Our Guide: How to Build the Right Delivery Model for Your Agile Project

Paul Thomas • July 14, 2026

Share this article

Most organisations don't fail at delivery because their project managers are poor. They fail because they're delivering the wrong things, in the wrong order, for reasons nobody can quite remember.


Walk into any large UK organisation and ask a simple question: "Which strategic outcome does this project serve?" You'll get one of three answers. A confident, specific response. A vague gesture towards "digital transformation" or "operational efficiency". Or an awkward silence, followed by "it was approved before my time."

If you're a CIO or operational director, the second and third answers should worry you far more than any red RAG status. A late project can be recovered. A project that shouldn't exist at all is pure waste, of budget, of scarce delivery capacity, and of the organisation's finite appetite for change.


This guide sets out a practical approach to closing the gap between strategy and delivery: how to make sure every project in your portfolio earns its place, and how to sequence delivery so the right things land at the right time.


Why the gap opens up in the first place


Strategy and delivery drift apart for predictable reasons, and it's worth naming them because most organisations exhibit at least two or three.


  • Strategy is written for the board, not the portfolio. Many strategies are compelling documents that describe ambition beautifully but offer nothing a portfolio manager can actually use. "Become the most customer-centric provider in our sector" is a fine aspiration, but it doesn't tell anyone which of the forty projects competing for next year's budget should win.


  • Projects are approved on momentum, not merit. Business cases get written to justify decisions already made. A senior sponsor wants it, a supplier has proposed it, or it's simply next in a queue that formed years ago. The investment committee becomes a rubber stamp rather than a filter.


  • Nobody owns the middle layer. The board owns strategy. Project managers own delivery. The translation layer between the two—where strategic outcomes get decomposed into a coherent, sequenced portfolio — is often nobody's explicit job. In organisations with a strong PMO it might live there; in many it simply doesn't exist.


  • The portfolio never gets pruned. Projects are far easier to start than to stop. Strategies change, markets shift, and priorities move, but the portfolio carries on regardless, accumulating initiatives like sediment. The result is the familiar "zombie project": still funded, still consuming people, long since detached from any current strategic intent.


None of this is a failure of intelligence or intent. It's a failure of mechanism. The fix is to build the connective tissue deliberately.


Start with outcomes, not projects


The single most useful discipline you can introduce is this: no project exists until a strategic outcome demands it.


That means doing the translation work properly. Take each strategic objective and decompose it into a small number of measurable outcomes — the specific, observable changes that would tell you the objective is being achieved. If the strategy says "grow our commercial energy business", the outcomes might be "reduce customer onboarding from six weeks to five days", "launch flexibility products to 200 industrial customers by 2027", or "cut cost-to-serve by 15%".


Outcomes have two properties that make them the right unit of currency. They're measurable, so you can tell whether delivery is actually moving the dial. And they're stable enough to plan against — outcomes typically survive for two or three years even as individual projects come and go beneath them.

Once the outcomes are defined, the question for every project — existing or proposed — becomes brutally simple: which outcome does this move, by how much, and by when? If the answer is unconvincing, the project doesn't belong in the portfolio, however elegant the business case or however senior the sponsor.


This is what's often called the golden thread: a traceable line from board-level strategy, through defined outcomes, down to individual projects and the benefits they deliver. When the thread exists, prioritisation decisions become debates about evidence rather than contests of seniority. When it doesn't, the loudest voice wins.


Selecting the right projects



With outcomes as your anchor, project selection becomes a filtering exercise rather than a beauty parade. A few principles make it work in practice.


  • Score everything against the same criteria. Build a simple scoring model: strategic contribution, financial return, risk, delivery confidence, and dependency on other work. Keep it light, five criteria scored one to five is plenty. The point isn't false precision; it's forcing every proposal through the same gate and making trade-offs visible. A project scoring highly on strategic contribution but poorly on delivery confidence prompts a useful conversation. A project scoring poorly on everything prompts a shorter one.


  • Make "no" a normal answer. An investment committee that approves 95% of what it sees isn't governing; it's processing. Healthy portfolios reject or defer a meaningful proportion of proposals, and they do it early — before business cases consume weeks of effort and sponsors become emotionally invested. A two-page strategic outline, assessed against the outcome framework, should be the first gate. Most bad ideas can be killed there for the cost of an hour's discussion.


  • Review the existing portfolio with the same rigour as new proposals. Run every in-flight project through the same filter at least twice a year. Some will have drifted from their original intent. Some will be serving an outcome the strategy no longer prioritises. Stopping a project mid-flight feels expensive — sunk costs, awkward conversations, disappointed teams, but continuing to fund work that no longer matters is more expensive still. The organisations that do this well treat project closure as a sign of portfolio health, not delivery failure, and they say so publicly.


  • Watch for the strategy-shaped hole. Filtering isn't only about removing weak projects; it's about spotting what's missing. Map your portfolio against your outcomes and you'll almost certainly find outcomes with little or nothing delivering against them, usually the harder, less glamorous ones. An honest gap analysis is worth more than another status report.


Delivering at the right time


Choosing the right projects is half the job. Sequencing them is the other half, and it's the half most organisations neglect.


  • Sequence by dependency, not by enthusiasm. Projects rarely stand alone. A customer analytics initiative depends on the data platform being in place; a new operating model depends on the systems consolidation finishing first. Map these dependencies explicitly and let them drive the roadmap. It's common to find organisations attempting phase-three work on phase-one foundations because the exciting project got funded before the enabling one.


  • Respect the organisation's capacity to absorb change. This is the constraint that delivery plans most often ignore. Your operational teams can only take so much change at once, new systems, new processes, new structures, before performance suffers and adoption fails. If three major initiatives all land on the same contact centre in the same quarter, at least one of them will fail, regardless of how well each was delivered. Build a change-impact view across the portfolio: who is affected, when, and how heavily. Then sequence accordingly, even if it means slowing something down.


  • Match ambition to real delivery capacity. Most portfolios are planned as if the organisation has 30% more capable people than it actually does. Architects, product owners, change managers, and experienced delivery leads are always the bottleneck, and spreading them thinly across too many initiatives guarantees that everything moves slowly. Fewer projects, properly staffed, will deliver more strategic value per year than a sprawling portfolio of half-resourced ones. This is uncomfortable arithmetic for stakeholders whose projects get deferred, but it's arithmetic all the same.


  • Deliver value in slices, not monuments. Long projects are where strategic alignment goes to die. A three-year programme approved against 2024's strategy will be delivering against a different world by 2027. Break delivery into increments that produce measurable outcome movement every three to six months. This does two things: it gets value flowing earlier, and it creates natural decision points where the organisation can redirect, accelerate, or stop based on what it's learned.


Governance that asks strategic questions


Most project governance is delivery governance: is it on time, on budget, are the risks managed? Necessary, but nowhere near sufficient. Strategic alignment needs its own questions asked at every review:

  • Is the outcome this project serves still a priority?
  • Is the project still the best way to move that outcome?
  • What has the last increment actually delivered, in outcome terms rather than milestone terms?
  • If we were making this investment decision today, from scratch, would we?


That last question is the most powerful and the least asked. It cuts through sunk-cost reasoning and forces a fresh look at whether the project still earns its place. Boards that ask it regularly run smaller, sharper portfolios.

It also matters who's in the room. Strategic portfolio reviews need the people who own the outcomes — typically your executive peers, not just delivery leadership. When operational directors sit alongside the CIO in prioritisation decisions, alignment stops being an IT problem and becomes what it always should have been: a whole-business discipline.


Measure outcomes, not outputs


Finally, close the loop. Delivering a system is an output. The 15% reduction in cost-to-serve it was supposed to enable is the outcome, and it usually arrives months after go-live, if it arrives at all.


Track benefits against the original outcome commitments, with named owners, well beyond project closure. Publish the results, including the misses. Nothing sharpens future business cases like the knowledge that someone will check whether the promised benefits actually materialised. Over time this builds something genuinely valuable: an evidence base showing which types of investment actually move your strategic outcomes, and which merely promised to.


Where to start


If your portfolio has drifted, and most have, resist the temptation to redesign everything at once. Three moves deliver most of the value:

  1. Define your strategic outcomes properly, with measures and owners. This is a matter of weeks, not months, and everything else depends on it.
  2. Run the current portfolio through the filter. Every project answers the question: which outcome, how much, by when? Expect to stop or reshape 20–30% of what you find. That capacity is your reinvestment fund.
  3. Rebuild the roadmap around dependencies and change capacity rather than approval dates. Fewer things, in the right order, properly resourced.


None of this requires new tooling or an army of consultants. It requires clarity about what the strategy actually demands, the discipline to say no, and governance willing to keep asking whether each project still deserves its place. Get that right, and delivery stops being a cost centre that occasionally disappoints the board, and becomes the mechanism by which the strategy actually happens.


Agile Minds works with UK organisations to recover struggling programmes and realign delivery portfolios with strategy. If your portfolio has drifted from your strategic intent, we should talk.


Recent Posts

Hand circling a date on a desk calendar with a pen beside a laptop
By Paul Thomas July 16, 2026
Planning a PMO setup? This guide walks through the realistic 3-6 month timeline phase by phase, plus a free assessment to find your own number.
By Paul Thomas July 15, 2026
Most consultancies won't answer the cost question directly. Ask, and you'll typically get "it depends" and a form to fill in. It does depend, but that's not an excuse to be vague about what it depends on. If you're a COO trying to build a business case for the board, three variables actually set the price, and none of them are secret. What actually drives the number Programme size is the first. A single at-risk workstream costs less to stabilise than a portfolio of twelve, scope drives everything downstream. A workstream with one team, one sponsor, and one clear deliverable is a contained diagnostic exercise. A twelve-workstream portfolio with overlapping dependencies and multiple sponsors is a different order of problem entirely, because stabilising one workstream in isolation without understanding its dependencies on the other eleven risks solving the visible symptom while the underlying cause resurfaces somewhere else three months later. Embed duration is the second: recovery engagements typically run three to six weeks, and a straightforward stabilisation, credible plan, honest reporting, a functioning team, costs less than a six-week turnaround with practitioners embedded across multiple workstreams. The difference between the two isn't really about calendar time so much as coverage, how many people need to be physically present, in the room, for how many of those weeks, before the organisation can run the plan on its own. Seniority is the third, and the one most likely to be quietly reduced to save money. We only deploy practitioners with 15+ years of experience, never graduates running a template, and that's a deliberate cost floor, it's also why the fix tends to hold. It's worth being honest about why seniority costs more and is still worth it: a junior consultant working from a template can produce a plan that looks credible on paper. A senior practitioner who has seen a dozen versions of this specific failure pattern before knows which parts of that plan won't survive contact with the actual organisation, and can say so in week one rather than week five. What that seniority actually costs, in public numbers You don't have to take a consultancy's word for what senior delivery talent costs in the UK market. The Association for Project Management's 2025 Salary and Market Trends Survey, the UK's largest annual study of the profession, puts the average project management salary at £52,500, up 10% on 2023, with practitioners in consultancy and energy/utilities sectors averaging £62,500, per the APM Salary Survey 2025 . Embedded recovery work commands more than that average because it's compressed: fifteen-plus years of pattern recognition delivered on a timeline measured in weeks, not spread across a year. There's a useful way to think about what's actually being purchased here. A permanent hire's salary buys a year of their time, spread across whatever the organisation needs that year. A recovery engagement's cost buys weeks of concentrated attention on one specific problem, from someone whose entire value is having seen that problem before. The hourly economics look expensive next to a permanent salary until the comparison is corrected for what's actually being delivered in that window. The cost that never makes it onto an invoice Every cost conversation focuses on what recovery costs. The number that's harder to see, and usually larger, is the cost of inaction. A project drifting for another quarter doesn't just slip its date, it compounds in several distinct ways. The team burns out chasing an unrealistic plan, because nobody has told them the plan is unrealistic, so they keep trying to hit dates that were never achievable in the first place. The best people quietly start looking elsewhere, because skilled practitioners can generally tell when a programme has lost its way faster than the sponsors funding it can, and they don't wait around to find out how it ends. Stakeholder trust erodes in a way that outlasts the project itself, a sponsor burned once by an over-promised, under-delivered programme brings that scepticism into the next funding conversation, and the next, long after the specific project is closed. And the eventual fix costs more because the drift has to be unwound before recovery can even begin, every extra week of a bad plan being followed is a week of decisions made on false assumptions that now need to be found and corrected. By the time a board asks how much it will cost to fix, the more useful question is how much it has already cost to wait. Why nobody publishes a rate card A three-week reporting fix and a six-week multi-team turnaround are different projects with different costs, and any headline number would be wrong for most people who read it. That's not evasion. It's the same reason a surgeon doesn't quote a price before the scan — not because the price is a secret, but because quoting it before the diagnosis would mean quoting the wrong number to almost everyone who asked. Where to start Cost climbs with the number of embedded practitioners, the complexity of stakeholder alignment, and how far a programme has already drifted before anyone calls for help. It does not climb because of methodology, tooling, or brand name. Agile Minds gives every prospective client a free, two-week assessment before any number gets discussed. If you need a figure you can actually take to the board, we should talk .
Jaguar and Land Rover logos on a gray building facade
By Paul Thomas June 1, 2026
How AgileMinds built JLR a working agile framework through a real product, not a workshop. A 16-week app and guilds that kept it alive after we left.
Halfords storefront with glass entrance, orange logo, and parked bicycles outside
By Paul Thomas June 1, 2026
How AgileMinds delivered Halfords' unified ecommerce platform on Salesforce Commerce Cloud. Two brands, one basket, over 100 integrations, nine months.
Three people in red helmets and jackets stand outdoors at dusk, looking serious.
By Bernadine Dela Cruz May 29, 2026
How AgileMinds helped E.ON take a virtual power station from proof of concept to commercialised, award-winning solution in six months.
Stressed person at desk with laptop, head in hand, reviewing charts in a bright office
May 24, 2026
Uncover the hidden costs of project team burnout, like knowledge loss & attrition. Contact us for strategies to mitigate these issues.
Team collaborating around a laptop with papers, notebooks, coffee, and a smartphone on a desk.
By Paul Thomas May 4, 2026
Most failing projects send warning signs months before the board notices. Here’s how to spot them, and how to recover before it’s too late. Free assessment.
Hands holding a tablet with a to-do checklist and red markings on a desk, lit by blue and warm light
By Paul Thomas May 4, 2026
More than half of PMOs are shut down within two years. Here are the five reasons mid-market PMOs fail, and what to do differently. Free PMO assessment.
Colleagues in a meeting, leaning over documents and pointing at a chart on a conference table.
By Paul Thomas May 4, 2026
Explore why 70% of digital transformations fail due to middle management issues. Contact us for tailored solutions today!