Kit Digital, Kit Consulting, AI applied with judgement, digital transformation of industrial and agri-food SMEs, AI governance under the EU AI Act. Modernising without losing identity or composure in the face of hype.
Executive summary · TL;DR
A Spanish SME in 2026 has three combinable digitalisation levers: Kit Digital (up to EUR 29,000 in Red.es vouchers for technology solutions), Kit Consulting (up to EUR 24,000 in strategic advisory) and applied AI, which has moved from experiment to productive technology in marketing, sales and operations. The rule in 2026 is no longer "which tool do I use" but "which processes do I automate with AI while respecting the AI Act and GDPR, without homogenising the brand and without blowing up operating costs". A typical end-to-end SME digitalisation project costs EUR 15,000-60,000 and is co-financed via vouchers covering 40-70%. My approach adds two layers rarely seen together: the rigorous one (compliance, governance, ISO 42001, AI Act) and the strategic one (brand judgement, honest metrics, no falling for technological fascination).
Operational definition
Digitalisation is the transformation of processes, products and business models to incorporate digital technology with a measurable objective — productivity, quality, customer experience or new revenue — without losing the company's operational identity.
The concept is saturated by two abuses. On one side, the digitalisation consultancy sells "transformation" when what it delivers is a CRM and a new website. On the other, the AI discourse asks adopters to take it all on the diffuse promise of "being more competitive". Neither serves a Spanish SME of 10-200 employees with a finite budget.
My approach starts from three principles. First, digitalisation is means, not end: if it does not improve a business metric (productivity, conversion, quality, customer satisfaction), it is dropped regardless of how modern it looks. Second, public-funding levers (Kit Digital, Kit Consulting, MAPA, regional) are legitimate but must not dictate the strategy — first decide what is needed, then look at which voucher co-finances it. Third, every AI deployment is paired with governance, compliance (AI Act, GDPR) and real impact metrics, not demos.
Structure of the area
SME digitalisation in 2026 organises around five axes that cover almost any reasonable project. Each one has its lever map, its criteria and its traps.
Kit Digital is the Red.es programme funded by Next Generation EU that puts vouchers in the hands of SMEs and self-employed workers to redeem with approved Digital Agents. There are five segments in 2026 by size: Segment I (10-49 employees, EUR 12,000), II (3-9 employees, EUR 6,000), III (self-employed 0-2 employees, EUR 2,000-3,000), IV and V for microenterprises. The procedure requires a digital certificate or Cl@ve, registration on acelerapyme.es, a digital-maturity test, choice of Agent, signing the agreement, execution and subsequent justification.
The challenge is not requesting the Kit but using it well and justifying it without clawback. I cover both in two guides: the complete Kit Digital 2026 guide for SMEs and self-employed workers with segments, eligible solutions and procedure; and how to justify Kit Digital without clawback with the documentary mistakes Red.es detects on review and how to harden against them.
Kit Consulting is the Red.es 2026 programme (Order TDF/38/2026) covering up to EUR 24,000 per SME in strategic advisory across six areas: cybersecurity, AI, internationalisation, data, sustainability and processes. The difference from Kit Digital is essential: Kit Consulting pays for the thinking and the plan, not for implementation. That is why it pays to combine it with own budget to implement what the advisory recommends — a voucher that ends in a report produces no ROI.
I unpack this in the 2026 Kit Consulting guide and, for the highest-demand case, in the specific entry on Kit Consulting applied to cybersecurity and ISO 27001 certification, where I explain how to combine it with private implementation budget to reach certification.
Productive AI in SMEs concentrates on five use-case families with measurable ROI: RAG assistants over proprietary documentation (cut internal Q&A time by 30-50%); administrative-task automation with n8n, Make or Power Automate (one FTE freed per 4-6 processes optimised); AI for marketing (personalisation, lead scoring, +20-35% conversion); AI for industrial operations (predictive maintenance, visual quality with vision models, -15-30% defects); and vertical agents (customer chatbots with guardrails, legal or clinical assistants). Average time to impact: 3-6 months if the use case is well chosen.
The recurring mistake is starting from the tool (Copilot, ChatGPT Team, GPT-4) rather than the use case. I recommend the reverse order: map processes with high repetition and low variability, calculate monthly hours, prioritise by ROI and only then select tooling. For the wineries sector specifically, see winery 4.0 · ERP, traceability and AI in Ribera del Duero.
The EU AI Act (Regulation EU 2024/1689) became fully applicable in August 2026. It defines four risk levels: prohibited, high, limited and minimal. Most SME use cases fall into "limited" — requiring transparency (informing the user they interact with AI), clear information and internal documentation. Annex III cases (HR, credit, education, critical infrastructure, law) fall into "high" and additionally require a management system, conformity assessment, registration and human oversight. Fines up to 7% of turnover. AESIA is the Spanish competent authority.
The operational instrument is ISO 42001 · the first international AI management-system standard, providing a framework for governance, life-cycle management, impact assessment and continuous improvement. For the urgent compliance plan see EU AI Act 2026 · SME compliance in 90 days with focus on Annex III cases.
When an SME tackles an end-to-end transformation — not a one-off project — it needs an 18-36 month roadmap with prioritised phases, KPIs per phase and project governance sustained by management. Without that frame, projects atomise into pilots without integration and the transformation remains in slide decks.
I cover the methodology in digital transformation of companies · roadmap. For the winery sector, where vertical ERP, IFS/BRC traceability, vineyard and cellar IoT, DTC e-commerce and predictive AI converge, the specific entry is winery 4.0.
Consultant entry phase
Three common entry modes in digitalisation projects. Sign one with clear expectations.
Digital-maturity diagnosis (3-5 weeks). The company wants an objective starting point and a 12-24 month plan. Output: process and data map, evaluation of current stack, use cases prioritised by ROI, indicative budget and co-funding plan (Kit Digital, Kit Consulting, regional aid). Typical price: EUR 3,500-8,000. The most useful entry format and a fit for most Kit Consulting tracks.
Project accompaniment (3-9 months). The company has a defined project (enterprise RAG, n8n automation, ERP, AI in operations) and needs independent judgement to govern vendors, validate deliverables and keep direction. Typical price: EUR 2,000-6,000/month depending on hours committed. The most cost-effective mode when the client wants protection from tech vendors with misaligned incentives.
AI and compliance audit (2-4 weeks). The company wants to validate that its current AI use complies with the AI Act and GDPR, identify uninventoried systems and design governance policies. Typical price: EUR 4,000-9,000. The usual entry when AI has been deployed tactically without a governance frame.
Go deeper
A selection of digitalisation articles from the site, organised by sub-cluster.
Segments, eligible solutions, procedure and mistakes to avoid.
Read guide → Kit DigitalKey documentation and mistakes that cost the already-collected voucher.
Read guide → Kit ConsultingEUR 24,000 voucher in strategic advisory for SMEs.
Read guide → Kit Consulting · CyberHow to combine the voucher with private budget to reach ISO 27001.
Read guide → EU AI ActOperational plan to comply with the European regulation focusing on Annex III.
Read guide → ISO 42001First international AI-management-system standard.
Read guide → Winery 4.0Vertical ERP, traceability, IoT, AI and GDPR for DO wineries.
Read guide → TransformationPhased methodology, KPIs per phase and project governance.
Read guide →Methodology
Five common steps in digitalisation projects, both in diagnosis format and in implementation accompaniment.
Process, data and current-stack inventory. Internal capabilities assessment. Lever map with Kit Digital / Kit Consulting where applicable. 2-3 weeks.
Use cases ordered by hour savings, quality improvement or commercial impact. Technical-feasibility and compliance filter. 1-2 weeks.
Identification of approved Digital Agents, proposal comparison, contractual fit and justification plan. 1-3 weeks.
Accompaniment during execution by vendor or internal team. Fortnightly sessions. Change management. 2-6 months by scope.
Rollout of AI Act policies, registries, human oversight and impact metrics reviewed quarterly. Continuous improvement.
Recurring mistakes
Failed digitalisation projects share five identifiable patterns. Cataloguing them lets the reader avoid them without paying for the consultancy that catches them.
Starting from the tool, not the use case. Adoption of Copilot 365 or ChatGPT Team is decided because "it's the next thing" without mapping which concrete processes will use the tool. Six months on, 70% of licences are underused. Fix: process inventory, ROI prioritisation, pilot with five users, gradual scaling.
Mistaking Kit Digital for digitalisation strategy. The SME applies for what the voucher covers (website, CRM, basic cybersecurity) and considers itself digitalised. But the voucher covers standard solutions, not transformation. Fix: use Kit Digital for tactical solutions and Kit Consulting for the real strategic plan, adding own budget where needed.
Deploying AI without governance. A chatbot, scoring system or content-generation tool is adopted without inventory, AI Act risk classification or internal policies. When the first inspection or first reputational incident appears, there is no frame of reference. Fix: AI inventory from the first adoption, risk classification, basic policies and ideally ISO 42001.
Outsourcing transformation to a vendor with misaligned incentives. The SME hires the integrator of an ERP to lead the "digital transformation". The integrator optimises for licence sales and billable hours, not for value to the SME. Result: over-scoped and over-sized project. Fix: independent consultant who governs the vendor, with interest aligned to the client not the seller.
Poorly justifying Kit Digital. The project is executed but the justification documentation does not meet Red.es formal requirements. Result: clawback of the collected voucher. One of the most expensive mistakes of recent years. Fix: plan justification from the start, with evidence generated in parallel with execution and a formal review before closing.
Relationship map
Digitalisation crosses every site practice. Three critical crossings.
Digitalisation + compliance. Every technology deployment creates duties: GDPR if personal data are processed, AI Act if AI is used, ENS if ICT services are supplied to the public sector, NIS2 if the company falls in scope, ISO 27001 where sensitivity exists. Digital-project design must include the compliance plan from day one. See the compliance area.
Digitalisation + marketing. A large share of current marketing productivity and precision comes from the technology stack — automation, generative AI, scoring, personalisation. Without the right digital base, marketing remains channel tactics with no scale. But digitalising marketing requires specific governance (consents, GDPR, AI Act limited). See the marketing area.
Digitalisation + operations. In industrial and agri-food SMEs, digitalisation runs through ERP, MES, traceability and, increasingly, AI in operations (predictive maintenance, visual quality control). It is the area with the largest demonstrable return when tackled methodically, and the most painful when done as fashion.
Frequently asked questions
Kit Digital is the Red.es voucher funded by Next Generation EU for digitalising SMEs and self-employed workers. There are five segments in 2026: Segment I (10-49 employees, EUR 12,000), II (3-9 employees, EUR 6,000), III (self-employed 0-2 employees, EUR 2,000-3,000), IV and V for microenterprises. Steps: digital certificate or Cl@ve, registration on acelerapyme.es, mandatory digital-maturity test, choice of an approved Digital Agent, signing the agreement, project execution and rollout, and subsequent justification. Red.es pays the Agent directly.
Productive AI in SMEs concentrates on five use-case families with measurable ROI. RAG assistants over proprietary documentation (cut internal Q&A time by 30-50%). Administrative-task automation with n8n, Make or Power Automate (one FTE freed per 4-6 processes). AI for marketing (personalisation, lead scoring, +20-35% conversion). AI for industrial operations (predictive maintenance, visual quality, -15-30% defects). Vertical agents (chatbots with guardrails, professional assistants). Average time to impact: 3-6 months if the use case is well chosen.
Kit Consulting is the Red.es 2026 programme (Order TDF/38/2026) subsidising up to EUR 24,000 per SME in strategic advisory across six areas: cybersecurity, AI, internationalisation, data, sustainability and processes. It covers 100% of the advisory but not implementation. Recommendation: combine it with own budget to implement what the advisory recommends. Beneficiaries: Segment A, B or C SMEs. Window: open until funds run out.
The AI Act (EU Regulation 2024/1689) became fully applicable in August 2026. Steps for a Spanish SME: inventory AI systems in use (chatbots, scoring, RAG, generative); classify by risk (prohibited, high, limited, minimal; Annex III lists the high-risk ones); apply obligations per classification (most marketing AI falls into "limited" — transparency + user information); document policies and governance; and ideally implement ISO 42001 for formal governance. Fines up to 7% of turnover. AESIA is the Spanish competent authority.
Four options by volume. Microsoft Dynamics 365 Business Central (EUR 25,000-200,000 + EUR 70-110/user/month), flexible mid-SME option. Sage 200/X3 (EUR 18,000-120,000), strong in the Spanish industrial sector. Odoo (open source, from EUR 4,000-15,000 + EUR 10-25/user/month), ideal for small SMEs with in-house IT. SAP Business One (EUR 50,000-300,000), only mid-market 100+ employees with international growth ambitions. For wineries and agri-food there are vertical solutions (VinoTEC, Isagri Vino) on top of Dynamics or Sage.
Yes, no exceptions. The voucher is redeemed exclusively with approved Digital Agents. The SME does not receive money — Red.es pays the Agent directly. If you go with a non-approved supplier, the cost is not subsidised. The official list is at acelerapyme.gob.es and is updated frequently.
Generally yes, because they cover different concepts: Kit Digital = technology solutions, Kit Consulting = strategic advisory. Regional aid (Canary Digitalisation line, Pyme Innova Castile and León, regional tax credits) usually requires not having received Next Gen funding for the same exact concept. Always review the specific terms of the regional programme.
Copilot 365 (Microsoft, EUR 28-30/user/month) integrates AI into Word, Excel, PowerPoint, Outlook and Teams. Ideal if you already pay for Microsoft 365. ChatGPT Team (USD 25-30/user/month) offers pure AI without native Office integrations. For everyday administrative cases, Copilot wins on integration. For analysis and free-form exploration, ChatGPT. For RAG over proprietary documentation, neither on its own — an enterprise RAG platform is needed (Plain Concepts, Sngular, custom integration with vector store).
Three bands. Quick wins (Copilot assistant + basic n8n automation): EUR 5,000-15,000 in 4-8 weeks. Productive solution (enterprise RAG over proprietary corpus + vertical AI agent): EUR 25,000-80,000 in 3-6 months. Operational transformation (AI in industrial operations with computer vision + MLOps): EUR 50,000-200,000 in 6-12 months. The first two bands are usually co-financed with Kit Consulting + Kit Digital + own budget.
First session at no cost. Tell me the context and, if we fit, I'll prepare a tailored proposal within 5 days.