How to Start a Hockey Analytics Function in Your Club

# How to Start a Hockey Analytics Function in Your Club Many clubs feel they "should" be doing more with data but are unsure where to start. Do you need a full-time hire, an external partner, a new tech stack—or all of the above? This guide walks through the practical steps to build a hockey analytics function that actually helps your coaches, scouts and front office make better decisions. The goal is not to create a flashy department. The goal is to create a **repeatable way** to turn data into decisions for your organization. --- ## 1. Start with hockey questions, not with tools Before thinking about software or hiring, clarify **why** you want analytics: - **Roster decisions** – Which players help you win more than their box score suggests? Who do you overpay? - **Scouting & recruitment** – How can you find undervalued players before other clubs do? - **Tactics & game prep** – How can you prepare for specific opponents and special teams matchups? - **Player development** – How do you track progression over months and seasons in a structured way? Sit down with: - GM / Sports director - Head coach + assistants - Head of scouting - Player development staff And ask: **"If analytics worked perfectly for us, what decisions would feel easier or more confident?"** Those answers become the starting backlog for your analytics function. --- ## 2. Decide what "analytics function" means for your club Not every club needs the same structure. In practice, most organizations end up in one of three models: ### A. Internal analyst (or small team) - **Pros**: Full-time focus, always available, embedded in the building. - **Cons**: Expensive, hard to hire the right hybrid of technical + hockey experience, risk of turnover. This model suits large pro clubs with stable budgets and long-term commitment to analytics. ### B. External partner (consultant model) - **Pros**: Senior expertise from day one, flexible scope, no full-time payroll risk. - **Cons**: Requires good communication routines and clear expectations. For many clubs, this is the fastest way to get real analytics impact without multi-year commitments. At The Hockey Brain Consulting, we operate in this model—based in Stockholm but working with clubs worldwide. Because we combine **technical data skills** and **real hockey backgrounds**, you don't have to translate between "data people" and "hockey people" all the time. ### C. Hybrid model - Internal liaison (coach, video coach, operations person) owns questions and context. - External consultant handles data engineering, modeling and tooling. This is often the most realistic starting point for junior clubs, national teams and growing organizations. If you want to see how our workflow fits into this, read [Our End-to-End Analytics Workflow](/insights/end-to-end-analytics-workflow). --- ## 3. The technical foundation you actually need You don't need an NHL-level budget to start. But you do need a **minimum technical backbone** so your function can grow. ### 3.1 Data sources List the data you already have today: - League or federation data feeds (events, rosters, schedules, standings). - Video tagging systems and tracking providers. - Internal scouting reports and ratings. - GPS/wearable data (if applicable). Key questions: - How do we **access** this data (API, CSV exports, manual files)? - How often do we **update** it (daily, weekly, per game)? - Who currently "owns" each data source? Even a simple spreadsheet or CSV pipeline can work as a first step—as long as it's consistent. ### 3.2 Storage and pipelines You need a place where all relevant data can live together: - A cloud database or data warehouse (e.g. PostgreSQL, BigQuery). - Or a structured folder setup with CSV/Parquet files to start. The key is to: - Standardize **identifiers** (player IDs, team IDs, league/season codes). - Keep **raw data** and **processed tables** separate. - Automate updates as much as possible (scheduled scripts instead of manual copy-paste). Our article [Our Modern Analytics Toolkit](/insights/modern-analytics-toolkit) describes one cost-effective stack that works for most clubs. ### 3.3 Analysis & modeling tools For most hockey clubs, a Python-based stack is enough: - Python + pandas / NumPy for data processing. - scikit-learn or similar for basic modeling. - Jupyter or scripted pipelines for repeatable analysis. You don't need every machine learning buzzword. Start with: - Expected goals (xG) and shot quality models. - Player similarity and role profiles. - On-ice impact analyses adjusted for teammates/opponents. ### 3.4 Delivery: dashboards and reports Analytics only matters if people see and trust the outputs: - **Dashboards** tailored to: - Coaches: line combos, matchups, special teams. - Scouts: player cards, draft boards, comparables. - Management: roster value, contract decisions, long-term trends. - **Static reports** before key milestones: - Draft meetings. - Trade deadlines. - Playoff series. Focus on **clarity and repeatability**. A simple, reliable report that arrives every week is more valuable than a one-off "wow" analysis. --- ## 4. The operational side: roles and workflows inside the club The biggest difference between clubs where analytics works and where it fails is not tech—it is **ownership and workflows**. ### 4.1 Assign clear ownership Someone must own the analytics function internally, even if you work with an external partner: - They collect questions from coaches and scouts. - They help prioritize requests. - They ensure outputs are actually used in meetings and decisions. This could be: - An assistant coach with strong interest in data. - A video coach or hockey operations coordinator. - A dedicated analytics liaison in the front office. ### 4.2 Define when analytics is used Map out your **recurring decision moments**: - Pre-game meetings and series prep. - Weekly player development reviews. - Scouting meetings before drafts or transfer windows. - Contract and roster planning checkpoints. For each, answer: - What information do we currently bring into the room? - What analytics outputs could improve these conversations? - In which format should they be delivered (dashboard, PDF, clips, bench cards)? ### 4.3 Communication cadence Set a light but consistent rhythm: - Weekly or biweekly check-in between analytics and coaching/scouting. - Short async updates (email/Slack) when something important changes. - Clear deadlines for when data needs to be ready (e.g. draft board freeze date). Without this, analytics becomes random side projects instead of an embedded function. --- ## 5. Effects you can realistically expect When a hockey analytics function is set up correctly, clubs typically see impact in three main areas. ### 5.1 Better decisions at the margins Analytics won't magically turn a bottom team into a champion. But it can help you: - Avoid overpaying for "name" players with declining impact. - Find undervalued role players that fit your system. - Make sharper decisions on borderline roster spots. Over multiple seasons, these edge decisions compound. ### 5.2 More structured internal discussions Instead of opinions vs opinions, you get: - Shared definitions of roles and playing styles. - Common language around shot quality, transition play, special teams. - Fewer arguments about facts—more focus on strategy and execution. ### 5.3 Faster learning cycles With good data and workflows you can: - See earlier when a tactical adjustment actually works. - Track development plans and identify when to intervene. - Evaluate whether your scouting and recruitment assumptions hold up. Clubs that review and adapt faster usually outperform those who change slowly. --- ## 6. Build vs partner: what makes sense for your club? For some organizations, the right move is hiring full-time staff. For many others, the best path is to **partner first**, learn what works, and then decide if an internal hire makes sense later. Questions to ask: - Do we have a clear enough roadmap to keep a full-time analyst busy with high-value work? - Can we realistically hire someone with both technical and hockey expertise? - How comfortable are we managing an analytics employee internally? Working with a specialist consulting partner can: - Give you senior expertise immediately. - Reduce risk if priorities or budgets change. - Help you design the function so that a future internal hire can step into a well-structured environment. Because we are based in Sweden and understand both the **European** and **North American** hockey contexts, we are used to adapting to different leagues, data sources and staff structures. --- ## 7. A 90-day roadmap to get started You don't need a five-year plan. You need a **clear first 90 days**. ### Days 1–30: Discover & prioritize - Clarify goals with GM, coaches, scouts. - Inventory data sources and current tools. - Identify 2–3 decisions where analytics could help this season (not in theory, but in practice). ### Days 31–60: Build foundation & first deliverables - Set up basic data pipeline (even if it's just structured exports). - Build 1–2 core reports or dashboards tied to real decisions (e.g. lineup optimization, draft list, special teams). - Test the outputs with staff, refine format and timing. ### Days 61–90: Embed & expand - Bake analytics into recurring meetings. - Document processes: how questions come in, how outputs are delivered. - Plan the next 1–2 projects based on what worked best in the first cycle. After 90 days you should be able to answer: - Which decisions improved because of analytics? - Which formats and workflows do our staff actually like? - What should we invest in next (people, tools, or external help)? --- ## Internal links & next steps If you want help designing or running your hockey analytics function: - Read about [our services](/services), including team & player analytics, scouting models and organizational dashboards. - See our [End-to-End Analytics Workflow](/insights/end-to-end-analytics-workflow) for how projects typically run. - Learn [How Remote Consulting Works](/insights/how-remote-consulting-works) if you want to understand our remote model. - Or simply [get in touch](/contact) to book an introductory call and see what an analytics function could look like for your club.