Choose the work before choosing the theme
Tommy Kroll's Great Query Clash gave a familiar training task the shape of a baking competition. Teams received exports as ingredients and worked through recipe-style clues to build a query. The challenge stayed close to something staff would actually need, rather than demonstrating everything a systems expert could build.
Start your own session by naming one useful outcome: find the right records, show the right information or adjust an existing query for a new question. Write the expected result before adding a theme. Then use the theme to make the instructions memorable. A good clue still gives people enough information to decide what to do, and the finished query gives you something concrete to check.
In the recording · 09:39 ↗Check the starting point, especially after roles change
A filter can seem basic to the person who uses it every day and be completely unfamiliar to someone whose responsibilities just changed. Tommy discovered that recruiters taking on transfer work had not encountered a transfer-specific filter included in his training. The exercise exposed a gap in prior experience.
Before the session, ask participants to explain a small example in their own words. Can they distinguish a filter from an export? Do they recognize the terms used in the task? Build a brief explanation or practice round around anything missing. Keep deeper system mechanics available for questions, while giving the whole group a clear path through the work they need to perform.
In the recording · 14:15 ↗Make room for people to help one another
Tommy describes teams gathering around one laptop, combining their knowledge of filters, exports and formatting. He also notices colleagues continuing to ask one another for help after the training. That is a useful outcome to look for beyond who wins the game.
Give each person a way to contribute. One can operate the query while another explains the choice and a third checks the result, then rotate. Watch for an expert doing all the work. Tommy moved systems and operations colleagues into judging roles after seeing the advantage they gave competing teams. Let facilitators check reasoning and offer a small hint without taking over the task.
In the recording · 19:54 ↗Practice editing without losing the query's purpose
With a substantial query library already available, Tommy shifted his training toward editing existing work. His relay used a prebuilt query and clues that asked pairs to diagnose an empty result, add information, format dates and sort the output. Colleagues checked each step before the next pair continued.
Use a training copy with fictional records and an expected result. Ask learners to explain whether a clue changes which records belong or which details appear. Before removing a filter, establish its purpose. A filter that is deliberately wrong in a puzzle is different from one that protects an authorized audience or preserves a business rule. Check the resulting records as well as the visible columns.
In the recording · 36:12 ↗Use the game to improve the next lesson
The relay revealed unfamiliar filters and exports, leaving Tommy with a practical question: should those items have been taught first, or did the exercise need a different level? A lively session can still uncover a mismatch between the task and the audience.
Keep a short note of where teams pause, what they misunderstand and which hints help. Debrief one confusing step together, then ask people to try a similar task without the game framing. That follow-up tells you more about usable knowledge than enthusiasm alone. Keep any competition accessible, with a seated option and no need to run through an office. Reward a correct, explainable result before speed.
In the recording · 42:00 ↗Leave something the next session can reuse
Tommy's advice to his younger self is to document recurring work and keep that documentation in a consistent place. Apply that to the training itself: save the starting query, fictional data, clues, expected answers and facilitator notes together. Record what changed after the session so another colleague can run it.
His proposed escape room makes repeatability especially clear. He wants stable dummy data so the same puzzle reliably produces the same outcome, with modules suited to different audiences. It remains an idea awaiting build time. Start smaller with one dependable challenge, and reuse the materials before adding more production. Tommy also describes exhausting preparation alongside supportive colleagues. Budget time and simplify the theme when needed so the training remains sustainable for its maker.
In the recording · 47:57 ↗