Why the grants are not landing

These are the factors the evidence supports most strongly. Between them they account for two of the three refusals outright and most of the third.

These reduce competitiveness. None of them is decisive on its own, and together they are what turns a near miss into a refusal.

• The reserves explanation is missing. Not the reserves. The explanation.
• The budget narrative does not reconcile. Two subtotals in the refused Funder E bid do not follow from the rows above them, and a third figure contradicts the Funder C application in the same pack. This takes two minutes to fix and it is the first thing a finance reviewer does.
• One organisation, several sets of numbers. The children's programme has two different starting years across the pack; the second designated fund two values; the legacy two values; Funder C three different annual figures.
• Limited partnership evidence. The neighbour charity won its Funder H award in partnership with another borough organisation. Partnership is a funder preference you are not using.
• Sustainability is asserted rather than argued. No application answers the question a funder is really asking, which is what happens to this work when the grant ends.
• No prior-impact section that reads like proof. You hold LeDeR mortality evidence most applicants would pay for and it appears in one bid.
• Repeat-application timing is not being tracked. Funder E's twelve-month rule is published; Funder A has one and two-year rules; Funder F has a five-year rule. A calendar solves all three.

Stated plainly, because the board is entitled to hear the negative finding as clearly as the positive ones.

• There is no evidence that high reserves caused any of your three refusals. Reserves are ruled out on four of your six applications and unsupported on the other two; and because two of those four were applications you won, the load-bearing finding is that two of the three real refusals are ruled out on the funder's own rules and the third is left honestly open.
• There is no published funder rule among your three refusers that penalises your reserves. Funder O requires a minimum and publishes no maximum. Funder I forbids only adding its grant to reserves. Funder E publishes no reserves rule at all and invites the explanation your bids did not give.
• Comparable funded charities hold similar reserves, and two hold more. Comparator 2 holds more than you, most of it designated, and was funded core costs by the funder that refused you.
• The applications did not fail on writing quality. The better-written Funder E application is the refused one.
• One real exception, stated so nobody else has to find it. Funder Q publishes a genuine reserves ceiling at twelve months of expenditure, and while the designation stands you are over it. Funder Q is named in your own paperwork as a funder you approach. It should come off the list rather than collect a refusal, and it is not one of the three refusals in question.

The seventeen funders of chapter 4, sorted into four tiers on the evidence in chapters 4 and 5: Priority, Conditional, Low priority, and Do not pursue. One caution the board should hold while reading it: funding rounds open and close, and a list dated today is a list dated today. Every entry carries the date we read it, and the tiers should be re-run at the start of any engagement rather than inherited.

You currently present as a local service provider, and that is accurate but incomplete. Three positions are available to you on the evidence, and the right one differs by funder.

• The borough's specialist institution. For the place-based funders and Funder C. Twenty years of the children's programme, one borough, irreplaceable local knowledge. This is your strongest position and it is under-argued.
• A learning-disability specialist, defined diagnostically. For Funder A and the specialist trusts. This requires stating who your users are in the funder's own definitional terms, which no current bid does.
• A resilience case. For Funder D, Funder B and any core funder. An established provider, four-fifths publicly funded in a borough changing how it commissions, three deficits deep, protecting a legacy for the long term, that needs independent income to stay independent. That is a true and fundable story and nobody has told it.

What you should not do is present as an innovation organisation. The refused Funder E bid leaned that way, on a model tested with about thirty families, against a funder that had just backed your twenty-year service.

1. Screen on the funder's own words before writing. Eligibility, minimum and maximum size, exclusions, themes and re-application dates, with a written reason for every fund rejected. Two of these three losses never reach a writer.
2. Tell the reserves story the accounts actually tell. The paragraph is in 9.4. Use it in every application.
3. Frame the ask in the words the funder leaves open. Core running costs at Funder E. Named posts over years at Funder C. Capital at Funder F. Whole programmes at Funder B.
4. Size the ask from the funder's demonstrated middle, not its published maximum.
5. Re-add every subtotal from its own rows before sending. Two minutes, and it is the first thing a finance reviewer does.
6. One set of numbers across every document, each figure dated, reconciled to the filed accounts before anything leaves.
7. Lead with the evidence you already own. The LeDeR mortality data belongs in every health-adjacent bid you write.
8. Name the partnership. Your neighbour won its borough award in partnership. Funders read partnership as reduced risk.
9. Answer the sustainability question with the public-income concentration, not with an assurance.

Do not apologise for the reserves and do not hide them. Explain them, in roughly this shape, adjusted to the year and to the fund. This is drafted from your own filed accounts and every figure in it is checkable.

Two sentences of that paragraph do the work: the one that says the legacy is a trustee designation made under a Letter of Wishes, and the one that says you are below your own target. Note that it gives both cost bases rather than the larger shortfall alone. A funder's finance reviewer will divide your expenditure by two and get the smaller target, and a paragraph that only offered the larger one would look like the subtotals section 6.3 criticises. Neither appears in any of your six applications. Both are already public in your filed accounts, so saying them discloses nothing a funder cannot already read, and there is only a cost in not saying them. One drafting note, because a funder may push on it. The paragraph says the trustees do not propose to release the legacy, which is true and defensible. It deliberately does not say they cannot, because on your own accounts a designation is a trustee decision and not a donor restriction, and that is a claim we would not ask you to sign.

• A ranked pipeline, re-run quarterly. Section 9.1 is the current ranking. It has a shelf life measured in weeks, not years, because rounds open and close.
• A realistic volume. Six applications produced three awards. That is a good conversion rate wasted on too few attempts. Twelve to eighteen properly screened applications a year, weighted toward the priority tier, is the shape.
• A balance between project and core. Five of the seventeen funders put core or unrestricted money on the table explicitly. None of your six applications asked for it in those words. At least a third of the pipeline should be core asks.
• Relationship building, not just submission. Funder E does not publish refusal reasons; a phone call to the funder, which in our experience they take, is the only route to the real answer on the second application.
• A feedback tracking system. Every outcome logged with the stage it reached, the reason if given, and the earliest date you may reapply. Three of the funders here have re-application rules that a calendar would enforce for you.
• A six-month review. Score this study against what actually happens. The Funder O prediction in chapter 6 is deliberately falsifiable and dated so that the first review has something real to mark.

Two things before the findings. First, the board was right to ask. The question that commissioned this study was a good one, it was asked about the correct part of the accounts, and the instinct underneath it holds: something about the reserves is costing money. The board had no way of knowing that the cost sits in the explanation rather than the amount, because that only becomes visible once you read six applications against seventeen funders' own rules. A cautious board that demanded evidence before spending is a board doing its job, and this study exists because it did. Second, almost everything here is a process finding, and those processes belong to people doing a hard job in a hard funding climate, with three of six applications funded. None of it is a judgement on anyone's competence and it should not be carried into a board meeting as one. The analysis indicates that the charity's grant difficulties are primarily explained by five things, in this order: applications reaching funders whose published rules exclude them; asks written in a form the funder does not buy; asks sized outside the funder's demonstrated range; a financial position that is never explained, so the balance sheet explains itself; and a market in which you are not competing where you are strongest.

The evidence does not support reserves as the cause. Reserves are ruled out on four of your six applications and unsupported on the other two, and since two of those four were applications you won, the claim that matters is that two of the three real refusals are ruled out and the third is left open. Of seventeen funders profiled, exactly one publishes a reserves ceiling you would fail, and it is not among the three that refused you. Charities holding more than you, in the same rounds, from the same funders, were funded.

Reserves may still affect how a funder interprets your financial need, and on the evidence they probably do, because your applications show a six-figure investment fund and never explain it. That is a presentation problem with a one-paragraph fix, not a balance-sheet problem with a five-year fix. Section 9.4 is the paragraph.

One honest qualification the board should hold. We can rule reserves out for a particular application or a particular funder, on that funder's own rules or conduct. We cannot rule them out of every human mind in every grants committee, and no study built on public evidence could. What we can say is that across six applications, seventeen funders, six comparator charities and 5,700 award records, we found no evidence for the board's hypothesis and a great deal of evidence against it.

1. Funder A, and we are recommending 31 December rather than 31 August. This is the learning-disability fund with no grant to you in the published record, and the best-fitting funder on the whole map on its published criteria, which is exactly why we would not gamble it. Here is the trade, both sides of it, because the case for August is not nothing.
   
   What August costs and what it buys. The 31 August round is decided at the November 2026 meeting; the 31 December round is decided in March 2027. Choosing December costs you four months of decision time, and you are in your third consecutive deficit, so that is a real cost and we are not going to hide it. Against that: 31 August is the summer bank holiday, so the only working days before it are Wednesday 26, Thursday 27 and Friday 28 August. In them you would have to open an account on the funder's portal, produce a statement of the proportion of your users who meet its diagnostic definition, which none of your six bids currently carries, and land two external references.
   
   The honest asymmetry. If the references simply arrive late, the funder says the application is held and put to the next meeting, which is where a December application lands anyway; lateness costs you nothing you would not already have spent. The risk that matters is different: a rushed application that is refused is one the funder says will not normally be considered again for a year, and that closes the best-fitting funder on this map until late 2027. So the question is not speed against delay. It is whether three working days can produce an application good enough not to be refused, at a funder with a diagnostic gate none of your current bids answers. We do not think so, and that is why we would take December.
   
   The critical path, with owners. Four things have to happen and only one of them is ours. You: confirm whether the charity has ever approached this funder, since the reapplication rule turns on it and no public data shows it (section 10.3). You: tell us the proportion of users of the children's programme and the adult skills centre who meet the funder's definition, because your own Funder E bid describes the cohort as autism, ADHD or other learning disabilities, and the funder funds the last of those three. You: line up two referees, which is the long pole in any month. Us: the funder research, the evidence base and the numbers reconciled to your accounts, then a hard internal challenge before it goes.
2. Before anything else goes out: adopt the reserves paragraph. Section 9.4. It costs nothing and it changes what every future funder reads.
3. Start Funder B now. Forty weeks to a decision, at your best-proven funder, at an average of £319,000 over three years. Your team writes it; we build the evidence behind it.
4. Decide on the Funder I round deliberately. It closes at noon on Tuesday 1 September, so the bank holiday leaves three working days and a Tuesday morning. Our read is in section 6.5 and it is a no, but it is the board's call and it should be a decision rather than a default.

“Reserves” is four different numbers in your own accounts, and the board's question changes answer depending on which one is meant. Reconciled for the year to 31 March 2025:

One arithmetic note we owe you. The designated funds plus the free unrestricted funds come to £2 more than the total in the accounts' own note. It is a rounding difference and it changes nothing, and we are pointing at it because a study that only shows the numbers that agree is not worth reading.

The regulator does not set one reserves level for all charities. It expects a charity to explain why it holds what it holds and how that relates to its own risks. That is a communication duty, and it is the duty your applications are not discharging.

This study leans on the fact that the legacy is a designation your trustees chose rather than a restriction the donor imposed. A trustee who follows that thought reaches the obvious next question within about ten seconds, and it deserves a straight answer rather than a silence.

The general position first. Designated funds are unrestricted funds that trustees have set aside for a purpose of their own choosing, so the decision that created them is in principle a decision the trustees can revisit. We are not your legal or accounting advisers, we have not seen anything beyond your filed accounts, and the ruling on your own funds belongs with your auditors and your board, not with us. On that footing the likely answer is yes, you probably could, and we would advise against it. What we can tell you with confidence is what un-designating would and would not achieve, because that part is answerable from the evidence in this study.

• It would fix nothing on the list in chapter 8. None of the five primary barriers in chapter 8 is the amount of money you hold. Two of the three refusals turn on subject matter and income band, and the third on how the ask was written.
• It would change your position at one funder in seventeen. Funder Q is the only published reserves ceiling on the map, and it is not one of the three that refused you.
• It would cost you the best sentence you have. A designation made under a Letter of Wishes, explained properly, is the answer to the question a funder is silently asking when it sees a six-figure investment fund. Release the designation and you still hold the money, and now you have no explanation for it.
• And it would set aside a donor's expressed wish. The Letter of Wishes is not a legal restriction, which is exactly why the choice is yours and why we will not make it for you. It is still a wish, from someone who left your charity a great deal of money.

Our recommendation is the opposite of unwinding it: keep the designation, and start saying it out loud. Section 9.4 is the paragraph.

The cash you can touch has been flat for three years, and slightly below your own target for the last two. Nothing in the trend supports a picture of an organisation accumulating money it does not need. What changed in that period was not the amount, it was the column: in the year to 31 March 2024 a single transfer moved £ out of free unrestricted funds and into a designated investment fund. Your financial position did not change on that day. What a funder sees did.

And the last piece of the trend is the one your bids never use: you have spent more than you raised in each of the last three financial years.

The hypothesis only works if funders can see something alarming. So we listed what was visible, per source.

• In the applications. Funder E bid 1 disclosed the legacy, the free reserves and a five-figure drawdown. Funder E bid 2 disclosed a larger legacy figure, the second designated fund, a higher free reserves figure and a £0 drawdown. The Funder H bid disclosed a five-figure allocation from reserves. The Funder I draft stated the six-month target and the free reserves figure quote.
• In your published accounts. Everything: the seven-figure headline, the designations, the Letter of Wishes sentence, and the three deficits.
• The reserve-policy explanation. Present in the accounts. Absent from all six applications.
• Committed expenditure and the proposed use of reserves. Answered once, with a five-figure sum, in the application that was funded. Answered £0 in the one that was refused.
• The need for grant funding. Never argued from the financial position in any of the six.

So a funder reading the second Funder E bid saw a larger balance than in the funded bid, no explanation of the designation, and no intention to contribute from reserves. That is the version of the board's hypothesis that survives contact with the evidence, and it is not about the amount held. It is about what was said about it.

The board's question, tested the only way it can be tested: do the organisations winning this money hold less than you? We downloaded the latest filed accounts of six recent winners and read their balance sheets and reserves policies, and set them beside your own.

Read it both ways, because the board should. On total unrestricted funds you sit third of the seven organisations in that table, below two organisations that win constantly. On free reserves, the measure the regulator and the funders' own assessment forms use, you sit mid-table at 5.9 months, below Comparator 2 at 6.3. Winners sit above you and below you on every measure. Reserves do not sort them.

The sharpest single fact is Comparator 2. In the same January 2025 round, the funder that refused you funded a charity holding more unrestricted money than you, most of it designated, whose trustees describe holding around six months of total expenditure. That is your structure, at a larger scale, funded for core costs, by the funder your board believes refused you over reserves. And in that same round the same funder gave core costs to Comparator 5, holding a small fraction of your funds. The top and the bottom of the reserves scale, in the same month, from the same committee.

Five verdicts were available: supported, partly supported, not supported, ruled out for this application, indeterminate. “Ruled out” is used sparingly and only where the funder's own rules or conduct exclude reserves as the cause for that application. It is never claimed for the market as a whole.

Across the six: reserves are ruled out on four, not supported on two, and supported on none. Two of those four are applications you won, so the claim that carries the weight is the narrower one: of the three real refusals, two are ruled out on the funder's own rules and one we honestly cannot close. As a general explanation for your grant performance, the hypothesis fails. There is one real reserves exclusion in this market, at Funder Q, and it is not among the funders that refused you. Reserves may still shape how a funder reads your financial need, which is a different and entirely fixable problem, and section 9.4 fixes it.

Nine dimensions, each scored 0 to 3, each drawn from the funder's own published criteria. 0 means no evidence of fit, 1 weak, 2 partial, 3 strong. Every score below 3 has its reason in appendix D, the application-level reading is in section 6.5, and where we could not see the evidence we have written a question mark rather than a number.

Three things to take from that table, and one thing not to.

• The funded three average 25.7 out of 27. The refused three do not. The gap is not effort and it is not writing quality; it is fit, and fit is knowable in advance from public documents.
• Reserves are not one of the nine dimensions, because on the funders' own published criteria there was nothing to score. Only one of the seventeen funders in chapter 4 publishes a test your reserves would fail, and you did not apply to it in this set.
• The Funder I application scores 23 and was still refused. That is what a priority queue does: it refuses eligible, well-made applications from organisations outside its priority band. Its single low score, organisation size, is the whole story.
• What not to take from it: that a score decides anything. Funder O fails on two zeros, not on a total. A single 0 on mission is fatal regardless of what the other eight columns say, and a total is only ever a way of seeing the shape.

Quality read separately from fit, against eleven things a funder's assessor looks for. The pattern is a competent, evidence-rich set of applications with two consistent weaknesses, both financial and both fixable.

• Clarity of need: strong. The refused Funder E application contains the sharpest need statement in the pack, setting the borough's official count of children with a diagnosed learning disability against the higher figure the charity's own casework shows quote.
• Strength of evidence: strong. The LeDeR mortality data in the Funder H bid is better evidence than most applicants ever hold.
• Specificity of outcomes: adequate. Funder H's published description of your award reads as a clear specification. The others are less concrete.
• Delivery plan: strong. Two decades of the children's programme is a real answer to the credibility question.
• Budget: weak. The second Funder E bid carries two different project cost totals across its own income and expenditure tables, two subtotals that do not follow from the rows above them, and a Funder C income figure that contradicts the Funder C application in the same pack.
• Value for money: adequate. The percentage-of-cost framing is defensible but it invites the wrong reading at a funder that excludes named-post salaries.
• Monitoring and evaluation: adequate. Outputs are counted. Outcomes are asserted rather than measured, and no application says who would measure them or when.
• Sustainability: weak. No application makes the case that this organisation, at 81% public income and three deficits, needs and will use trust income to become durable.
• Risk management: not addressed. No application names a delivery risk or what would be done about it, and several funders ask.
• Explanation of reserves: weak, and this is the finding that matters. The refused Funder E bid gestures at the donor's wishes. The Funder I draft states the six-month position. No application states that the legacy is a trustee designation under a Letter of Wishes rather than a donor restriction, and none states that accessible reserves sit below your own target.
• Consistency with your public accounts: weak. The pack gives the children's programme two different starting years; the second designated fund at two different values; the legacy at two different values; Funder C at three different annual figures. Each is probably a different snapshot in time. A funder reading two of them side by side sees carelessness.

An eligibility failure is a different event from losing at the final competitive stage, and treating them the same is how a fundraising strategy stops improving. Sorted properly:

Sorting them the way a grants strategy has to sort them: by the gate you have to pass, not by how appealing they sound.

• By geography. Borough only: Funder H. London-wide: Funders C, J and L. South of England: Funder I, and Funder K's southern round. UK or England-wide: Funders A, B, D, E, F, G, M, N, O, P and Q.
• By issue. Learning disability as the defining purpose: Funder A alone. Disability among named priorities: Funders C, F, L and O. Everything else reaches you only through a broader theme such as poverty, marginalisation, health inequality or children and young people.
• By beneficiary age. Under-18s only: Funder G. Adults only: Funder K's current grants. All ages: the rest.
• By grant size. Under £15,000: Funder F's small grants, Funder K, Funder H's micro strand. £15,000 to £60,000: Funders E, J and I, and Funder H's delivery strand. £60,000 upward: Funders C, D, B, L and O.
• By organisation size. Funders that prioritise smaller charities than you: Funder I (£250,000 and below), Funder P (under £500,000). Funders whose ceiling is far above you: Funder F (£10 million turnover), Funder M (£2 million). Funders with a floor you clear: Funder D (£100,000 turnover), Funder O (three months' free reserves).
• By cost type. Capital only: Funder F. Revenue only in practice: Funders M and K. Both: Funders A, E, C, B and D.
• By restriction. Explicit core or unrestricted money on the table: Funders D, C, E, A and G. Project money only: the place-based funds and Funder K.
• By term. One year: Funders K, F and H. Two to three years: Funders I, E and G. Three to five years: Funders C, B and D.

Full profiles for all seventeen are in appendix A. Six of the seventeen carry a live date; the nine below are the ones that decide your next twelve months.

A funder's strategy page tells you what it would like to be. Its award record tells you what it buys. Where the two differ, the record wins, and the difference is where applications are won and lost.

• Funder E says projects and gives core. Its own record-year announcement reports nearly 3,000 grants of which “Half of these were made towards organisations' core running costs” and “94% were in our regular grants category, grants under £100,000”. Its exclusion page says it will not fund the salaries of specific posts, while its core costs page says operating or core cost grants can be used to help pay bills or rent and towards salaries. The same money, described two ways, with only one of them fundable.
• Funder C says core and buys posts. Its policy language is core and unrestricted; 90 of the 240 awards in our library describe a named post, a salary or staffing costs, over terms of two to five years. Comparator 1's three part-time trainers. Another London charity's five-year award for two named posts. Both are true at once: it funds people, for years, and calls that core.
• Funder B says communities and buys programmes. The current average is £319,000 over three years. Its recent learning-disability awards are whole programmes, not contributions: Comparator 3's three-year award, a national charity's similar one, and the neighbour charity's largest grant.
• Funder F says infrastructure and means it. Every eligible cost is a thing you can photograph. Its 259 records are dominated by refurbishment and equipment. A revenue ask here is not a weak application, it is an ineligible one.
• Funder A says learning disability and enforces it diagnostically. The gate is not the cause, it is the IQ-and-onset definition, and applications from mixed-need services fail when they gesture at the client group instead of stating the proportion.

For every funded organisation we could identify from a recent award by one of your funders, we collected: name, charity number, geography, beneficiary group, issue area, annual income and expenditure, the grant amount, the grant purpose in the funder's own words, the duration, any prior relationship with the funder, and the reserves position from its own filed accounts. Six organisations had everything available. They are carried in full in appendix B and set beside you in section 7.4, because reserves are the reason the dataset exists.

Sorting the funded organisations by kind rather than by cause shows where you sit and where you do not.

• Medium-sized specialist charities in one place. Four of the funded organisations we identified are exactly this. This is your category, and it is winning.
• Small grassroots and community-led organisations. Consistently funded, and often at sizes that would matter to you. The neighbour charity is one of them.
• Large national organisations. Present in the record but a different market.
• Service-delivery charities. The bulk of the disability awards we studied, including yours.
• User-led and disabled-people's organisations. A distinct and growing category, and the one funders name most often in current policy language. This is the category that Funder L's main disability fund is gated to.

Counted across 5,700 award records from 25 funders publishing to the 360Giving standard, of which 454 match our disability test and 142 match our learning-disability test. The exact tests are printed in appendix H.

• Which organisation sizes get awards. All of them. In the same Funder E round of January 2025, Comparator 2, holding seven figures in unrestricted funds, and Comparator 5, holding five figures, were both funded core costs. Size does not sort this market.
• Which locations are represented. Of the 240 Funder C records, 19 of the 33 disability awards carry a London borough in the text; of the 240 Funder L records, 20 of 34. Place-based funders fund their own place, hard. Your borough advantage is real and it is currently being used by the neighbour charity rather than by you.
• Which beneficiaries are prioritised. Learning disability is a small but persistent share: 142 of 5,700 records, 2.5%. It is never the largest category at a general funder and it is never absent either. That is why the specialist funders matter so much: at Funder A the share is the whole portfolio.
• Typical award size, by funder, on the learning-disability records only. Funder C £57,100. Funder M £29,100. Funder F £25,000. A co-operative retailer's foundation £20,000. Funder J £17,413. Funder E £15,000. Funder D £122,000. Funder G £74,240. These are medians of the matching records in our library and they are how you should size an ask.
• Restricted against unrestricted. Core and unrestricted language appears at Funder E (half of all grants), Funder C (policy commitment), Funder D (explicit), Funder A (explicit) and Funder G (core stream). It is available far more often than your applications assume.
• One-off against repeated. Repeat funding is normal at Funders C, B and J, and unusual at Funder E: across our 240 Funder E records, 240 distinct recipients appear and none is funded twice within the set.
• Existing relationships. Comparator 1 appears in the Funder C record in 2009 and again in 2022, and in the Funder J record in 2020. The organisations doing best are the ones being renewed, not the ones starting over.

These are the recurring characteristics, in the order the evidence supports them. Nine of the ten are things you can change this month.

• The work matches a programme the funder actually runs, verified against the funder's own current pages rather than its reputation.
• The problem is narrow. Comparator 1 asked for three named trainers, not for youth disability work in south London.
• The need is evidenced locally. The winning applications quote numbers about their own borough and their own caseload.
• The outcomes are measurable and stated. Funder H's own description of your award reads like a specification, and that is why it reads well.
• Local legitimacy. Place-based funders fund organisations that are visibly of the place.
• A credible delivery model, already running. Funder E funded the children's programme, twenty years old. It refused the early-years extension, tested with about thirty families.
• The ask is in the funder's own form. Core costs at Funder E. Named posts over years at Funder C. Capital at Funder F. Whole programmes at Funder B.
• The amount sits inside the funder's demonstrated range. Not its published maximum: its actual middle.
• Budget justification is specific, and the arithmetic in the budget agrees with itself.
• The financial position is explained rather than left to be inferred. Comparator 2's accounts describe the same design as yours, a large designated fund held apart and a six-month free-reserves policy, and it was funded core costs by the funder that refused you.

This is the half of the map most funder research skips, and it is the half that would have saved two of your three refusals.

• Work outside the stated subject, however good. Funder O does not fund learning-disability services; it funds transport. It has funded three national disability charities, and every one of those grants was about journeys.
• Organisations above a priority band. Funder I will consider you and will prioritise charities at £250,000 income or below. Funder P will not consider you at all above £500,000.
• Revenue costs at a capital funder. Funder F excludes salaries, overheads, training and rent by name.
• Salaries of specific posts at Funder E, alongside core running costs that include salaries. The distinction is entirely in how the ask is written.
• Autism, ADHD, dyslexia and dyspraxia without an intellectual disability at Funder A, plus advocacy projects and arts and theatre projects, which are outside its normal policy even for learning-disabled beneficiaries.
• Adults at Funder G, which funds beneficiaries aged 18 and under only.
• Organisations holding more than twelve months of expenditure at Funder Q. The only genuine reserves exclusion we found in the whole market, and one that currently applies to you.
• Applications written by consultants at Funder B. A rule about who holds the pen, not about the organisation.

Two of your three refusals sit in this list, and both were knowable before a word was written. That is the difference between an organisation that is weak and an organisation that is approaching the wrong part of the market, and it is the central finding of this study.

• Legal status. Private company limited by guarantee, Companies House number no., registered office address.
• Registration. Registered charity no.. Reporting recorded as up to date on the register.
• Mission and beneficiaries. People with a learning disability and their families, described on the charity's own register entry through children's and adult services: behavioural support, activity sessions, advocacy, job coaching and day services.
• Geography. One London borough. Local, not regional and not national. This matters more than it looks, and chapter 4 shows why.
• Main activities. The children's programme and its early-years extension for children and families, the adult skills centre, plus advocacy, employment support and day services.
• Size. Income £ for the year to 31 March 2025. On the funders' own bands that places you as a medium-sized local charity: above the small-charity thresholds most place-based funds prioritise, well below the ceilings the large trusts apply.
• Governance. no. trustees on the register. One authorised signatory, the chief executive.

Three consecutive years of spending more than you raise, on a rising income line, behind a balance sheet that looks larger than the money you can touch.

Read the last column and then the one before it. The money you can actually spend has been essentially flat for three years, rising gently. Everything above it is either held as investments and fixed assets or formally designated. In the year to 31 March 2024 a single transfer of £ moved out of free unrestricted funds into a designated investment fund. Nothing was earned and nothing was spent to make that happen. It was a decision about where the same money sits, and it is the decision that changed what a funder sees.

For the year to 31 March 2025 the register analyses the charity's income across donations and legacies, charitable activities and investments. Against that same total it records most of the money as received under government contracts and government grants. We have not reconciled those two public-money lines to each other in the notes, so we state them as the register states them. On any reading of them, public money is the large majority of your income.

That is the strategic fact behind the board's question, and it is more urgent than the reserves. You told us in July that the borough's commissioning is shifting towards Public Health. A charity whose income is four-fifths public, running its third consecutive deficit, needs trust and foundation income to become a real column rather than a hopeful one. In the year to 31 March 2021 your income was far above your expenditure. That was the legacy year. It is also the year that created the balance sheet the board is now worried about.

• Published grant history. A published history running back to 2006. Of its total value, 85% came from one funder, Funder B. The rest came from six further funders in five-figure and four-figure amounts.
• Recent record. Three of the six applications in this study were funded, worth £ in confirmed first-year value: Funder C, Funder H and Funder E. The Funder C application asked for roughly five times its first-year value over five years; we have not seen a document confirming the later years, so we report the year we can prove and flag the rest.
• Average request size. Of the five applications carrying a stated ask, the requests run from a low five-figure sum to a six-figure multi-year one, with a five-figure median first-year ask.
• Funding types requested. Overwhelmingly project and part-salary. None of the six asks for unrestricted or general core support in those words, although the funded Funder E bid comes closest by framing its request as a contribution to project costs.
• Relationships your own paperwork discloses. Beyond the five funders in this pack, your bids name fifteen further funder relationships: four live or planned, three received and eight pending, reconstructed from two of your own documents. The names are withheld in this edition.

What a funder can see from your applications: a long-established local learning-disability charity with a strong local evidence base, delivering direct services in one borough, asking for percentages of named staff posts inside project budgets, and holding a seven-figure balance sheet in unrestricted funds with no explanation attached.

What is true and does not reach them: that the legacy is a designation your trustees chose under a Letter of Wishes, not a restriction imposed by the donor; that the money you can actually spend is a fraction of that headline; that this is below your own board's reserves target; that you have run three deficits in a row; and that your own bid holds evidence most applicants would pay for, the LeDeR finding that the median age at death for an adult with a learning disability is 63 against 82 for men and 86 for women in the general population, with 38% of deaths from avoidable causes.

A charity running three deficits, slightly below its own reserves target, holding a legacy its trustees have chosen to protect, is a fundable proposition with a clear story. A charity sitting on a million pounds is not. Both descriptions fit the same balance sheet. Only one of them is true in substance, and it is not the one your applications leave a funder to infer.

The purpose is to diagnose why the charity's recent grant applications have not succeeded, and to test one specific hypothesis the board holds, by analysing eight things together: each funder's published strategy, each funder's eligibility rules, the profiles of the organisations those funders actually fund, the pattern of their awards, the fit of your applications to the funders you sent them to, the quality of those applications, your financial and organisational presentation, and the reserve hypothesis itself.

This is a diagnosis, not a bid-writing critique. We do not comment on anyone's prose, and chapter 6 assesses the applications on fit, arithmetic and what each funder's own form asks for, never on style. The reason is in the evidence: the better-written of your two Funder E applications is the one that was refused, so telling you to write better would be advice that the facts contradict.

• Who do these funders support, and what do their portfolios reveal that their strategy pages do not?
• What types of project, cost and organisation do they prioritise?
• What characteristics recur among the organisations they fund?
• How does the charity compare with those organisations, on mission, size, geography and reserves?
• Is the charity applying to funders whose demonstrated preferences match its own profile?
• Are reserves a material barrier, a presentation problem, or an unrelated factor?
• What should change in the funding strategy, and in what order?

We tested five explanations against the same evidence rather than setting out to confirm one. The verdicts below are the study's conclusions, evidenced in chapters 6, 7 and 8. They are stated here so that a reader who disagrees knows exactly which chapter to attack.

Five pieces of work, done in this order, each one able to falsify the last.

• A document-based case study of the charity, built from filed accounts, the register record and your own applications.
• A grant-market analysis of the funders in your orbit, built from each funder's own published guidance and from award data published to the open 360Giving standard.
• A comparative grantee analysis: charities that recently won money from the same funders, read from their own filed accounts.
• A structured review of each application outcome against the funder's own rules, scored on nine dimensions.
• A financial review testing the reserve hypothesis directly, per application and across the market.

• Your six applications, forwarded by your team on 5 August 2026.
• Your annual report and accounts for the years ended 31 March 2023, 2024 and 2025, downloaded from the Charity Commission register.
• Your reserves policy, as stated in the Trustees' Report in those accounts.
• Funder websites, application guidance and published forms, each read on a stated date.
• Funder annual reports and announcements.
• Published grant lists, including Funder C's committee papers and Funder H's round results.
• Award data published to the 360Giving open standard, searched through GrantNav and read through the funders' own data feeds.
• Charity Commission register records for you and for every comparator charity.
• Companies House, for your corporate form: private company limited by guarantee, number no..
• The filed accounts of six funded charities, downloaded individually.
• Your own emails of 27 July, 5 August and 11 August 2026.

The 360Giving data matters more than any single funder page, because it records what a funder did rather than what it says. Each record carries the recipient, the amount, the date and the funder's own description of what the grant was for. Read across hundreds of records it produces a profile of the organisations a funder actually chooses, by size, type, geography and purpose.

Three levels, kept separate throughout, because conflating them is how funder research goes wrong.

• Funder level. What the funder's strategy and portfolio reveal about what it buys.
• Grant level. What each individual award was intended to support, in the funder's own words.
• Organisation level. What characteristics successful and unsuccessful applicants appear to share.

Two appendices exist so that a sceptical reader can test this study rather than trust it. Appendix H prints the exact search rules behind every counted figure, so anyone with the same open data gets the same numbers. Appendix K lists the things we got wrong and withdrew while writing, including an explanation of one refusal that your own funded bid disproved. If the board wants to know how much weight this document deserves, those two pages are the honest place to start.

Published award data shows who was funded. It does not show the full pool of unsuccessful applicants, and it does not show any funder's internal scoring. So this study can identify patterns, fit and rule breaches with confidence, and it cannot claim certainty about any individual funder's reasoning where that reasoning is unpublished. Where we reach the edge of the evidence, we say so on the line rather than writing around it. Six specific limits apply throughout:

• Our award records are drawn from each funder's data feed in the order the feed returns them, not as a random sample. They show what a funder pays for. They do not show how often, and they are not adjusted for inflation.
• Coverage by year differs sharply between funders. Our Funder E records run December 2024 to February 2026; our Funder C records cover 2009 and 2021 to 2023 with nothing in between; our Funder B records predate its current main programme. So the record counts in chapter 5 are not comparable between funders as measures of volume, and no claim in this study rests on comparing them that way.
• Two of the three refusal files are not applications. The Funder O file is that funder's blank template. The Funder I file is that funder's example form carrying your draft answers, and says so in its own header. We can therefore read those funders' rules and your prepared answers, but not what was finally submitted.
• The Funder O refusal feedback has not yet arrived. Our reading of that refusal is registered in advance, in section 6.5, so it can be scored rather than assumed.
• Comparator reserves are year-end snapshots from the accounts nearest each award, not the balance on the day the funder decided.
• Two funder pages refused our reader on 25 August 2026 and are cited from earlier reads or from the funder's own public announcements instead. They are named in appendix J.