AI Tools: Track Cost Per Developer Per Month and Justify the Spend

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-tools-cost-per-developer-month-r8a1-creators

1. What is the main reason to set up per-seat usage telemetry for AI tools in a development team?

   1. To automatically disable tools that developers don't use frequently
   2. To track exactly how many hours each developer works each day
   3. To justify tool spending with data rather than relying on intuition
   4. To predict which developers will leave the company soon

2. Which task can AI tools currently perform when analyzing AI tool spend?

   1. Directly measure whether code quality has improved because of an AI tool
   2. Aggregate per-seat usage data from multiple tools and convert it to monthly costs
   3. Decide which developer seats should be cut based on usage data
   4. Replace the need for a formal productivity study entirely

3. A developer shows up in the bottom 10% of AI tool usage for the month. What should happen next?

   1. Immediately reclaim their seat to save money
   2. Assume the tool isn't valuable and stop paying for it
   3. Report them to their manager for low productivity
   4. Investigate whether they're on PTO, on a project that doesn't fit the tool, or simply efficient

4. What does 'per-seat cost' refer to in this context?

   1. The total monthly cost of an AI tool divided by the number of developers using it
   2. The cost of hosting AI models on your own servers
   3. The price a company pays for each license or subscription to an AI service
   4. The budget for training developers on AI tools

5. When correlating AI tool usage with output, what metric might you use as a proxy for productivity?

   1. Time spent in meetings
   2. Length of commit messages
   3. Number of coffee breaks taken
   4. Number of pull requests shipped

6. What is 'seat reclaim' in the context of AI tool management?

   1. Reinstalling software on a new computer
   2. Filing a refund for an AI tool that doesn't work
   3. Taking back unused or underused tool licenses from developers who don't need them
   4. Recovering a developer's workstation after they leave the company

7. Why is it important to check multiple data sources when building a cost-per-developer dashboard?

   1. The finance team only accepts data from one specific platform
   2. Each tool has different usage APIs and metrics, requiring aggregation from all sources
   3. Data sources don't matter as long as you have one good source
   4. Most tools provide identical data, so you need to verify consistency

8. What is a key limitation of using AI tool usage data to prove ROI?

   1. AI tools don't track usage accurately enough to be trustworthy
   2. The cost data from AI vendors is always wrong
   3. Developers will lie about how much they use the tools
   4. Usage data shows how much a tool is used but not necessarily the quality of output it produces

9. You notice a developer has zero AI tool usage for three months. Which explanation should you consider BEFORE reclaiming their seat?

   1. They have been fired but their account wasn't deactivated
   2. They may be working on a project where the tool doesn't apply or may be on extended leave
   3. They must have found a better alternative tool on their own
   4. They are likely trying to avoid using company resources

10. What should a cost-per-developer dashboard be able to answer for finance?

    1. Which developer writes the best code
    2. Whether the money spent on AI tools per developer is generating sufficient value
    3. Which programming language is most popular
    4. How many bugs were introduced by AI-assisted code

11. What does the term 'usage telemetry' refer to?

    1. Manual surveys asking developers how often they use tools
    2. Physical tracking of developer computers using GPS
    3. Automated collection of data about how much and when a tool is used by each user
    4. The process of sending usage data to AI models for training

12. If an AI tool costs $20 per developer per month and you have 50 developers, but only 30 actively use it, what is the effective monthly waste?

    1. $50—the cost of the unused portion
    2. $1,000—total cost for all developers
    3. $0—only pay for active users
    4. $400—paying for 20 unused seats

13. Why might a developer show low AI tool usage even if the tool is valuable?

    1. They likely don't understand how to use the tool properly
    2. They may be highly efficient and need fewer prompts to get results
    3. The tool must be malfunctioning
    4. They are deliberately avoiding using it

14. What happens if you skip the investigation step and immediately cut tools showing low usage?

    1. The AI will automatically add better tools
    2. You will definitely save money on every cut
    3. You might remove tools that developers actually need but aren't using due to project timing
    4. Developers will be happy to lose tools they don't use

15. When the lesson says AI 'cannot replace a real productivity study,' what does it mean?

    1. AI is not smart enough to run studies
    2. Productivity studies are too expensive to conduct
    3. AI tools will only be useful for small teams
    4. Formal studies with controlled conditions are needed to truly measure tool impact on productivity