Small business grants reviews: how to compare advisers

A credible grant-provider review says what was tested, which evidence was checked and what remains unknown. It compares like with like, separates provider claims from independent records and discloses every commercial relationship. A polished website or large claimed award total does not prove that a service fits one English business.

This guide to small business grants reviews sets out a review method rather than endorsing a supplier. The same approach applies whether the work is a grant application, an R&D tax credit claim or Patent Box support.

What to take away

A credible review states what was tested, which evidence was checked and what remains unknown.
Set criteria and pass or fail conditions before contacting any provider.
A Companies House record confirms filed information, not competence, solvency or grant success.
A provider's claimed success rate stays a provider claim unless its method was independently verified.
Report findings by evidence type and avoid one universal score across different sectors.

Define the reader and decision

Start with a specific use case: an English manufacturing SME seeking Innovate UK application review, a retailer researching local capital grants, or a funded company needing claim-management support. State business size, location, grant type, project stage and deadline.

Name the funder, not just the scheme type. For an English business that could mean an Innovate UK competition, Start Up Loans delivered by the British Business Bank, or a local authority capital grant; in the devolved nations, Business Wales, Scottish Enterprise and Invest NI run their own programmes. Claim work is separate — HMRC administers R&D tax credits, the Patent Box and the Research and Development Expenditure Credit — and strength on one funder's process says nothing about the others.

The decision might be whether to shortlist, request a proposal, run a paid trial or sign a contract. A desk review can support a shortlist. It cannot establish delivery quality that was never observed.

Write the exclusions. If the comparison covers innovation-grant advisers, do not imply that it assesses community, arts, charity or agricultural grant work.

Build the candidate set transparently

Publish how providers were found, the search date and the rules for inclusion. Useful minimums include a current public service page, an identifiable trading entity or practitioner, relevance to the defined grant type and evidence of serving UK organisations.

Do not create a "best" list from whichever companies appeared first in a search. Search visibility is not service performance. Keep an exclusion log for providers outside the scope, those with unverifiable claims and those that did not supply information.

If a provider pays, gives free access or supplies data, disclose it beside the relevant review. The CMA's reviews and endorsements guidance covers genuine reviews, incentives and the responsibilities of publishers, brands and content creators.

Before you act

Define the use case, business size, grant type and deadline.
Set pass or fail conditions before any contact.
Test the service with the same realistic brief for each provider.

Verify the entity

Record the contracting name, company number where applicable, registered jurisdiction, address, ownership, named contacts and public service terms. The Companies House search service is an official source for UK company records.

A Companies House record confirms filed information, not competence, solvency or grant success. Note the filing date and conduct proportionate financial, insurance and identity checks separately.

Distinguish a brand from the legal party. A group website may route the contract to a subsidiary or overseas entity. Review the entity on the proposed quotation and terms.

Separate evidence types

Label each item as:

first-party service description
contract or quotation
official register or professional-body record
direct test observation
customer reference checked by the reviewer
public user review
reviewer judgement

Do not merge them into one confident paragraph. A customer testimonial selected by the provider is not a representative sample.

Use official funder guidance to check assertions about process. UKRI's Innovate UK application guidance states that relevant applications must use the Innovation Funding Service and makes the lead applicant responsible for reviewing answers and submitting on time. An adviser cannot take away that responsibility through marketing wording.

Set criteria before contact

Choose criteria that matter to the use case. They may cover scheme experience, named team, method, capacity, evidence controls, conflicts, information security, fees, contract terms, post-award support and exit.

Define the scoring scale. A score of five for capacity might require named availability through the deadline and a replacement plan; a score of one might mean that staffing is unknown. Record evidence beside the mark.

Score	What it means for named team capacity	Evidence to record
5	Named consultant available through the deadline and a named replacement if they leave	Written confirmation, staff plan, contract clause
3	Named consultant available now but no replacement plan	Email confirmation, no formal plan
1	Staffing not identified or only a generic team named	No names, no availability statement

Set pass or fail conditions for integrity, legal entity, security or a binding procurement rule. A high total score should not compensate for advice to submit false information.

Test the service where possible

Use the same realistic brief for each provider. Remove sensitive information unless a proper agreement and secure route are in place. Ask the provider to identify poor fit, missing evidence, risks and the work required from the applicant.

For software, test account access, permissions, version history, collaboration, exports, backup and deletion. For advisers, a paid sample review can test clarity, challenge, factual discipline and usefulness. Record the test date, plan, data, people, duration and constraints.

Set a minimum test you can repeat: the same written brief, the same deadline, the same evidence pack and the same question set for every provider. Anything shorter compares sales manner rather than advice.

Do not claim an end-to-end test after a sales call or free demonstration. If no purchase or live delivery occurred, use "desk review" and leave performance unscored.

Check specialist competence

Ask which schemes and stages the named team supports. Verify relevant qualifications where tax, accounting, legal, technical or evaluation advice is offered. A successful application in one competition does not prove expertise across all funders.

Where tax credits, accounts or legal drafting are involved, check the individual or firm on the relevant professional register, such as ICAEW or ACCA for accountancy or the Chartered Institute of Taxation for tax advice.

Request a redacted work sample or talk through the method. Look for evidence mapping, eligibility discipline, financial consistency, challenge and a clear boundary between applicant facts and adviser drafting.

Check capacity through the actual deadline and post-submission period. Ask who replaces a named consultant and whether subcontractors will access data.

Examine fees on the same basis

Obtain a written scope and full fee schedule. Identify fixed fees, day rates, retainers, success fees, expenses, VAT and post-award charges. Define what counts as success and when payment becomes due.

Model no award, reduced award, delayed payment and full award. Compare the same tasks. One quotation may include financial modelling and revisions while another prices them later.

Check whether adviser cost is eligible under the specific scheme. Do not assume a fee becomes fundable because the provider has supported similar work.

Review claims and reviews critically

For a claimed award total or success rate, ask for the period, population, denominator, excluded cases, provider role and independent assurance. Record when the provider declines or cannot support the claim.

Public review platforms can reveal themes, but identity, selection and commercial incentives may be uncertain. Read recent positive and negative records, provider responses and the platform's method. Do not calculate a fresh rating from a handful of chosen comments.

Named platforms matter because each applies its own verification and moderation rules. Trustpilot, Feefo and Google reviews let businesses invite and respond to feedback, so check the platform's method and the provider's invitation practice before treating a score as evidence; none of them confirms competence or regulatory standing.

The CAP Code section on misleading advertising requires documentary evidence for objective advertising claims within its scope.

Assess security and data protection

Map the financial, personal and commercially sensitive information the supplier receives. Check authentication, access, hosting, subprocessors, retention, export, deletion and incident notification. Review the actual contract and product plan.

Use the ICO UK GDPR guidance and NCSC small organisations guide to frame current UK checks. A badge on a website is not enough evidence for the configured service.

Test how a user leaves and whether the business can retrieve applications, comments, evidence and audit history in usable form.

Review the contract

Check scope, deliverables, applicant responsibilities, fee triggers, expenses, timing, accuracy, intellectual property, confidentiality, personal data, conflicts, subcontracting, liability, cancellation, complaints and exit.

The applicant should retain approval of every submission and declaration. Do not sign terms that permit material changes without review or imply that funding is guaranteed.

Where the customer acts mainly outside a trade or profession, consumer law may apply. The CMA's fair-contract guidance addresses trader-to-consumer terms. Many grant advisory engagements are business-to-business, so obtain legal advice for the actual relationship.

Write the verdict narrowly

State which use case the provider appears to fit, which conditions remain and which areas could not be tested.

A useful verdict might say that public information supports shortlisting a provider for innovation-grant application work, while pricing, named capacity, security and contract terms remain subject to proposal. That is more honest than "best overall".

Give the provider a chance to correct factual errors without allowing it to buy, approve or suppress the editorial conclusion. Log changes and publish material corrections.

Keep the review current

Record the research date, reviewer, plan tested, sources and conflicts. Set an expiry based on change risk. Recheck ownership, staff, service scope, fees, terms, privacy and material claims before publication.

Monitor complaints and credible new evidence after release. Update the review visibly when the conclusion changes. Do not refresh the date without repeating the stated checks.

Preserve the review record

Store the candidate list, criteria, evidence captures, correspondence, test plan, raw observations, scores, reviewer notes and published version. Keep commercial records secure and respect lawful retention and confidentiality requirements.

The record should allow another editor to reproduce each material factual finding. It should also show why a provider was excluded or left unscored. If the evidence disappears from the web, retain the permitted dated capture and full citation rather than silently replacing the claim.

Separate corrections from later reassessments. A correction fixes an error in the original review. A reassessment reflects new evidence, a changed product or another test period. Readers should be able to tell which occurred.

Common questions

Can a desk review establish how well a grant adviser delivers?

No. A desk review can support a shortlist, but it cannot establish delivery quality that was never observed. If no purchase or live delivery occurred, the article says to use desk review and leave performance unscored. Do not claim an end-to-end test after a sales call or free demonstration.

What should a reader ask for when a provider claims a high success rate?

Ask for the period, population, denominator, excluded cases, provider role and independent assurance. Record when the provider declines or cannot support the claim. A customer testimonial selected by the provider is not a representative sample, and a claimed success rate remains a provider claim unless independently verified.

Who stays responsible for submitting an Innovate UK application?

The lead applicant. UKRI guidance states that relevant applications must use the Innovation Funding Service and makes the lead applicant responsible for reviewing answers and submitting on time. An adviser cannot take away that responsibility through marketing wording. The applicant should also retain approval of every submission and declaration.