Seamless Onboarding

Your new hire shows up on Monday morning, eager to make a difference 😊. Then they spend their first week hunting down passwords, scrolling through outdated wikis, and watching months-old recorded onboarding and process presentations that leave them 😵‍💫.

A key team member drops out mid-project 😬, and their replacement has to piece together what's happening from skimming scattered documentation, reading through old Slack conversations, and hurried handoff meetings 😵‍💫.

Onboarding, whether for new hires to your company or for new team members on a project, is frustratingly inefficient. For the new hire or team member, it's a overwhelming experience that delays their ability to contribute. For the team they're joining, it's a distraction from their core work. For the business, it's a costly period of low productivity that lasts weeks for new team members and months for new hires.

Generative AI can transform onboarding both for the people creating onboarding materials and for the employees consuming them, giving your team the ability to contribute to the business, faster. 

Here's how.

Step 1:
Document Your Current Processes & Projects

Gather these materials; basically, everything you currently use to onboard a new employee:

• Existing onboarding documents and checklists
• Team/department wikis or knowledge bases
• Process documentation
• Training materials (including recorded or transcribed sessions)
• Onboarding feedback from recent new hires (this one is key!)

If you don't have much formal documentation, don't stress. Turn on a transcription service and start by asking yourself and current employees (both new and tenured) these questions:

• What are the critical systems new employees need access to?
• What key information do they need to understand about our business?
• What specialized knowledge is required for their specific role?
• Who are the key people they need to connect with?
• What common questions do new hires typically ask in their first weeks?
• What feedback have recent hires given about their onboarding experience?
  
  

For project-level onboarding, take stock of how you're organzing project files today. Ideally, you're storing everything in a unified place with clearly labeled folders. Given current AI model integrations (rows 17-21 in the AI Models for SMBs Comparison Chart), this is going to be quite easy with Google Workspace and a bit frustrating with anything else. 

Step 2:
Design Better Onboarding

Let's build a jig that helps you create better onboarding materials for both new employees and specific projects. Let's call it an Onboarding Architect. 

Start by creating a new GPT (ChatGPT), Project (Claude), or Gem (Gemini) with custom instructions like:

You are my Onboarding Architect. Your purpose is to help me create engaging onboarding materials for both new employees joining our company and existing team members transitioning to new projects.

When helping with onboarding content:
1. Organize information into logical learning sequences
2. Highlight critical information vs. "nice to know" details
3. Convert complex processes into clear, step-by-step instructions
4. Suggest knowledge check questions to confirm understanding
5. Identify content gaps that might confuse the person being onboarded

Approach different content types with these specific strategies:
- Company/Project Overview: Focus on mission, values, goals and how the person's role contributes
- Technical Documentation: Simplify without losing accuracy; create glossaries for specialized terms
- Access Instructions: Create clear steps for all necessary systems and resources
- Role Expectations: Balance responsibilities with context on how success is measured
- Team Dynamics: Highlight communication norms, meeting cadence, and key relationships
- Project History: Summarize key decisions, milestones, and learnings to date (for project onboarding)
- Stakeholder Map: Identify key relationships inside and outside the team

Format your content as:
- Overview: Brief summary of what the person will learn
- Critical Information: Must-know facts and processes
- Detailed Guidance: Step-by-step instructions where needed
- Knowledge Checks: Simple questions to confirm understanding
- Next Steps: Clear direction on what to do or learn next

Use a conversational, welcoming tone that builds confidence while maintaining clarity. Avoid jargon unless explaining essential terminology.

Always ask me:
1. The person's role, experience level, and whether they're new to the company or just the project
2. Company/project-specific terminology they should learn
3. Any known pain points in the current onboarding process

Upload your existing materials from Step 1 to provide context. Even if they're incomplete or outdated, they'll give your jig valuable information about your company's structure, roles, processes, or projects - just clarify in your instructions which are accurate and should be followed precisely and which are simply examples of what good looks like but are not current / active. 

Now you can use this jig to enhance any aspect of your onboarding process. Try prompts like:

Help me create a "First Day Experience" guide that could work for both:
1. A new software developer joining our company
2. An existing team member transitioning to our e-commerce redesign project

Our current process consists only of basic introductions and assigning an initial task. How could we create a structured, supportive experience that works in both scenarios?

The jig will help you develop an adaptable plan that might include:

• Pre-start communication with tailored information based on whether they're new to the company or just the project
• A structured first-day schedule with clear objectives for both scenarios
• A terminology guide that explains company or project-specific language
• Setup checklists for all necessary systems and tools
• Suggested small wins they can achieve in week one
• For project transitions: key documentation and decision history

The goal of this step is to take your current onboarding and level it up before you create tools to empower your team to self-serve as they onboard. You'll likely observe a plethora of opportunities to improve onboarding; take them in this set-up stage. For the sake of simplicity and clarity, try to align project-level and new employee onboarding as much as possible. 

Step 3:
Unlock Self-Guided Learning

Now let's build AI tools that people can interact with directly during onboarding. These "front-of-house" jigs work offer several advantages:

• Team members can learn at their own pace
• They can ask questions without feeling like they're bothering busy colleagues
• The experience can be personalized to individual needs and learning styles
• Questions can be asked privately without fear of appearing uninformed

AI as a Unified Knowledge Assistant

Build a flexible AI assistant that serves as both a responsive company handbook or a project guide. 

Given that you want to share this across your team, your choices are a bit more limited. As of this writing, Gemini Gems are not shareable (womp womp), but custom GPTs in ChatGPT and Projects in Claude are for Team and Enterprise accounts. It's likely a "when" not "if" Gems are shareable within a workspace given other Gemini product announcements. 

Try instructions like:

You are [COMPANY/PROJECT NAME]'s Onboarding Guide. Your purpose is to help people quickly become productive whether they're new to our company or joining an established project.

Your knowledge includes:
- Company/project background, mission, goals, and structure
- Relevant procedures and workflows
- Required systems and access points
- Team structures and key relationships
- Common terminology and concepts
- For projects: history, key decisions, and current status

When interacting with someone being onboarded:
1. Start by asking if they're new to the company or just joining a specific project
2. Tailor your guidance based on their specific situation and role
3. Provide concise, accurate information with links to resources
4. Encourage questions and clarifications
5. Acknowledge when you don't know something and suggest who to ask

You should NOT:
- Provide performance feedback or evaluations
- Share sensitive information outside their role's scope
- Make promises about policies or timelines
- Pretend to be a human colleague

Always begin by asking what specific aspect they'd like to learn about today, and adapt your responses based on whether they're new to the organization or just the project.

Upload the relevant set of materials to this assistant:

• For company-wide onboarding: handbook, policies, org charts, and general info
• For project-specific context: project plans, status reports, decision records, and team documentation
  
  

This creates a responsive guide that adapts to different onboarding needs.

A new employee might ask: "How do I set up my benefits?" or "What's our vacation policy?" A new project collaborator might ask: "Why did we choose this technology stack?" or "What were the key design decisions made in Phase 1?"

Using NotebookLM for Onboarding 

Google's NotebookLM is a compelling choice for employee and project onboarding, but comes with complications. Its strength as a research assistant makes it perfect for creating a unified knowledge base that can handle complex questions about your company or specific projects.

The tricky part is that NotebookLM is not - well - otherwise a typical chat-bot. It has amazing features like a built-in FAQ generator and timeline builder, but you can't interact with it as freely as you can other AI models. It also doesn't accept spreadsheets or images; just text and presentations. Before going down this path, I highly recommend you identify a limited expeirment as this has been a highly polarizing choice among my clients. 

If you're going to experiment with NotebookLM, I recommend you try it for a single project plus one fairly contained component of onboarding, such as benefits. For each, upload all of the documents associated with that project or process to the NotebookLM and then share it to your team. They'll be able to ask it anything they wish; heck, they can even turn it into a podcast that they can join. But they may find the format of the answers frustrating; the interface seems to get a love/hate reaction. Make sure to collect feedback on this scrappy experiment.

Step 4:
Build Interactive Learning Experiences

Static documentation, no matter how well-organized, can't match the engagement of interactive learning. Use generative AI to create experiences that actively involve people in the learning process, whether they're new to your company or joining a project mid-stream.

Practice Makes Perfect 

Claude's artifact capabilities as well as "vibe-coding" tools like Lovable or Bolt (watch for an upcoming edition of AI for SMBs Weekly on vibe-coding!) allow you to create and share interactive scenarios that simulate workplace situations for both new hires and project transitions.

Try something like:

Create an interactive customer service simulation for new team members. The simulation should:
1. Present realistic customer scenarios
2. Allow the employee to type their response
3. Provide feedback on their approach
4. Offer guidance for improvement
5. Suggest alternative responses that follow our service philosophy

Claude will generate HTML artifacts that function as simple but effective role-playing tools. These scenarios let people practice handling situations they'll encounter in their specific context - whether it's a new hire dealing with customers or an experienced employee navigating the politics of an established project.

Pop Quiz 

Create a flexible system of knowledge checks for new hires. Build a jig with a name like Onboarding Quiz, upload all of the documents from your self-serve jigs above, and try instructions like this:

Create an adaptive knowledge check system that can test understanding for new company employees. Each check should:

1. Include 5-7 scenario-based questions that test genuine understanding
2. Provide helpful explanations for both correct and incorrect answers
3. Link to relevant resources for further learning

For company-wide onboarding, include modules on:
- Company history, mission and values
- Organizational structure and key relationships
- Essential systems and processes
- General policies and procedures

Make the interaction conversational and encouraging rather than test-like.

This unified approach creates a flexible assessment system that adapts to different onboarding needs while maintaining a consistent experience. The questions automatically adjust based on context - new hires might get questions about company-wide systems, while project transfers get questions about project history and current status.

Both the simulations and knowledge checks serve dual purposes: they reinforce learning through active practice, and they identify areas where additional guidance is needed.

Step 5:
Feedback + Iteration

The final step is establishing a process that continuously improves your AI-enhanced onboarding system based on feedback from both new hires and project transfers.

Set up a consistent feedback mechanism with adaptable questions:

For everyone:

• Which aspects of the AI tools were most helpful?
• Were there questions the AI couldn't answer adequately?
• What additional information would have been helpful?

For new hires:

• Did the onboarding properly prepare you for your role in the company?
• Which company processes were still unclear after onboarding?

For new project members:

• Did you get sufficient context about project history and decisions?
• Were you able to contribute effectively within the first week?

Use this segmented feedback to refine your AI tools. For example, if multiple project transfers mention confusion about stakeholder relationships, that's a knowledge gap you should address in your project onboarding materials.

Here's a quarterly checklist to ensure you keep these jigs in tip top shape:

1. Analyze all onboarding feedback, segmented by type
2. Update both company-wide and project-specific information
3. Add new common questions and answers to all relevant AI tools
4. Refine simulations based on real workplace situations
5. Measure improvements in time-to-productivity for both scenarios

And - of course - you should leverage generative AI to help in the synthesis of the feedback you collect over time :)

Human + AI Onboarding ‧₊˚ ☁️⋅♡𓂃 ࣪ ִֶָ☾.

Effective onboarding isn't about replacing human connection with AI, it's about using AI to handle the information-heavy aspects so that human interactions can focus on relationship-building, culture, and nuanced guidance.

Techniques that help new hires understand company culture can be adapted to help project transfers understand project culture. Documentation approaches that effectively communicate project history to new team members can be repurposed to explain company history to new hires.

By transforming both employee and project onboarding from information dumps into interactive, personalized learning experiences, you're creating a culture of continuous knowledge sharing that benefits your entire organization - from the newest hire to the most experienced team member transitioning to their next challenge.

✨ ✌🏻 ✨

If you have ideas for future newsletters, I'd love to hear them.