern that makes the week fit together, is going to push back on being asked to give it up. Both positions are defensible, which is what makes this difficult, and these are hard trade-offs for organisations and employees alike.
If you are anywhere near this decision, the thing worth settling first is what your team needs to be together for. Name the sessions that work better in a room, protect those, and reduce clock-watching.
Steering committee roles go largely undefined, small study finds
This month’s Project Management Journal carries a new study of project steering committees. Stoppels, Bosch-Rekveldt, Mooi and Bakker interviewed twelve committee members and six project managers across six Dutch steering committees, budgets from €8m to €250m, spanning ERP, business process redesign, engineering, a plant upgrade and IT infrastructure.
It isn’t a large study, but what it does give you is a well-documented look inside a governance body most of us take for granted. The authors propose five member roles:
- Representing the funder or funders: those who decide the budget and carry any overrun back to their own constituency.
- Representing the users: the groups who will work with what the project delivers.
- Representing the suppliers: those providing the capability and resources to build it.
- Representing the support organisation for the deliverables: the people who keep the thing running afterwards, whose interest in standardisation differs from the users’.
- Quality assurance: added where other members lack the expertise to interrogate what they are told.
Three of those map onto PRINCE2 and two do not, which the authors are explicit about. They split the support organisation out of PRINCE2’s senior user on the grounds that “use” suggests putting something into service once, while the people who maintain it afterwards want standardisation and simplicity and will argue for different things. The quality assurance role is one PRINCE2 advises against, and they observed it working anyway.
Regarding the roles: everyone agreed on the funder, described by one interviewee as the member who “has skin in the game”. Past that it got vaguer. At the ERP implementation nobody was clear who represented the staff departments, and the paper records that their business impact was underestimated and partly ignored as a result. On the same project, directors from two external suppliers sat on the committee and used it for escalation and sign-off, sometimes hijacking the meetings and stopping the business case from being discussed at all. A board consultant was added as a quality advisor because members were struggling to follow the project’s status and the impact of their decisions. It helped, and it also left him making decisions inside other people’s domains.
Two structural findings sit underneath those stories. Decision authority followed a member’s position in their own organisation rather than their seat at the table, which is why one project owner needed a supplier representative’s agreement every time budget or planning moved. And in all six committees, the project manager set the agenda.
If you run a PMO in a large organisation, consider replicating the study inside your own environment. Interview a handful of your steering committee members and their project managers, separately, and ask each what they think their role is, what they are responsible for, and where they believe decisions actually get made. Then hold your answers against the paper’s. You would learn more about your governance in a fortnight than a maturity assessment will tell you. Steering committees stay oddly under-researched for something we all depend on, which is why a modest study still earns its place, and why your own evidence would be worth having alongside it.
For accidental project managers, training matters more than templates
How do PMOs best support accidental and non-professional project managers? Lejeune and Burnay surveyed 257 of them across French and Belgian hospitals – people who were running projects because of their specialised domain knowledge – not because of their project management background.
They found training strongly predicts whether people use project management tools and templates, and it is the strongest relationship in the model by some distance. But tool use on its own has no measurable direct effect on how the project performs. The effect appears only when tool use turns into working in a way that lines up with the recognised performance domains, and that alignment is what predicts operational and strategic performance.
For PMOs who hand these PMs a SharePoint site full of templates and hope for the best – this report should make you re-think. Issuing a template pack to your accidental project managers achieves very little by itself. Templates only start paying back when training has turned them into a way of working. The usual caution applies: a self-selected sample of hospital staff in two countries, self-reported performance, but the logic holds, and for me at least, it matches patterns I’ve seen in multiple organisations and sectors.
The question I would put to your PMO: how much of your support offer assumes a trained project manager, in an organisation where a good share of projects are run by people who are not one? If the honest answer is most of it, the area to focus on is training and development support for those who suddenly find themselves leading projects rather than acting as subject matter experts.
September and October: worth leaving your desk for
The House of PMO has a run of free in-person evenings this autumn, and the thread through them is data, AI, reporting and decision-making.
17 September, London: AI for PMOs: Practical Applications & Strategy, hosted by Nationwide in Holborn, 5.30pm to 8pm. I am speaking at this one, so come and disagree with me in person.
24 September, Leeds: Blue Light PMO returns for the emergency services and wider public sector, at St James’s University Hospital, 1pm to 5pm.
1 October, Edinburgh: AI in the PMO: Useful, Uncomfortable and Unavoidable?, hosted by Computershare, 5.30pm to 8pm.
22 October, Newcastle: AI and the PMO Service Catalogue, on how AI could reshape what a PMO offers and where human judgement still earns its place, at Newcastle Building Society in Wallsend, 5.30pm to 8pm.
You can watch the full session, including the news segment all of this came from, right here:
This roundup is based on my monthly PMO News segment from the PMO HotHouse, run by the House of PMO. If you’re not already part of that community, it’s well worth a look.
August’s research is fairly blunt about template libraries. Lejeune and Burnay found that training drove tool use, and that the tools themselves changed nothing measurable until that use became working practice. That is the thinking behind the PMO SuccessHub: the PMO Vault is in there, alongside an honest assessment of where your PMO actually stands, experienced PMO people to call on through the year, and a plan for building capability in the team you have. If that is the gap in yours, it is worth a look.
