How to set and track goals with AI in 2026
or your own safe AI Personal Assistant and Coach (the full downloadable setup inside)
It is almost the end of January. If you are like 90% of the population, the shiny “New Year, New Me” energy is already starting to fade. You’ve set the resolutions. You’ve bought the planner. And now, the reality of “maintenance mode” is setting in.
Hey! I’m Alena, former AI startup CEO ($2M raised), Yandex and Acronis Sr. PM with 10k LinkedIn followers, 57% of my followers are senior leaders from big tech, good company ;)
There are tasks that should be delegated to AI.
And then there are tasks where you need to build a custom project – one that works for you, and only you.
(I’ve included the full free template for AI personal assistant at the end of this post. I spent over 50 hours debugging the prompts so you can set this up in 5 minutes.)
There are about 100 popular AI-assisted to-do lists popping up right now.
Teresa Torres, author of Continuous Discovery Habits and an internationally acclaimed coach, explains this proliferation simply:
“How we manage our tasks is so idiosyncratic that this is exactly the type of thing that you should build for yourself because it can work exactly the way you want it to work.”
I couldn’t agree more.
I’ve been setting personal and professional goals for the last 10 years. I rarely achieve even half of them. Usually, things lose relevance, or I simply forget they exist. This year, I decided to outsource the executive function to AI.
Imagine having a personal assistant with the skills of an executive coach, ready to help you 24/7 for $0 a month. Top CEOs have Chiefs of Staff. Professional athletes have performance coaches. Why do we expect ourselves to manage our complex lives alone?
The assistant that doesn’t just manage your calendar. It manages you. It knows when you’re lying to yourself about “just checking email.” It knows when you’re catastrophizing. And it knows exactly how to get you back on track without making you feel guilty.
But first, I did some research to find the most scientifically validated techniques for goal setting and tracking.
The science
Part 1: How to set and track goals
It turns out, simply “set SMART goals” isn’t enough. Here are the four evidence-backed frameworks I’m using:
1. Implementation Intentions (Gollwitzer, 1999)
The structure is simple: “If [Situation], then I will [Action].”
Example: “If it is 9:00 AM on Monday and I open my laptop, then I will write for 25 minutes before checking email.”
2. WOOP (Oettingen, 2014)
Wish → Outcome → Obstacle → Plan
Wish: What do you want?
Outcome: What will it feel like?
Obstacle: What internal barrier stops you?
Plan: If [Obstacle], then [Action].
Fun Fact: Positive visualization without visualizing the obstacles actually reduces the likelihood of achieving the goal. You need to prepare for the resistance.
3. Learning vs. performance goals (Locke & Latham, 2002)
I personally love this concept.
Performance Goals: Good for familiar tasks (e.g., “Increase conversion rate by 2% in Q1”).
Learning Goals: Essential for new or complex territory.
Bad: “Hit 10k subscribers.” (You can’t fully control this).
Good: “Run 5 experiments to understand what drives subscription growth.” (I will learn/understand X).
Fun Fact: The SMART methodology lacks a strong evidence base for complex, ambiguous goals. It often forces false certainty.
4. Written Goals + Accountability (Matthews, 2015)
The formula: Write the goal → Share with 1 person → Schedule weekly check-ins.
Part 2: How to track goals & maintain motivation
The Weekly Check-in (Harkin et al., 2016)
Action: Every week, ask yourself: “Am I closer to the goal than I was a week ago? If not, why?”
If you lose motivation → Goal Gradient (Kivetz et al., 2006)
Action: Break the goal into 5-10 stages. Show progress as “X% done” (not “Y% remaining”). Make the first 10-20% easy wins to build momentum.
If you forget to act → Implementation Intentions
Action: Write “If [Situation], then [Action]” and place a reminder at the physical trigger point.
If you have too many goals → Sequential Focus
Action: Maximum 3 goals per quarter. 1 MIT (Most Important Task) per day. Finish one before adding another.
If you’re stuck and don’t know why → WOOP Diagnostic
Action: Ask yourself: “What INSIDE ME is stopping this? What feeling am I avoiding?”
Part 3: How I picked goals
To start, I wrote down 100 wishes. It was hard. My imagination ran dry after the first 20, but that’s when the interesting stuff started coming out (e.g., “start taking golf lessons” or “take a course at the Art Institute”).
For the final list, I divided all goals into 3 groups:
One-off Projects: Finite tasks with a clear end.
Pillars: Ongoing areas requiring constant engagement (Startup, Consultancy, Personal Brand).
Side Quests: Everything else – from learning to make pizza to visiting my 41st country.
Important: I intentionally did not separate “Work” and “Life.” Work-life balance is one of my goals this year, and I plan to achieve it by freeing up more time for rest using AI.
The build
Here is how I actually built it:
Created a new project. I use Cursor, but you can use Obsidian, Antigraviti, Claude Projects, ChatGPT Projects or just a folder on your desktop. The magic is that it’s just text files – no vendor lock-in.
Uploaded my goals. I literally took a photo of my notebook and had Claude transcribe it.
Created a Context File. This defines work, family, time constraints, and cognitive distortions.
Uploaded Frameworks. I fed the AI the scientific frameworks for goal setting and tracking (WOOP, etc.).
Tasked Claude Code. I asked it to organize the project structure based on all inputs.
Tasked Gemini 3 Pro. I asked it to review the structure.
...Several iterations of edits...
...2 weeks of debugging... (I added a simple database for task tracking and a calendar database).
Profit.
The final structure
goals-template-2026/
├── Vision_2026.md # The main map (generic categories)
├── Memory.md # AI Instructions (Persona & Rules)
├── README.md # Setup Guide
├── Pillars/
│ └── pillar-template.md # Template for recurring systems
├── One-off projects/
│ └── project-template.md # Template for finite projects
├── Side quests/
│ └── quest-template.md # Template for hobbies/fun
└── frameworks/
└── project_template.md # The master template to copyThe context
I added two critical layers of context to my system:
Cognitive Distortions: The psychological patterns I’m currently working on.
Life Constraints: My reality – like having a small child, which means plans often change unexpectedly.
I also applied the concept of Progressive Disclosure when designing my AI coach.
The result
I now have a Coach/PA that helps me:
Set goals in a way that makes them achievable.
Track those goals without the manual headache.
Identify blockers (both external and internal).
Overcome demotivation.
Work through my cognitive distortions in real-time.
Keep all related notes and information related to each project.
Balance work and life.
Work through my cognitive distortions in real-time.
My most favorite feature so far: organize meeting transcripts into tasks and allocate to appropriate projects.
