Why Spain Startup Visa Applications Get Refused in 2026: The Full Typology

A Spain Startup visa can be refused twice over, by two different bodies, on two completely different kinds of grounds, and the single most useful thing after a refusal is knowing which one you got. ENISA (Empresa Nacional de Innovación) issues the innovation report and judges the business itself: is it innovative, is it scalable, is it early-stage, and can this team execute it. UGE-CE (Unidad de Grandes Empresas y Colectivos Estratégicos) processes the residence authorisation and judges the file: your means, your criminal record certificate, your insurance, your legal stay and the form of every document, with the authorisation formally granted or refused by the Dirección General de Migraciones.

The distinction matters because the two produce almost opposite prognoses. A UGE-CE refusal on documents is usually a fixable problem with a short path back. A negative ENISA verdict on the nature of the business is frequently terminal for this route, no matter how well the plan is rewritten. This article is the full typology of both: every recurring ground, what in the dossier actually triggers it, and which grounds are worth appealing rather than rebuilding. The procedural mechanics of appealing (deadlines, competent bodies, courts, silence rules) are a separate subject, covered in recurso de alzada versus reapplying; this article is about the reasons.

What ENISA is actually testing

ENISA does not require narrowly high-tech businesses. Under Ley 28/2022, the criteria are technological innovation or a differentiated business model, genuine scalability (revenue that can grow faster than costs), early-stage status, and founder capability to execute. Innovative is assessed relative to what already exists in the market, not in absolute scientific terms.

A software company that automates scheduling for restaurants can qualify. A restaurant cannot. That distinction, what the business does versus what market it serves, is where most refused applications go wrong.

ENISA ground 1: the activity is not innovative (HoReCa and lifestyle)

The most consistent refusal category in 2026 is hospitality: cafés, restaurants, boutique hotels, wine bars and similar lifestyle operations. These are not innovative by ENISA's definition however they are positioned.

The recurring pattern is a "concept café" submitted as a "food-tech platform" or an "AI-powered hospitality experience". Evaluators have seen it often enough to identify it quickly. The underlying test is whether core value is created by the physical service (preparing food, hosting guests) or by technology that could be deployed without the physical asset.

Typical wording: the described activity does not meet the innovation threshold required under Article 4 of Ley 28/2022; the business model replicates existing commercial activity without demonstrable technological or business-model differentiation.

Fixable? Only if the core product is genuinely software (a platform sold to other restaurants, a food-waste logistics tool, a reservation SaaS) and the hospitality element is one sales channel among several. If the business is a restaurant, the Startup Visa is not accessible.

ENISA ground 2: the concept is licensed (franchises)

Franchises fail on two grounds simultaneously: the concept is licensed rather than original, and the model is proven and replicated by design rather than early-stage. Either disqualifies independently.

Even a franchise from a strong brand in a fast-growing segment (plant-based food, co-working, language schools) will not pass. Applications here often present large addressable markets and favourable projections; evaluators check the legal structure first.

Typical wording: the described business arrangement is a licensed franchise; ENISA does not certify franchise operations as innovative startups under Ley 28/2022.

Fixable? Rarely. If you fully control the intellectual property and are expanding a proprietary concept internationally, and there is no franchise agreement in the legal sense, the application can be restructured. If a franchise agreement exists, the answer is no.

ENISA ground 3: real estate dressed as PropTech

Real estate investment or brokerage described in PropTech or SaaS terminology. This category grew after the Golden Visa was discontinued in April 2025 and founders began looking for another way into Spanish residence; the route map for that audience is in Spain residence permits after the Golden Visa.

Common shapes: a property management company as "an AI-driven real estate platform", a buy-to-let business as "a PropTech marketplace", a co-living operator as "a decentralised housing network". ENISA reads the revenue model behind the terminology.

Typical wording: the described activity is primarily real estate investment or brokerage; technological elements are ancillary to the core revenue model, which is property appreciation or rental income.

Fixable? If the technology is the actual product (software sold to property managers, a data platform for institutional investors, a compliance tool for short-term rental operators) and real estate is the vertical rather than the revenue model. The practical test: could the software be sold to a different industry? If yes, the technology is genuine. If it is inseparable from owning or managing specific properties, the business is real estate whatever the branding.

ENISA ground 4: linear cost structure

The least obvious category, and a meaningful share of refusals. Service businesses (consultancies, import/export traders, marketing agencies, offline retail) that present ambitious revenue projections on a fundamentally linear cost base.

The scalability test: can revenue grow substantially without proportional growth in headcount, physical assets or variable costs? A software company selling licences passes. A consultancy billing hours does not. An importer that grows by adding SKUs and warehousing does not. An agency that grows by hiring account managers does not.

Typical wording: the financial projections are commercially reasonable, but the business model is inherently linear; the company does not demonstrate the scalability characteristics required under Article 4 of Ley 28/2022.

Fixable? Sometimes. If the firm genuinely develops and sells proprietary IP (a methodology, a software tool, a training programme deliverable repeatedly without proportional cost growth), repositioning as an IP company can succeed. The diagnostic ENISA implicitly applies: if you signed ten clients tomorrow, would costs grow tenfold?

ENISA ground 5: not early-stage, or the wrong company shape

Less discussed and genuinely surprising to applicants: a company can be refused for being too established. The Startup Act framework is built for early-stage companies, and an entity that is already mature, profitable on a conventional model, or effectively a foreign group's local subsidiary does not fit the definition it is being measured against. Founders who present years of stable trading as their strongest evidence sometimes hand the evaluator the refusal ground.

The corporate shape can also break the case independently of the business: a structure where the applicant is not genuinely a founder or does not hold a substantive role, an entity incorporated purely as an immigration vehicle with no operating substance, or a co-founder appointment with no defined role and no real equity arrangement behind it.

ENISA ground 6: the plan is unevidenced

A genuinely innovative business can still be refused because the file does not demonstrate it. This is the ground that most rewards a rebuild, because the business does not have to change at all.

The recurring evidential gaps: founder capability asserted rather than shown (a CV with no connection to the sector or the technology); market analysis built from generic global figures rather than the Spanish market ENISA is comparing you against; no competitive moat, meaning nothing explains why this is defensible rather than merely first; financial projections without cost lines, which make scalability unassessable; and no Spanish economic contribution, no hiring plan, no local partnerships, nothing that answers why Spain. The section-by-section anatomy of a file that closes these gaps is in what each part of the ENISA business plan must contain.

What in the dossier actually triggers a refusal

Evaluators do not primarily react to prose. They react to artefacts, and the artefacts override the narrative. The recurring triggers:

• The CNAE activity code. Registering under hospitality, real estate or retail codes and then describing a technology company is the fastest self-inflicted refusal in the set.
• The objeto social in the incorporation deed. If it describes running a restaurant or managing properties, that is the legal statement of what the company does.
• A franchise or licence agreement in the annexes. One document, one ground.
• Letters of intent signed by related parties. Demand evidence from your own other company or from a co-founder's employer reads as fabricated traction rather than validation.
• Projections with no cost model. Revenue curves without a matching cost structure make the scalability test impossible to pass.
• Revenue lines that are rent or appreciation in an otherwise technology-flavoured plan.
• Traction claims that cannot be verified (user numbers, pilots, contracts) with nothing behind them.
• A templated plan. Files that read as generic, or as generated boilerplate with the country name swapped, are recognisable, and evaluators discount everything in them.
• Translation defects. A plan translated so that its technical claims no longer parse is judged on what it actually says in Spanish.

UGE-CE grounds: the residence layer

Even with a favourable ENISA report, the residence file can fail on its own terms. These grounds have nothing to do with your business.

Wrong means-test benchmark. The Startup and entrepreneur authorisation is indexed to IPREM, not to the minimum wage: 100% IPREM, about €600/month (€7,200/year) for the main applicant plus about 50% IPREM (€300/month) per family member. The widely quoted "200% SMI, about €34,188" belongs to the Digital Nomad Visa. A practical buffer of €30,000 or more makes the solvency review easier, but it is a recommendation, not the legal floor.

Wrong form of financial evidence. Property equity does not count. Nor does anything the reviewer cannot trace: unverifiable statements, accounts opened days before filing, or a balance that appears from nowhere with no history behind it.

The criminal record certificate. It must cover the countries where you have lived for the past five years, carry an apostille or full consular legalisation, and come with a sworn translation into Spanish. Validity windows are short, so this document should be timed to the filing date rather than obtained a year ahead. Format rules are strict, and an electronic copy where a legalised paper original is expected is a refusal, not a discussion.

Health insurance. Private cover from an insurer authorised to operate in Spain, with full coverage and no co-payments or waiting periods, is the safe format. Travel policies and policies with excesses are among the most common defects.

Legal stay and sequencing. The inland UGE-CE route requires you to be legally in Spain when you file. Filing after a Schengen stay has expired, or filing the residence authorisation before the favourable ENISA report exists, breaks the case on procedure rather than merits. The consular route grants a one-year visa; the three-year authorisation exists only on the inland filing.

Family documents. Marriage and birth certificates run the same apostille-then-sworn-translation chain as your own. A family member's broken document chain refuses that member, and sometimes stalls the whole file.

An unanswered requerimiento. Where the file is incomplete, the administration issues a subsanación requirement under article 68 of Ley 39/2015, generally ten working days. Missing that window converts a fixable gap into a refusal, and practitioners report that UGE has moved noticeably from issuing requerimientos toward issuing outright refusals through 2026.

Post-approval audits. An approval is no longer the end of scrutiny. In a dated assessment of 23 February 2026, the Spanish firm IG Abogados estimated that up to 70% of files in its own caseload were being audited retrospectively, with authorisations revoked where the holder had never registered with Social Security. That is a practitioner estimate from private caseload, not official statistics, and no separately named control unit appears in UGE's published structure. What the reporting converges on is the behaviour: stricter verification, and post-approval checks on whether you actually did what you said.

Appeal or reapply: the ground decides

Match the route to the ground you were actually given, not to how strongly you disagree.

The clocks: a recurso de alzada must be filed within one month of notification, and judicial review within two months of an express decision. Reapplying has no deadline, and a fresh cycle runs ENISA 10 to 20 working days plus UGE-CE around 20 working days with positive administrative silence. That arithmetic is why a rebuilt file usually reaches an answer before an appeal does. The full procedure, including which court actually hears these cases, is in recurso de alzada versus reapplying.

What your realistic chances actually are

There is no official published breakdown of ENISA or UGE-CE outcomes by refusal ground. Any article quoting a precise success rate for appeals or resubmissions is quoting something it cannot source, and you should discount it accordingly.

What can be said honestly is directional, and it follows the typology above. Documentary and procedural grounds at UGE-CE are the most recoverable: the eligibility was always there, and the correction is mechanical. Evidential weakness in a genuinely innovative business is usually recoverable too, because the rebuild is a writing and evidence problem rather than a business problem. Substantive innovation and scalability verdicts are the hard cases: overturning one means persuading a reviewer that a judgement call was wrong, and reviewing bodies rarely reverse themselves on judgement. Grounds that go to the nature of the business (a franchise agreement, revenue that is rent, value created by a physical location) are not odds at all; they are a wrong-route diagnosis.

Three questions give you a more useful estimate than any percentage. Did the refusal cite a document or the business? If you removed the disputed element entirely, would the business still exist? And has anything material changed since the first filing, other than the wording?

When the refusal is final

Some businesses cannot qualify however the application is structured:

• Genuine franchises where a franchise agreement exists
• Businesses whose revenue derives from a physical location or asset (cafés, hotels, most conventional retail)
• Businesses that depend entirely on the founder's personal reputation or specific relationships rather than a transferable product
• Real estate businesses where appreciation or rental income is the primary return

For these profiles the Startup Visa is not accessible, and continuing to file is a way of spending months to reach the same answer. The usual alternatives are Portugal D2, which applies a viability test rather than an innovation test, France Talent for project-driven founders, or, where the income profile fits, a different Spanish route entirely: the Startup Visa against the Digital Nomad Visa and the self-employed route sets out which of those matches which profile.

What a strong submission looks like in 2026

For contrast, the business types that consistently receive ENISA certification in 2026: early-stage SaaS with clear unit economics, hardware-software combinations with defensible IP, biotech and medtech with clinical validation, marketplaces addressing genuine matching inefficiencies at scale, and fintech or legaltech platforms. All share one structural characteristic: unit economics that improve with scale, and revenue that grows faster than cost.

If your business fits one of those profiles, the application is won or lost on how the submission is structured, section by section. That walkthrough is ENISA business plan for the Spain Startup Visa 2026, and the route overview is the Spain Startup Visa guide.

Relovisa reads the actual refusal grounds, maps them to the fastest viable path, and files the rebuilt case, including routes outside the Startup Visa when the innovation test is simply the wrong test for your business. Get a Spain Startup refusal review from Relovisa