"Unable to assess impact model" is not a verdict on whether your work creates change. It is a translation failure. The funder opened your proposal, looked for the model that links what you do to what changes, and could not find it in the form they evaluate. Decode that one sentence into the four specific things it actually means — no Theory of Change, no outcome data, no intervention logic, no scalability markers — and a rejection letter becomes a communications diagnostic you can act on.
The email arrives on a Friday afternoon. Two lines. "We appreciate the work your organisation is doing. However, we are unable to assess your impact model and cannot take your proposal forward at this time."
The director forwards it to the program team with a question mark. The program team forwards it to the fundraising lead with the same question mark. Nobody knows what to fix, because the sentence does not say what is wrong. It says what the funder could not do.
This is the most common rejection line in Indian institutional funding — and it is the least understood. CSR teams, foundation program officers, and family philanthropy offices use a version of it because it is polite, defensible, and avoids naming a specific weakness that could be argued with. It is also, almost always, the wrong diagnosis the NGO hears. The phrase "NGO impact model" sounds like a technical assessment of your impact. It is not. It is a comment on whether your model was legible in the document.
The funder did not say your impact does not exist. The funder said they could not find it.
Those are different problems, with different fixes. The first is a program problem. The second is a communication problem. We've written about why NGOs with the strongest field work lose funding to organisations with sharper communication — this is the mechanism, one rejection letter at a time.
Because Indian funders are now legally and structurally required to assess impact, and most proposals do not give them anything to assess.
Under Rule 8(3) of the Companies (CSR Policy) Amendment Rules, 2021, every company with an average CSR obligation of ₹10 crore or more in the preceding three financial years must conduct an independent impact assessment for projects with an outlay of ₹1 crore or more. The assessment has to be done by an external agency. The result goes before the board and into the Annual Report. A CSR program officer who funds your project is, in effect, betting their own compliance position on your model being assessable.
If your proposal does not let them see the model, they will not take that bet. The polite version of that refusal is "unable to assess impact model."
The sector data tracks this. The CSRBOX India CSR Outlook Report 2023 found that insufficient reporting and impact communication was the most common funder-side complaint about implementation partners — cited by roughly 60% of companies surveyed. An ICOR 2023 summary by a CSR practitioner puts the same finding more bluntly: 70% of CSR heads say lack of transparency and reporting is the biggest barrier to funding grassroots NGOs.
The 2024 and 2025 editions show the bar rising. Funders now explicitly demand scalable, replicable models (66%) and data-driven impact assessments (61%). The assessment muscle on the funder side has grown. The communication muscle on the NGO side, in most cases, has not.
A funder writing "unable to assess impact model" is collapsing four separate gaps into one sentence. They will not separate them for you — doing so would amount to free consulting. But the gaps are separable, and each has a specific fix.
| What the funder wrote | What they meant | What to fix |
|---|---|---|
| "Unable to assess impact model" | "I could not find your Theory of Change" | Articulate the causal chain: when X, then Y, because Z |
| "Unable to assess impact model" | "I could not see outcome data, only activity counts" | Report outcomes with baselines and targets, not outputs |
| "Unable to assess impact model" | "I could not tell what is specific to your intervention" | Show your intervention logic — the mechanism that makes your model work |
| "Unable to assess impact model" | "I could not see how this scales" | Add replicability and cost-per-outcome markers |
Each row is a different failure. Each has a before/after. Each is fixable without touching the program.
This is the most common reading, and the easiest to fix. The funder looked for a causal chain — when this happens, that changes, because of this mechanism — and found a list of activities instead.
A 2026 guide for CSR heads on how to choose NGO partners frames the checklist question funders actually ask:
"does the NGO's theory of change reflect a genuine understanding of how the development challenges you are addressing actually work — including their interconnections with other dimensions of deprivation?"
If your proposal does not answer that question, in those terms, the funder moves on. They do not have the bandwidth to reconstruct your Theory of Change from your activity list.
Before — what got rejected:
Our intervention empowers adolescent girls in rural Rajasthan through life skills education, menstrual hygiene awareness, and community engagement, leading to lasting change in their lives and communities.
After — what gets assessed:
When out-of-school adolescent girls in rural Rajasthan complete a 12-month life-skills and menstrual health curriculum delivered by trained community facilitators (not external trainers), their school re-enrollment rises from a baseline of 22% to 68% within one academic year. This happens because the curriculum simultaneously addresses three barriers — household permission, safety of commute, and access to hygiene materials — that earlier programs addressed in isolation. The facilitator model replicates at ₹4,200 per girl per year, one-fifth the cost of external-trainer models.
Same program. One is a platitude. The other is a model a funder can assess. The five most common Theory of Change mistakes and how to fix each one are here.
The funder needed baselines, targets, and outcome indicators. The proposal gave them workshops conducted, beneficiaries reached, kits distributed. That is an output log, not an impact model.
The Sambodhi and Dasra "Measuring What Matters" study (2025) found that 73% of MEL data collected by Indian NGOs flows upward to donors — but only 18% informs internal decisions. Most NGOs are not short of data. They are short of the right data, structured the right way, in the proposal.
The ThinkCap guide on CSR impact assessment names the failure precisely:
"An assessment that reports '78% of beneficiary children are now attending school' is not an impact assessment; it is a status survey." — ThinkCap Advisors, CSR Impact Assessment in India: The Complete Guide 2026
A status survey tells you what is. An impact assessment tells you what changed, and what produced the change. Funders assessing your model need the second. Proposals that ship the first get the one-line rejection.
Before — what got rejected:
We distributed 1,200 sanitary pad kits and conducted 48 menstrual hygiene workshops across 16 villages, reaching 3,000 adolescent girls.
After — what gets assessed:
Baseline (2025): 22% school attendance among adolescent girls in 16 target villages; 14% reported regular menstrual-health product use. Target (2027): 65% attendance; 70% regular product use. Mechanism: school-based distribution plus peer educator model. Cost per girl per year: ₹4,200.
Even when outcomes are present, the funder needs to see why your intervention — and not the general drift of the village, the government scheme, or the next NGO — produced them. This is the attribution question, and it is where most proposals go quiet.
The IDR piece on reimagining fundraising for nonprofits captures the funder-side shift:
"Donors are also starting to demand rigorous due diligence and more measurable outcomes, besides looking into long-term theory of change and sustainability."
"Rigorous due diligence" includes the question: what is your model's specific contribution to this change, and how do you know? The Smile Foundation guide makes the same point from the funder side: outputs are what was delivered, outcomes are what changed, impact is the portion of that change attributable to the programme. A proposal that reports outcomes without an attribution argument reports impact it cannot defend.
Before — what got rejected:
Our women's collectives have led to a 40% increase in household income across our program villages.
After — what gets assessed:
Women in our 18-month collective program reported a 40% median income increase. We attribute this to the collective savings mechanism plus market-linkage training, not general economic conditions, because control villages in the same blocks without collectives reported a 6% income increase over the same period. The counterfactual is small but it is there.
Honest, specific, and assessable. The funder now has something to evaluate.
Institutional funders are investors. They want to know whether the model works somewhere else, at what cost, and under what conditions. The 2025 ICOR data shows 66% of funders now explicitly demand scalable, replicable models. A proposal that proves a pilot but does not show the scale logic gets the one-line rejection even when the pilot is excellent.
The Smile Foundation guide again, from the funder side:
"The ability to scale, that is, to maintain programme depth while extending geographic reach, is a specific organisational capability"
It is a capability. It needs to be demonstrated in the proposal — not asserted.
Before — what got rejected:
Our model can be replicated across India.
After — what gets assessed:
What replicates: the facilitator training protocol (6 weeks, validated across three states with different languages). What is context-dependent: the livelihood curriculum (co-designed with local markets). Replication cost: ₹3.5 lakhs per new district in year one, declining to ₹1.2 lakhs by year three as facilitator training cascades internally.
If the gap was specific, why did the funder not name it? Two reasons. The program officer reviewed 80 to 200 proposals that cycle and does not have the bandwidth to write a custom diagnostic for each rejection. And naming a specific gap creates an argument — "your Theory of Change is unclear" invites a reply defending it. "Unable to assess impact model" closes the conversation cleanly and moves to the next proposal.
This is throughput, not malice. The decoding work falls on you. Treat the rejection letter as a free communications audit and reverse-engineer which of the four gaps it most likely names. This is also why a 90-day fundraising sprint cannot fix the problem — a sprint raises cash from retail donors who decide on emotion, and does nothing for the CSR program officer who needs to assess a model.
1. Re-read your last rejection letter against the four-row decode. Map the one sentence to the gap it most likely names. You will usually find one or two rows where the proposal was visibly weak. Start there. The field-stories-to-funder-metrics framework is here.
2. Rewrite your Theory of Change in the "when X, then Y, because Z" form. If a funder cannot find your causal chain in 60 seconds, that is the gap. The five most common Theory of Change mistakes are here.
3. Add baselines and targets to every outcome you report. If you report a 40% income increase, add what they earned before, the target, and the counterfactual. Why NGOs with rich field data still lose funding explains why this data usually exists but never reaches the proposal.
4. State your attribution argument explicitly. Say what your intervention produced, what it did not, and how you know. A model that claims everything claims nothing.
5. Add a replication cost and a context-dependency section. Show what scales, what is local, and what it costs to replicate. This is the section that turns a pilot into an investment case.
6. Run your website through the free Website Impact Analyzer. The proposal is one surface. The website is the surface a program officer visits after reading it. Run the analyzer here →
Re-read the proposal against the four-row decode in this post. The sentence usually means one of four things: no Theory of Change, no outcome data, no intervention logic, or no scalability argument. Your impact being real is not in question. Whether the funder could see it in the document is. Fix the document, not the program.
It can be either, and the letter will not tell you which. If the funder has funded similar work before, re-apply in the next cycle with a rewritten proposal. If the funder has no history with your sector, treat it as a no and move on. The decoding work is the same either way — the next funder will use the same sentence.
A one-page Theory of Change with causal links, a baseline-endline outcome table with targets, an attribution argument (including a counterfactual if you have one), and a replication cost. Most of this fits in a two-page annexure. If your proposal does not have these as separate, findable sections, the funder does not have them either.
Theory of Change is the causal logic — when X, then Y, because Z. Impact model is the broader package a funder assesses: Theory of Change plus outcome data plus intervention logic plus scalability. A strong Theory of Change with no outcome data still leaves the funder unable to assess the model.
Sometimes, but rarely. The proposal is the document the program officer scores against. The annual report is corroboration they reach for if the proposal is borderline. If the proposal does not surface the model, most program officers will not go hunting for it in a 60-page annual report. Fix the proposal first.
"Unable to assess impact model" is the most useful rejection line in Indian institutional funding, once you stop reading it as a verdict. It is a communications diagnostic in one sentence. It tells you the funder could not find your Theory of Change, your outcome data, your intervention logic, or your scalability argument — in some combination — in the form they evaluate.
The work is real. The model is real. The translation is what failed. Fix the translation, and the next funder does not need to write that sentence.
If this post decoded a rejection letter you have received, the next step is translation — taking the model your program team already runs and structuring it in the language funders assess. Read the field-stories-to-funder-metrics framework →
Related Reads: