Change capacity: two numbers transformation leaders need

Capacity to deliver and capacity to receive: the two change capacity types every transformation leader needs

Nov 15, 2026 | Change Measurement

Change capacity is two numbers, not one

Ask “do we have capacity for this?” in an investment committee and listen to who answers. The portfolio office answers with a resource plan: project managers, architects, business analysts, change leads. The COO hears a different question, which is whether the operation can take this in the second quarter on top of everything else. Both readings are legitimate. In most organisations only the first has a number attached to it.

Capacity to deliver

Capacity to deliver is the supply of people and governance attention needed to design, build and land change. It covers programme and project management, business analysis, change practitioners, learning design and training, testing, data migration, communications, and the subject matter experts and super users seconded from the line. It also covers the attention of steering committees and the executives who chair them, which is finite and almost never counted.

Three properties matter. It is measured in person-days by role family and month. It is largely pooled, because a change lead or a business analyst can be moved from one programme to another. And it is partly purchasable, through contractors and integrators, although with lag and with limits that are getting tighter. PMI’s 2025 Global Project Management Talent Gap report estimates that up to 30 million additional project professionals will be needed worldwide by 2035, with demand for project talent growing by as much as 64% over the decade. Buying your way out of a delivery shortfall is getting slower.

Capacity to receive

Capacity to receive is the room that a specific group of people has, in a specific period, to take in change while still doing the job they are paid for. It is made up of the hours per person per week available for training, briefings, testing and dual running, the performance headroom to work more slowly while new processes are learned, and the attention of the leaders who have to explain, coach and correct.

Its properties are the reverse of the delivery side. It is local: spare room in finance cannot be lent to the contact centre. It cannot be bought quickly, because backfill takes months to become competent in most regulated operations. It is seasonal, tied to trading peaks, year-end, renewal cycles and event seasons. Our guide to building a change capacity model sets out the dimensions that make up receiving capacity, so this article does not repeat them. The subject here is what happens when you read that number next to its delivery-side twin.

Why a single score misleads

The two capacities differ on four things that a composite index averages away:

Owner. Delivery capacity sits with the transformation office, the CIO and the head of change. Receiving capacity sits with the COO and the heads of each business unit.
Unit of measure. Delivery is counted in person-days by skill. Receiving is counted in hours per person per week by impacted group.
Timing. Delivery effort for an initiative peaks in design and build, often two or three quarters before go-live. Receiving load peaks at go-live and in the six to twelve weeks after it. The same initiative draws on the two capacities in different quarters.
Elasticity. Delivery capacity is pooled and partly purchasable. Receiving capacity is local and close to fixed in the short term.

An organisation that reports one “change capacity” figure is usually reporting one of these and implying the other.

Why the two capacities fail independently

When delivery capacity runs out first

Delivery-side failures would impact initiatives, because they leave a trail of missed milestones and cost variations. A recent example is the UK National Audit Office’s November 2025 report on National Savings and Investments’ business transformation programme, which found that the government savings bank “underestimated the scale of this challenge and overestimated its ability to deliver the Programme, despite warning signs about its lack of skills and capability”. The constraint was the ability to run a programme of that type, not the readiness of the people receiving it.

Private-sector evidence points the same way. Bain’s 2024 research on transformations reported that only about 12% achieve their original ambition, and that those that succeed avoid overloading top talent. In that study, 76% of respondents from organisations with a strong transformation record said they understood which roles were mission-critical, against 58% from poor performers.

There is also a planning error underneath many delivery shortfalls. In a 2012 Harvard Business Review article that remains the clearest treatment of the subject, Stefan Thomke and Donald Reinertsen reported that the average product development manager in their surveys kept capacity utilisation above 98%, and explained why high utilisation of people doing variable work creates queues and lengthens every project that has to wait in them. Transformation delivery is variable work. A delivery pool planned at full utilisation will be late even when every individual in it is performing well.

The less visible symptom is what gets dropped when delivery people are short. It is rarely the build. It is the change work: the impact assessment that is done once and never refreshed, the training that is compressed from three sessions to one, the readiness work completed by a project manager from a template. That impacts on what is felt by the receiving side.

When receiving capacity runs out first

This failure is harder to see because every delivery metric stays green. Releases ship on time. Training completion sits in the nineties. Then usage is thin, workarounds persist, and handling times and error rates stay elevated for longer than the business case assumed. The evidence is in operational data, which programme governance does not usually review.

The background level of receiving capacity is lower than most plans assume. Microsoft’s 2025 Work Trend Index, based on a survey of 31,000 workers in 31 countries, describes a capacity gap in which 53% of leaders say productivity must increase while 80% of the workforce say they lack the time or energy to do their work. Deloitte’s 2025 Global Human Capital Trends research found that respondents spend 41% of their day on work that does not contribute to the value their organisation creates, and that only 22% rate their organisation as highly effective at simplifying work. Neither study is about change programmes. Both describe the starting position of the people your portfolio will land on.

Rose Hollister and Michael Watkins, writing in Harvard Business Review on initiative overload, named “impact blindness” as a root cause: with no mechanism to identify and measure the demands of active initiatives, senior leaders are unaware of the load they have created. In our experience that mechanism is still missing in most large organisations.

The people who are counted on both sides

The two capacities are separate numbers, but they are connected through specific people. Suppose the most experienced assessor in a claims team is seconded to a platform programme at 60%. Delivery capacity rises by 0.6 of a person. The receiving capacity of that team falls by more than that, because the same person was the one colleagues asked for help, the one who checked difficult files and the one who brought new starters up to speed. Super users, team leaders and process owners are in the same position, and the load on people leaders deserves separate measurement in its own right.

The coupling runs in both directions over time:

A delivery shortfall becomes receiving load. Compressed training, thin support and defects at go-live all raise the effort a business unit must spend to absorb the change.
A receiving shortfall becomes delivery load. Slow adoption extends hypercare, triggers rework and second rollouts, and holds delivery staff who were planned onto the next initiative.

Fixing only the visible side therefore tends to shift the shortfall to the other side a quarter or two later.

The four combinations of change capacity and what to do in each

Put the two numbers on two axes and you get four positions, assessed per business unit and per quarter. An enterprise is never in one quadrant. Its contact centre may be receiving-bound in the third quarter while its branch network has room and the delivery pool that serves both is overcommitted.

	Capacity to receive is sufficient	Capacity to receive is constrained
Capacity to deliver is sufficient	Room to move	Receiving-bound
Capacity to deliver is constrained	Delivery-bound	Overcommitted

Room to move

Both sides have headroom, which is rarer than executives assume and usually temporary. Use it on purpose: pull forward a change the unit will need anyway, or protect the period as recovery time after a heavy release. Check the two quarters ahead before committing, because delivery effort booked now becomes receiving load later.

Delivery-bound

The business unit has room and is waiting, and the delivery pool cannot serve it. The cost here is decay: readiness built through early engagement fades, leaders who cleared time lose patience, and the window closes as the unit’s peak season approaches. Useful responses are to start fewer initiatives and finish them faster, to fund specific role families rather than general headcount, to keep mission-critical roles from being spread across five programmes, and to let capable business units run lower-risk changes themselves with a light support model. The common mistake is adding more work to the pipeline because the business is asking for it.

Receiving-bound

Delivery is staffed and on schedule, and the impacted group cannot absorb what is coming. This is the most dangerous position because programme reporting shows nothing wrong. The standard reflex, adding change managers, does not help, since it increases the push on a group that has no room to respond. What helps is changing the demand or the local supply: resequence go-lives for that group, split a release so the heaviest process change lands outside the peak, combine training from separate initiatives into one session where it makes sense, take other discretionary work off the unit for the period, and give its line leaders practical support for the weeks when they carry most of the explaining. Spotting which initiatives stack up on the same group in the same weeks is covered in our piece on change conflict detection.

Overcommitted

Both sides are past their limits, and resequencing will not resolve it. The portfolio is larger than the organisation’s ability to build and absorb it, and the honest options are to stop, defer or reduce scope. The contribution a head of transformation can make is to bring the decision forward: show both ratios, name the three or four candidates for deferral with the consequence of each, and ask the executive committee to choose. Otherwise the choice gets made anyway, through slipped dates and partial adoption, without anyone deciding.

How to get a first honest number for each side

Neither number needs a new system to get started. Both need a consistent time grain (months are enough for a first pass), recorded ranges and a named owner.

A first delivery number

List six to eight delivery role families, for example programme and project management, business analysis, change, learning, testing, and seconded subject matter experts.
For each family, count supply by month for the next four quarters. Count named people, not approved positions. A vacancy counts as zero until someone is hired and onboarded. A secondee counts at the share of time the line manager has really released, which is often lower than the secondment letter says.
Apply a planning cap. We suggest treating around 80 to 85% of nominal availability as the usable figure. The exact percentage matters less than having one, for the reasons Thomke and Reinertsen set out.
Estimate demand from the initiative pipeline by month, using a simple effort profile by initiative size. A full resourcing model is a separate exercise. A rough profile is enough to find the pinch points.
Express the result as demand divided by usable supply, by role family and quarter.

A first receiving number

Pick the ten to fifteen impacted groups that carry most of the portfolio, defined by business unit and role.
For each group, add up the hours per person per week that each initiative will ask for, by month: training, briefings, testing, dual running and an allowance for slower work after go-live. If you have not done this before, our method for measuring change saturation covers the demand side in detail.
Ask the head of each unit one question: how many hours per person per week can your team give to change in each month without service levels moving? Record the answer, the peak periods behind it and who gave it. That figure is your first capacity line. Thresholds derived from historical performance data are better and can replace it later.
Express the result as demand divided by agreed capacity, by group and month.

Keeping both numbers honest

A first number is a forecast, and it improves only if it is checked. Record a range and a confidence level with each figure. After every major go-live, compare what the unit said it could take with what happened to its service levels, and adjust. Expect self-reported figures to be shaded: delivery leads understate spare capacity to protect their teams, and business units learn to declare themselves full. The controls for that are covered in our article on protecting the integrity of change impact data. Finally, keep the secondee rule visible: anyone counted as delivery supply must be subtracted from their home team’s receiving capacity.

Presenting both change capacity numbers to executives

The purpose of showing two sets of numbers is to make the binding constraint obvious, because the remedy depends on which one it is. A page that works has three parts.

First, a delivery panel: demand against usable supply for each pooled role family, by quarter, for the next four quarters. Second, a receiving panel: demand against agreed capacity for each major business unit, by month. Third, and most important, a short list of the places where either ratio exceeds 100%, each labelled as delivery-bound, receiving-bound or overcommitted, with two or three options and their consequences. That third part is the so what, and it turns the page from information into a request for a decision. We have written separately about framing change data at decision level in executive language.

A hypothetical example: a regional bank with two different capacity problems

Consider a hypothetical regional bank preparing its third-quarter portfolio review. The delivery panel shows business analysts and change leads at 118% of usable supply, and trainers at 70%. The receiving panel shows the branch network at 65% of agreed capacity, finance operations at 90%, and the contact centre at 150% for six weeks, because a customer relationship system release, a regulatory change to complaints handling and a new knowledge tool all land between weeks five and ten.

A single amber rating would have merged two different problems that call for different decisions. The contact centre is receiving-bound: more change managers would achieve nothing, and the real choice is between moving the knowledge tool to the fourth quarter and splitting the system release. The branch initiative, which is waiting for analysts, is delivery-bound: the branches have room, so the choice is between funding two contract analysts for a quarter and letting the network run the change with lighter central support. The executive committee is asked for two specific decisions, each with an owner.

Three presentation traps

Averaging across the enterprise. An organisation-wide figure of around 80% can contain one unit at 150% and three at 60%.
Letting the change function own both numbers. The transformation office should own the delivery figure and unit heads should own their receiving figure. The change function is steward of the method and the page.
Showing capacity without options. Executives who are shown a red cell and no choices will approve the overlap and move on.

How The Change Compass shows delivery and receiving capacity together

Spreadsheets work for a first pass and become hard to sustain once the portfolio passes a few dozen initiatives, because every date change alters both pictures. The Change Compass was built around the two sets of data this article describes.

On the receiving side, Change Portfolio plots the load each role and business unit is due to absorb, week by week, as a heatmap and a timeline, and draws a capacity line for each group. A receiving-bound group shows up as load above its line before go-live, and the initiatives colliding on it are identified, so the sequencing conversation starts from evidence. Change Driver also works on the receiving side, by coaching the business leaders who have to land the change with their teams. On the delivery side, Change Automator gives hours back to a delivery-bound change team. A scope document goes in, a first impact assessment comes back as a draft, and the practitioner reviews and approves it, so time moves from typing to judgement. Because the tools share one dataset, the two capacities are read from the same record and do not have to be reconciled by hand. The platform does not replace your methodology or your rostering tool. To see both sides on one screen, join the weekly demo or book a demo.

Start with the side you cannot see

Most organisations already have a rough delivery number, because resource plans exist and delivery shortfalls are noisy. Far fewer have any receiving number, and that is usually where the unmeasured risk sits. If that describes your organisation, start there: ten groups, one question to each unit head, one month of effort. If you already have a mature view of business load and a delivery pool nobody has counted, start with the role families and the planning cap. Either way, put the second number next to the first as soon as you have it, and label each pinch point by the constraint that is binding. Change capacity reported as two numbers gives executives something they can act on: fund delivery here, resequence there, cut the portfolio where both have run out.

Frequently asked questions

What is change capacity?

Change capacity is the amount of change an organisation can carry in a given period without delivery slipping or operational performance degrading. It has two parts: the capacity to deliver change, which is the supply of programme, change and specialist effort, and the capacity to receive change, which is the room impacted teams have to absorb it.

What is the difference between capacity to deliver and capacity to receive change?

Capacity to deliver is pooled, measured in person-days by role, and partly purchasable through contractors. Capacity to receive is local to each business unit, measured in hours per person per week, seasonal, and close to fixed in the short term.

Why not combine both into a single change capacity score?

A combined score hides which constraint is binding, and the remedies are opposite. A delivery shortfall is addressed by funding roles, protecting critical people or starting fewer initiatives. A receiving shortfall is addressed by resequencing, splitting releases or removing other work from the unit.

How accurate does a first capacity number need to be?

Accurate enough to separate groups at 60% from groups at 140%. A figure agreed with the unit head, recorded as a range with its source, is enough to start making sequencing decisions. Accuracy improves when you compare each forecast with service levels after go-live and adjust.

Who should own each capacity number?

The transformation office or portfolio office should own the delivery number, since it controls the delivery pool. The COO and unit heads should own the receiving numbers for their areas, since they are accountable for operational performance. The change function maintains the method and the executive page that brings both together.