# How should a lab split one budget across programmes and projects?

> Funds, programmes and projects, and which way round to split them. One small biotech's year: two funds, three projects and £1,577.20 nobody could spend.

- Source: CCA Bio (Cell Culture Automation Ltd), Cambridge, UK
- Canonical URL: https://cca.bio/articles/research-program-budgets
- Published: 2026-11-17
- Topic: Lab finance
- Tags: lab budget, research programmes, grant management, lab finance, CCA Labs guide
- Series: One programme's year, from the budget split to the August review (overview)

## Short answer

Give each source of money its own fund, group work into programmes by what it is for, and allocate funds to projects rather than to programmes. A programme's budget is then the sum of its projects' allocations. Keep part of each fund unallocated as contingency, and when a project closes, hand back what it did not spend.

At the end of August our example lab finds £1,577.20 sitting on a project that finished in May. Nobody had spent it. Nobody could, either, because a closed project was still holding it.

That's a small sum, and a common way for a lab budget to quietly lose its accuracy. Splitting money across programmes and projects is really about who can see what: which work is drawing on which grant, and what's left for decisions nobody has made yet. I think most labs split it the wrong way round. Below is the way I'd do it, then one biotech's year of it, then where it breaks.

## What are funds, programmes and projects?

A fund is where money comes from, a programme is why the work is done, and a project is the piece of work that spends it. Three layers, and each answers a different question.

- A **fund** is one source of money with a fixed total and usually a term: a grant, an internal R&D budget, a customer contract.
- A **programme** is a line of work with a purpose and goals that outlive any one project, such as "get our cell line into a bioreactor".
- A **project** has an owner, a start and an end date, and an **allocation** from one fund. Orders and experiments are charged to it.

In engineering terms it's a power budget. The supply is fixed. Each rail gets an allocation, the rails add up to less than the supply, and the difference is headroom you keep on purpose.

## Why allocate money to projects and not to programmes?

Because a project has an owner and an end date, and a programme has neither in the same way. Money parked on a programme is a pot that everyone in it can draw on and no one person answers for.

So the programme doesn't hold money. Its budget is the sum of its projects' allocations, which CCA Labs calls the programme's **Draw**. Add a project and the programme grows; close one and hand the money back and it shrinks. Nothing is typed in twice, so the programme figure can't drift away from the projects underneath it.

## A worked example: one programme, two funds, three projects

The example lab is a small biotech. Its HEK293 cells grow stuck to plastic in medium with 10% fetal bovine serum, and it wants them growing in suspension, without serum. A stirred tank scales and a stack of flasks doesn't, and serum is dear and varies from lot to lot. On 2 March the head of R&D sets up a programme for it, **HEK293 Suspension**.

Two funds pay for it:

| Fund | Total | Term | Charges bench time |
| --- | --- | --- | --- |
| Process R&D 2026 | `£10,000` | calendar 2026 | No, salaries come from payroll |
| Process Development Grant | `£20,000` | 1 Jun 2026 to 28 Feb 2027 | Yes |

The first is this programme's share of the lab's internal R&D money, set up as a fund of its own. The second is a grant. Both belong to HEK293 Suspension, so every pound has one owner. The split across projects:

| Project | Fund | Allocation | Dates |
| --- | --- | --- | --- |
| `PRJ-MEDIA-26` Serum-free media screen | Process R&D 2026 | `£2,500` | 2 Mar to 29 May |
| `PRJ-ADAPT-26` Suspension adaptation | Grant | `£3,000` | 1 Jun to 30 Sep |
| `PRJ-SCALE-26` 2 L bioreactor scale-up | Grant | `£14,000` | 3 Aug 2026 to 26 Feb 2027 |
| Held back | Grant | `£3,000` | |

The programme's draw is `£19,500`: `£17,000` from the grant (87%) and `£2,500` from internal money (13%).

Notice the order. The media screen ran first, on internal money, before the grant existed. It cost `£922.80`, tested three serum-free media on a suspension-adapted reference line and picked one. Cheap evidence first, then the bigger commitment. It's staged funding, the same logic as building a prototype before you order the tooling.

## How much of a budget should you hold back?

Enough to cover the thing you already suspect will go wrong. For this grant that's £3,000 of £20,000, 15%, unallocated on purpose, because adapting a cell line to serum-free suspension is slow and hard to predict.

It did go wrong. At 0.5% serum in early August the cells clumped and viability fell to 82%. The adaptation took three weeks longer than planned. In money, the slip cost `£371.10`: five extra passages and a bottle of anti-clumping agent. The £3,000 wasn't touched. The three weeks were the expensive part, because the grant ends on 28 February whatever the cells do.

That's the case for holding back unallocated money rather than padding every project. Padding hides inside each project and gets spent. Headroom sits where the lead can see it and decide.

## Where it goes wrong

**Closed projects keep their money.** The media screen closed on 29 May with £922.80 spent of £2,500. The other `£1,577.20` stayed on its allocation, still counted as committed and still inside the programme's draw, until the August review lowered the allocation to what was spent. It's a one-line fix. It just needs someone to do it when the project closes, not three months later.

**Splitting by programme first.** Give a programme £20,000 and let its projects work out their share, and the first project to order takes the most. Allocate to projects and the argument happens once, at the start, in the open.

**One fund in two spreadsheets.** A shared internal budget shows its whole balance on each programme's sheet, and two leads plan to spend it. Each fund needs one owner and one total, which is why CCA Labs gives every fund to exactly one programme.

**Comparing projects on different funds.** The internal fund doesn't charge bench time and the grant does, so a run on the screen looks cheaper than a run on the adaptation for reasons that have nothing to do with the science. I go through that in [what one experiment really costs](/articles/cost-per-experiment).

## How CCA Labs holds the split

A fund belongs to one programme. A project belongs to exactly one programme too, and it can only draw on that programme's funds. Every organisation has a default programme, **General**, for work that isn't filed anywhere yet; it never gets a pace verdict and reads **Unfiled**.

The programme page adds everything up from the projects when it loads: **Forecast at Term**, with what's spent and projected against the draw, a **Funds** view of how much each fund is drawn, spend by month, and each project's share of its allocation spent against the programme's own.

On 31 August the adaptation has spent 72.9% of its allocation while the programme as a whole has spent 33.3%, so the adaptation reads **Behind**. That's a pace signal, not an accusation: part of what it paid for, about 1.8 litres of medium and 48 flasks, is still on the shelf. Changing an allocation, including handing back a closed project's leftovers, is **Edit Project** on the project's **Spend** tab.

When you [create your lab](https://lab.cca.bio/), a fund per source of money and a programme per line of work is the whole setup. The planning tools are on the [all-in-one lab page](/all-in-one-lab#plan).

## The programme's year, part by part

The rest of this example is told floor by floor, each part for the people who work there.

| Part | For | What happens |
| --- | --- | --- |
| [Research goals and milestones](/articles/research-goals-milestones) | Heads of R&D, PIs | Three goals, and the rule that flags one At Risk in late August |
| [Seeing which projects slip](/articles/lab-project-roadmap) | Project leads, lab managers | The 10 August re-plan, baselines and a +21 day slip |
| [What one experiment costs](/articles/cost-per-experiment) | Scientists, whoever prices work | `£61.42` a passage, and why the programme says `£138.28` |
| [Burn rate and runway](/articles/lab-burn-rate-runway) | Grant holders, finance | A grant that is under-burning because of a slip |

The split is only the starting position. What happens to it depends on whether the work goes to plan, and the first thing that tells you it isn't is a goal falling behind. That's where to go next: [how to set research goals and milestones that show when you're behind](/articles/research-goals-milestones).

## Part of: One programme's year, from the budget split to the August review

One example programme, followed through a small biotech: getting HEK293 cells to grow in suspension without serum, then in a 2 L bioreactor. Leadership splits two funds across three projects and sets the goals, one project slips three weeks, the bench counts what each run costs, and the August review shows what the slip did to a grant with a fixed end date. Each part is written for the people on that floor.

- (this article) [How should a lab split one budget across programmes and projects?](https://cca.bio/articles/research-program-budgets) (Overview)
- [How do you set research goals and milestones that show when you're behind?](https://cca.bio/articles/research-goals-milestones) (Leadership, for heads of R&D and PIs who set the goals)
- [How do you see which lab projects are slipping?](https://cca.bio/articles/lab-project-roadmap) (Operations, for project leads and lab managers who keep the plan)
- [How much does one lab experiment really cost?](https://cca.bio/articles/cost-per-experiment) (Lab, for scientists who run the work, and whoever prices it)
- [How do you work out a lab's burn rate and runway?](https://cca.bio/articles/lab-burn-rate-runway) (Leadership, for grant holders, heads of R&D and finance)

## Frequently asked questions

### What is the difference between a programme and a project in a lab?

A programme is a line of work with a purpose and goals, such as getting a cell line to grow in suspension. A project is one piece of that work, with an owner, start and end dates and an allocation of money. Programmes run for years; projects finish. Money sits on projects, and the programme adds them up.

### How much of a grant should a lab keep unallocated?

There is no fixed rule; it depends on how uncertain the work is. Our example lab keeps £3,000 of a £20,000 grant, 15%, because adapting a cell line often takes longer than planned. It decides at each review whether to release the money, and to which project.

### Can one fund pay for projects in more than one programme?

Not in CCA Labs. A fund belongs to one programme, and only that programme's projects can draw on it. If two lines of work share a pot of money, give each its own fund from it, so each total is clear and the same pound cannot be planned twice.

---

Cite as: CCA Bio. "How should a lab split one budget across programmes and projects?". 2026-11-17. https://cca.bio/articles/research-program-budgets
