Internal or External Trainer? Research Data, Pros and Cons

Why this choice matters (especially in IT)

In IT, training is not a “nice-to-have” - it’s a way to maintain productivity in a world where tools change almost every day: Microsoft Copilot, Copilot Studio, Google Workspace Flows, Gemini, new features in Microsoft 365, Power BI, AI Agents, RAG, new security standards (AI-assisted attacks, deep fake), and constant changes in marketing (e.g. GEO) and analytics (e.g. filtering bot activity).

Right now, we’re seeing a true "arm race" for dominance in AI adoption between companies like Microsoft, Google, OpenAI, and Anthropic, which makes the pace of change overwhelming even for experts.

That’s exactly why decisions about a training strategy and how to onboard people into new technologies have become critical again for the future of companies. Historically, many organizations have relied mainly on internal training - but with changes happening this fast and this often, it’s worth honestly reassessing when that model still works best, and when it makes sense to add external support.

Research data

Before we go into the pros and cons, let’s first look at the data showing what this split looks like today:

• World - according to the Association for Talent Development (one of the largest professional industry organizations), the State of the Industry 2025 report (source) divides employee learning-related spending into three buckets:
• • 53% - internal costs (also including administrative costs of delivering external training),
  • 29% - external providers (learning suppliers),
  • 18% - reimbursement of formal education and certifications (tuition reimbursement).
• Poland - The most recent nationwide data from Statistics Poland (GUS) that directly split internal vs external courses come from the CVTS survey from 2020 (the next wave is 2025, but publications usually come with a delay). That’s why for 2024–2025 we will rely on newer sectoral and regional studies.
• • Civil service (dane.gov.pl, 2024): in the training data for the civil service, for the IT topic area the share of internal training is 26%, and external 74%. This is not “the whole economy,” but it’s a current signal that for fast-changing technical topics, organizations more often buy competence from the market.
    (Dataset: “Training in the civil service” → “Percentage share of internal and external trainings by topic area in 2024”)
  • Employer study (Institute of Sociology, University of Zielona Góra, September 2025, source): the quantitative stage used CAWI, with 159 companies from the Lubuskie voivodeship (micro 53.5%, small 23.3%, medium 15.1%, large 8.2%). Among the most frequently indicated forms of competence development (multiple answers possible), the leading ones are:
  • • external training 70.1%,
    • internal training 49.6%,
    • online courses / e-learning 45.3%.
  • “Whole economy” benchmark (GUS, CVTS 2020): among companies that provided training, external courses occurred in 54.0% of companies, and internal in 44.6% (often in parallel). This is still the only broad, comparable reference point - but it’s worth remembering it’s from before the pandemic and the AI boom.

It should be noted that ATD data refers to % of costs, while the Polish data refers to % of trainings, and additionally ATD’s internal costs include not only the L&D team (including trainers) but also the costs of handling external training.

However, the large discrepancy suggests that in Poland we use external training more often (around 50–74%) than globally (29% of costs according to ATD), which - contrary to appearances - may turn out to be our strength given today’s pace of change.

In the end, the question is not “what is better”, but “what is better in a given context.”

Internal trainer (in-house) - when it works best in IT

Biggest pros

• Knowledge of your environment: specific Microsoft 365 tenants, security policies, processes, data architecture, SQL/BI standards, real constraints (e.g., permissions, licenses, network).
• Better transfer to the job: it’s easier to do follow-ups, consultations, fixes using “live examples,” and reinforcement after training (and this often determines the final outcome).
• Scalability for recurring topics: onboarding, “company standards,” recurring process changes, training new people every month.
• Faster response: mini “here and now” trainings after rolling out a new tool or changing a process.

Biggest cons

• Risk of a “bubble” and outdated habits: in IT it’s easy to teach “how we always did it,” not necessarily “how it’s best now.”
• Narrow perspective: fewer comparisons across companies, fewer benchmarks, fewer “best practices” from the market.
• Hidden cost (opportunity cost): a good engineer/analyst as a trainer often stops doing key project work.
• Not every expert can teach: lack of methodology, exercises, working through mistakes, and a sensible structure (a common killer of technical training).

Typical IT scenarios where in-house wins

• Onboarding into your data stack (Power BI + sources + governance + naming conventions + repositories).
• “How we do it here”: SQL standards, code review, deployments, security, M365/Entra policies.
• Product/process training: internal tools, company dashboards, company marketing playbooks.

External trainer - when it gives the biggest advantage in IT

Biggest pros

• Up-to-date market knowledge and experience from many implementations: especially for M365, Power BI, AI, marketing automation, where changes are fast.
• Faster start in new areas: “from zero to a solid level” in a short time, often with ready-made materials and labs.
• Benchmark and objectivity: an external trainer can say “this is a bottleneck,” “this is an anti-pattern,” without internal politics.
• Easier to ensure a standard and teaching quality: external trainers earn their living by teaching, so they usually have a better-developed methodology.

Biggest cons

• Less fit to your context if you don’t do a good brief / needs analysis (which is why personalizing the program and examples is so important).
• Organizational and security risks: access to data, NDA, working on “production-like” examples or needing a sandbox environment.
• Risk of a “nice presentation without implementation”: if there are no exercises on your cases and no follow-up / change implementation process.
• Higher unit cost: usually higher per day/group, though it can save the team’s time.

Typical IT scenarios where external wins

• Getting started with Power BI (DAX, modeling, performance, governance), especially if you want to avoid bad practices early.
• Microsoft 365 / Entra / security / collaboration: when you’re implementing a new approach and need market “best practices.”
• AI tools in office and analytical work: prompting, automations, usage policies, real use cases, risk assessment.
• Internet marketing / analytics / advertising tools: the market and platform UIs change quickly - external specialists are often “fresh” on the newest versions.

The most practical approach: a hybrid model (and why companies naturally move toward it)

ATD data shows that organizations fund both internal L&D functions and external providers. In Poland, employers also often declare using both forms.

Hybrid works especially well in IT:

1. An external trainer does a kick-off + organizes standards + provides “best practices.”
2. Internal “champions” (change ambassadors) take over implementation in daily projects, consultations, and reinforcement.
3. After 4–8 weeks you return for an advanced workshop and consultations with real problems from day-to-day work and putting knowledge into practice.

This also aligns with our experience: without environmental support and practical tasks after training, the effect quickly drops.

A simple decision: 8 questions that choose the model for you

1. Is the topic new for the company (e.g., AI, Copilot Studio, new Power BI governance, new M365 setup)? → more often external / hybrid
2. Do you need company standards and work in your environment? → more often internal / hybrid
3. Is it recurring training (onboarding, cyclical)? → more often internal
4. Is the risk of mistakes high (security, compliance)? → often external + strong brief + sandbox / NDA
5. Do you have someone internally with time and teaching competence? → if not, external
6. Do internal politics make honest feedback difficult? → external
7. Do you need a benchmark and market comparison? → external
8. Does success depend on implementation in a project after training? → hybrid + follow-up

Summary

• An internal trainer wins where context, repeatability, and implementation in daily work matter most.
• An external trainer wins when you enter a new area and need up-to-date knowledge, objective perspective and market standards.
• A hybrid approach in IT very often delivers the best effect: an external trainer “sets the direction,” and an internal trainer “delivers the change.”