It’s not magical to think this way; instead of worrying about how AI might replace you, how about if AI took over all those tasks you don’t like to do?
We dive into how to identify and automate those time suckers with two use cases!
Listen on Apple || Spotify || YouTube
0:00 - Introduction
2:00 - The 4 Keys to Intelligent Workflows
5:00 - Setting Clear Goals
7:00 - Identifying Tasks to Automate
9:00 - Effective Communication
11:00 - Continual Review and Refinement
13:00 - Example 1: Using GPTs to Analyze Business Data
15:00 - Example 2: Automating Grant Proposal Writing
17:00 - Example 3: AI-Enhanced B2B Targeting
19:00 - Starting Small When Implementing AI Workflows
21:00 - Providing Ongoing Guidance to AI Like an Employee
23:00 - Conclusion - AI Can Free Up Time If Thoughtfully Implemented
Here is a summary of the key points about using AI to free up time and increase productivity from the podcast:
1. AI tools like GPTs, assistants, and agents can save 20-40% of the time wasted on administrative tasks and things you don't want to do. They act as a "co-pilot" to help streamline workflows.
2. The 4 keys to creating an "intelligent workflow" with AI are:
- Set clear, focused, and achievable goals
- Identify and automate the most tedious tasks
- Ensure good communication between people and AI
- Continuously review and refine the AI system
3. Examples of intelligent workflows enabled by AI include:
- Using GPTs to analyze business data and suggest actions
- Automating the writing and sending of grant proposals
- Helping target ideal B2B customers based on their needs
4. When implementing an AI workflow, start small by picking one frustrating task and figuring out how AI could simplify or eliminate it. Focus on augmenting human capabilities, not replacing jobs.
5. Approach AI workflows the same way you would an employee - provide ongoing training, give periodic feedback, and refine goals over time. AI takes continual guidance.
In summary, AI tools can free up significant time if thoughtfully implemented to enhance workflows. The key is setting clear goals, communicating effectively, and continuously guiding the AI.
How You Already Use AI – Maps and Netflix
The funny thing about AI is how it penetrated our lives in the background way before the ChatGPT craze made the overused term AI a part of our vocabulary.
When you drive, how often do you use Google Maps, Waze, or whatever map program to find your way around?
It’s so typical these days that people forget when we had to use maps, memory, and a sense of direction to find our way home.
Now, you type in where you’re going, and the directions appear, guiding you.
That’s what AI agents and assistants will do for us. Sure, using a map and finding your way seems more nostalgic, much like adding numbers in your head was lost to many with the invention of the calculator.
The way maps are a co-pilot, showing us the way, is how AI is taking over.
Netflix suggests movies and shows you might like; Spotify prompts musical alternatives and playlists. All this seems normal compared to 20 years ago when you had to do this yourself.
In this pod, we’re exploring how to pick out those time wasters in your life and business hours. Seeing where AI can not only fill in those gaps but also give you time to do what you like or do on your job.
I’ll show how to plan and roll out your GPT for those who won’t consider this pod a 4-Step AI Roadmap to streamline life and work’s boring, mundane, and exhausting aspects.
1. Goals: Setting a clear purpose, focused and doable, is vital. Who else will be involved in achieving this goal if this is your job?
Bonus points for considering if the goal is even the right one and adapting it.
After all, goals are a dream with a deadline. Sometimes, the original goal isn’t the right one to help you!
2. Efforts: What tasks do you hate, and which can you automate with AI?
Many waste time and energy doing things AI could do better, like emailing the Team with an update.
Identify what you can streamline and free time to do better work and engage more.
3. Communicate: Remember that communication is the art of sharing ideas and emotions.
Yet, instead of connecting us, we often detour through the land of overcomplication and missed cues.
Learn how to navigate this quirky communication landscape with AI, clarity, and a bit of humor.
4. Review and Revisit: AI needs reviews like anyone. It’s not automatic and always correct.
For example, employees get quarterly and yearly reviews, so look at AI.
Is it helping people, or do they need training, or does the AI?
Let’s walk through these steps to create a GPT. I used my example from last week, the DataWise Director I made to help small businesses get the most out of their Google Analytics data.
Even writing that is a mouthful and keeping it simple is critical.
DataWise Director Case Study:
Freeing Time with Google Analytics for Those Lost in Spreadsheets
Creating a new GPT for the first time, I see many doing funny tricks like Simpsonize Me GPT, which takes someone’s photo and makes a Simpsons-like cartoon image.
Cool and fascinating, though more playing with GPT Builder than solving a problem or achieving a goal.
[Great article from VentureBeat outlining these examples and more]
Rowan Cheung creates a Twitter social engagement GPT,
“AI influencer Rowan Cheung, creator of The Rundown AI newsletter, created X Optimizer GPT, which automatically analyzes his proposed text for posts from his account on the social network X (formerly Twitter) and suggests improvements and optimal times to post for maximum engagement from the network.”
Now, that tool takes your Twitter/X data and guides you. You can look at the data and do this; many social media pros do.
Yet most people are not pros, don’t have time, and are overwhelmed -PERFECT for GPTs.
These inspirations led me to my own DataWise Director because:
The goal is to get valuable data from your Google Analytics instead of the tsunami of spreadsheets shared. Many digital marketing pros love to read these, and many I know want to find the essential data for them.
To meet a specific goal, I focused on organic content marketing targeted toward creators, bloggers, podcasters, and others who seek to share content to spread awareness and get traffic.
I left out the high conversion elements for this run – like e-commerce and ad-buying businesses – because it likely would complicate matters. I could have added some conversion elements and may do so in the future, but for this run, I kept it simple.
Efforts were the next piece to solve; is this truly a problem?
So many small businesses I’ve interviewed around The AI Optimist podcast point out that these numbers aren’t helpful, primarily how they are delivered.
They need to know what articles are working and where traffic is coming from, including digital sources and geographic locations.
Communicate: how can I share this information with people using simple prompts, uploading a PDF or CSV (Excel) file of their data, and getting specific answers for content marketers?
With this goal, I stepped into creating the GPT quickly because I had a set of instructions to share with the GPT builder at ChatGPT.
I could have done this without the chat, adding instructions and tabs, but I let the GPT builder guide me, which helped finish the initial work.
I did upload Google Analytics training PDFs I’d shared with a company. However, I removed the names as somebody may share this data you upload to ChatGPT.
Be careful to understand this with any AI automation; if you upload private data, be sure it is anonymized and protected from public view.
Here’s how it developed; follow along and consider this only as an example.
What time wasters do you have, and how can you outline a solution following this process?
1. GOALS
My goal is to get people usable information from their Google Analytics. Still, I also want them to get a 30-day Action Plan and suggestions for improvement. Just sharing the data doesn’t solve the problem.
The goal must be a specific and unified approach to getting actionable data designed for a single individual to get some good ideas for now and into the next 30 days.
I used my AI Optimist traffic, which Google is only partially reporting on. But it’s an excellent basis to see what a new site, only 60 days old, is doing in terms of traffic and engagement.
By doing this, I aim to increase repeat usage and give people something to do instead of just sharing the data and letting them decide what to do next.
Here’s the GPT creation process I used.
Remember, if you are creating a GPT or just a solution to your problem, what elements can you include upfront, and what will people do next?
Summary of DataWise Director Development:
Initial Configuration: The DataWise Director analyzes organic traffic, page quality, and search traffic from social channels. It focuses on natural engagement and content strategies, excluding ad spend and data.
Enhanced Reports: We added more detailed information in reports, such as page titles, URLs, and meta descriptions, that help figure out which pages are working best.
Action Plan Development: Provide a 5-step weekly action plan with each report.
CSV Report Generation: The capability to generate and download CSV files summarizing findings is an option.
Automatic Email Functionality: Automatically email summarized findings and CSV reports to specified recipients. This function didn’t work, though GPT Builder indicated it would.
4 Custom Choices: Finally, the interface offers four specific tabs:
‘Analyze my Traffic,’
‘What Improvements Should We Focus On,’
‘Send CSV Analysis to My Team,’ and
’30 Day Action Plan.’ The ’30 Day Action Plan’ tab includes a summary of key data points, recommendations, and a structured 5-step action plan.
The goal is to give people valuable directions and insights from the DataWise Director; in my tests, it achieved these goals, though it could not email.
The CSV files created are basic and not helpful.
2. Efforts
The next step when creating a GPT or any automated AI solution is to define what tasks – in this case, what Google Analytics data – should be delivered. Usually, a person would sift through Google Analytics or create Custom Reports to view this.
My solution requires them only to download a PDF or CSV and upload it to the DataWise Director to get the following guidance.
Visitor Segmentation: Determine who visits the website and where they are coming from, analyzing segments to understand different user groups and their behaviors.
Channel Analysis: Examine which channels (like search, direct, and referral traffic) drive visitors to the site, understanding the effectiveness of various marketing efforts and where to focus resources.
User Engagement: Investigate detailed searches, time spent on the site, and repeat visits. Events tracking in Google Analytics is crucial here, especially considering extensive search events.
Conversion Tracking: Assess how well the site converts visitors into leads or sales. Currently, The AI Optimist uses goals in Google Analytics primarily to track leads via email.
However, my tracking still needs to be improved since the website is in Substack. Expand this to include events as goals, which can help measure and value activities like searches.
Bounce Rate and Exit Rate Analysis: Understand the behavior of users in terms of how they interact with different pages. Analyze which pages users are leaving from and whether they leave the site after viewing only one page (bounce rate) or after multiple pages (exit rate).
Goal Improvement: The current goals in Google Analytics are primarily for measuring leads as emails for AI Optimist signups. Use events as goals to connect visitor activities with revenue-generating actions better.
Event Value Assignment: Assign monetary values to events, especially searches, to define their weight and page value to understand which products are popular and potentially more profitable.
Content and Page Value Analysis: Analyze page values and understand which pages contribute more to the site’s revenue.
Survey Funnels and Lead Qualification: Use survey funnels instead of email forms. Allow users to self-identify, providing valuable data for segmentation and targeted marketing.
Even this is more detailed than most people will use. As a long-time CEO and entrepreneur in the digital market, my choice to add more layers could confuse some.
Yet, learning what to do immediately and how to grow is also essential for efforts to succeed in building digital businesses.
When designing any AI automation, consider TMI – too much information – is involved. Keep things lean and simple, yet being too simple can undermine effectiveness.
Also, ask AI – in this case, ChatGPT’s GPT builder – for advice and insights. Given the topic, it had lots of information and helpful guidance for my design.
3. Communicate
I’d love to have the DataWise Director automatically email the Analytics summary and 30-Day Action plan to all involved, but it wasn’t possible.
It’s important to use AI to summarize, focus people on the most critical information, and send this out to the right people.
We often rely on people forwarding emails, which has caused sloppy communication. But I couldn’t solve this problem with ChatGPT, and while it’s possible, automation is tricky.
Asking ChatGPT’s GPT Builder to improve my DataWise Director is possible. Communication is critical for all involved, yet it’s excellent to have AI guide you when you lack people to give you feedback initially.
Plus, it’s always available and never considers it a burden to give you responses.
Here are a few questions and feedback from ChatGPT that updates the DataWise Director:
THE AI OPTIMIST
When assessing most viewed pages and screens, the only thing that shows is the title.
I want more than a title, including the page URL and meta description. With each report, I’d like the recommendations followed by a 5-step weekly action plan to implement improvements.
I want the ability to download a CSV with the summary of findings put into a spreadsheet with each report.
GPT BUILDER
DataWise Director has been upgraded! It can now generate and provide a downloadable CSV file summarizing its findings with each report.
This feature is in addition to its detailed analysis of the most viewed pages, including titles, URLs, and meta descriptions, as well as its tailored 5-step weekly action plans for improvement.
Would you like to make any other changes or additions?
THE AI OPTIMIST
Can I get these automatically emailed to people within the DataWise Director?
GPT BUILDER
DataWise Director has been further enhanced! It can now automatically email summarized findings and CSV reports to specified recipients.
This feature complements its existing abilities to analyze digital marketing data, provide detailed insights on the most viewed pages, and offer tailored 5-step weekly action plans.
Is there anything else you’d like to adjust or add to DataWise Director?
THE AI OPTIMIST
For the tabs when you open DataWise Director, please include Analyze my Traffic, What Improvements should we focus on, Send CSV analysis to my Team, and 30-Day Action Plan under the 30-Day Action Plan, including a summary of the critical data points, recommendations, and a 5-step Action Plan to implement.
GPT BUILDER
DataWise Director now features a more user-friendly interface with four primary tabs: ‘Analyze my Traffic,’ ‘What Improvements Should We Focus On,’ ‘Send CSV Analysis to My Team,’ and ’30 Day Action Plan.’ Each tab is specifically designed for different aspects of marketing analysis and action planning.
The ’30 Day Action Plan’ tab, in particular, includes a summary of key data points, recommendations, and a detailed 5-step action plan.
Would you like any further modifications or additions to the DataWise Director?
Several other questions and responses happened during the discussion. But when I asked to download the history of the conversation, it was yet to be available for GPT Builder.
I assumed it would save all my chats like ChatGPT does with my paid subscription.
4. Review and Revisit
Any AI automation must be regularly reviewed and evaluated. Revisit whether the answers are accurate, if problems arise like hallucinations, and where improvements are needed.
In my initial evaluation, users had to copy/paste their answers into a separate doc; they couldn’t save them. All GPTs share that weakness now.
A few users I talked with didn’t have their business model covered, as the DataWise Director only addresses organic content businesses.
If I choose to develop this as a business model, I will have to run this on my server and pay for the GPT API. Which isn’t expensive, but each answer costs money.
Expanding the DataWise Director to cater to a broader range of companies, including e-commerce sites and ad buyers, is a great idea. Here are some considerations for enhancing its capabilities:
1. Customization for Different Business Models:
E-commerce Sites:
Integrate metrics crucial for e-commerce, like conversion rate, average order value, cart abandonment rate, and customer lifetime value.
Implement tracking for funnel analysis to understand where customers drop off in the purchasing process.
Add features to analyze the effectiveness of product pages, including heatmaps and click tracking.
Ad Buyers:
Focus on metrics like click-through rate (CTR), cost per click (CPC), cost per acquisition (CPA), and return on ad spend (ROAS).
To compare performance across channels, provide analytics for different ad platforms (Google Ads, Facebook Ads, etc.).
Include tools for A/B testing ad copy and creatives.
2. Data Integration and Management:
Ensure the DataWise Director can integrate with various data sources and platforms (e.g., Google Analytics, Shopify, ad platforms).
Implement a robust data management system to efficiently handle large volumes of data from diverse sources.
3. User-Friendly Dashboards and Reporting:
Create customizable dashboards that allow users to view the most relevant metrics for their business type.
Offer automated report generation tailored to business needs, including detailed insights and actionable recommendations.
4. Predictive Analytics and Machine Learning:
Employ machine learning algorithms to predict trends, customer behavior, and potential outcomes based on historical data.
Use predictive analytics to suggest optimization strategies for ad campaigns and e-commerce sales.
5. User Segmentation and Personalization:
Implement features for advanced user segmentation to understand different customer groups’ behavior.
Offer personalization strategies for e-commerce sites to enhance user experience and increase conversions.
6. Training and Support:
Provide comprehensive training materials and support for users to maximize the potential of the DataWise Director.
Offer consultation or advisory services for complex implementations and strategies.
7. Feedback and Iterative Improvement:
Implement a feedback system to gather user input on the tool’s usefulness and areas for improvement.
Regularly update the DataWise Director based on user feedback and emerging market trends.
2 More Examples for Creating AI Automation for Other Business Models
For someone inspired to create a tool like DataWise Director for a different industry, such as healthcare or education, the first step is to identify the specific problem or need the tool will address. Here’s a breakdown of this crucial initial step:
1. Define the Purpose and Scope:
Healthcare Example: Analyze patient data to improve treatment plans or predict health trends.
Education Example: Analyze student performance data to personalize learning experiences or identify areas needing additional support.
2. Understand the Data:
What Data Will You Use? The data could be patient records, treatment outcomes, or demographic information in healthcare. It might be test scores, attendance records, or student feedback in education.
Data Accessibility and Privacy: Ensure compliance with privacy laws like HIPAA in healthcare or FERPA in education. Data security and ethical considerations are critical.
3. Identify Key Features:
Healthcare: Features might include disease prediction, treatment optimization, or patient risk assessment.
Education: Look at features like personalized learning plans, curriculum effectiveness analysis, or early detection of learning difficulties.
4. Consult with Industry Experts:
Gain Insight: Talk with healthcare professionals or educators to understand the nuances of the field and validate the relevance and usefulness of your proposed tool.
Collaboration: Collaboration provides practical insights and helps fine-tune the tool’s objectives.
5. Set Realistic Goals:
Start Small: Begin with a manageable aspect of your larger vision. For instance, focus on a specific type of disease in healthcare or a particular grade level in education.
Plan for Growth: Design your tool to be scalable and adaptable as you expand its capabilities.
By starting with a clear purpose and understanding the industry’s specific needs, you lay a solid foundation for developing a practically useful and ethically sound tool.
This focused approach ensures that your tool will effectively address the unique challenges in your chosen field.
Free Your Time with AI - GPTs, Assistants, and Agents
Begin with the end in mind; what time is wasted in your life or work?
Are AI automations possible for you to execute?
For example, I wanted to create a podcasting GPT that would transcribe podcasts, give timestamps for posting, analyze the text, and share a summary, plus include a takeaway action plan for each episode.
The task is possible, but doing this within ChatGPT would require everyone to integrate their podcast platform, which is likely impossible.
It would also require people to upload their audio, which ChatGPT cannot yet do.
Still, there are tools out there doing this; I use Substack for The AI Optimist. It’s a free platform, allows me to gather emails for subscribers, and does many things.
Among them is transcribing my podcast automatically and allowing me to email my list through the tool.
So, I use LLMs like ChatGPT and Claude to help with my podcast summaries and time stamps, but other than that, everything is done through Substack.
While not automated, how powerful is that?
Take that approach when creating AI automation to help you as a personal assistant, AI Agent, or GPT.
Nothing is perfect, but it’s accessible and available. Plus, it will only improve.
AI Black Friday – Freeing 20-40% of Time Wasters with GPTs, AI Assistants, and AI Agents