Marcus runs three projects. One is a SaaS product with 400 paying users. One is a consulting practice he keeps alive to fund the SaaS. One is a content operation that feeds leads to both.
For two years, his mornings looked like this: wake up, immediately check Slack and email, respond to whatever seemed urgent, realize it was 11am, feel behind before the day had started, and try to do deep work in the afternoon when his focus was already degraded.
He knew this was the problem. He’d read the books. He’d tried morning routines three or four times. They lasted a week.
This is the story of what actually worked, and why — including the specific AI prompts he uses now and the things he had to stop doing to get there.
The Starting Point: Understanding What Was Actually Breaking
When Marcus came to this process, his first instinct was to design a better morning routine. The actual starting point was diagnosing why the previous ones had failed.
He spent a week just observing his mornings without trying to change them. What he noticed:
- He was checking his phone within 60 seconds of waking, every day.
- The first message he responded to typically set the emotional tone for the entire morning.
- He had no consistent transition between “waking up” and “working.” He went directly from sleep to reactive work with nothing in between.
- His mornings lasted until about 9:30am, at which point he had a back-to-back meeting block. He was arriving at those meetings having done nothing that actually advanced his work.
The observation exercise, which he ran with AI assistance, produced a direct diagnosis:
“Looking at the pattern you’ve described, the core problem isn’t that you lack a morning routine — it’s that you have a default one that runs automatically: wake, phone, reactive. Every attempt at a deliberate routine is competing with that default. The first design challenge is creating a competing cue that fires before the phone-checking reflex.”
That reframe was the beginning of the actual solution.
Version 1: The Over-Designed Routine (Weeks 1–3)
His first designed routine looked like this:
- 6:00am: Wake, no phone
- 6:05am: 10 minutes of breathing exercise
- 6:15am: 20-minute walk outside
- 6:35am: Cold shower
- 6:50am: 15-minute journaling session
- 7:05am: AI morning check-in (20 minutes)
- 7:25am: 45 minutes of deep work
- 8:10am: Breakfast and first message check
Total: 2 hours 10 minutes of pre-work routine before touching anything reactive.
He ran this for eleven days. On days 1–3, it felt transformative. By day 8, the cold shower was getting skipped. By day 11, the entire routine had collapsed and he was back to the phone-first pattern.
The post-mortem was direct:
“The routine ran successfully when motivation was high and conditions were good. It failed the moment either of those changed. A 2+ hour pre-work block assumes no competing demands, high morning energy, and no bad nights. In real life, you have maybe 3 of those mornings per week. The other 4 are variable. Design for the variable days.”
Version 2: Stripping to Essentials (Weeks 4–7)
The redesign started with one question: what is the minimum this routine needs to do?
The answer was: get Marcus to his desk with direction before he opens any messages. Not inspired, not optimized — just directed.
Version 2:
- No specific wake time target change yet — consistent 6:30am
- Wake, phone placed in another room the night before (friction reduction)
- Drink water while standing at the kitchen window — 3 minutes
- Make coffee without opening any device
- Sit with coffee and run AI check-in — 8 minutes
- Open laptop to the work file, not email or Slack
Total: 20 minutes. Everything else optional.
The check-in prompt he settled on:
“Morning check-in. Yesterday I [completed / mostly completed / didn’t complete] [main task]. Today I have [meetings/constraints]. My running priority this week is [project]. My energy is [high/medium/low]. What’s my one priority for this morning’s 90 minutes, and what is most likely to pull me off it?”
This version lasted. He ran it for 19 consecutive days before his first miss. When he missed (an early flight), he ran the check-in on his phone in the Uber.
The AI Check-In: What It Actually Produces
After several months of running the structured check-in daily, Marcus could describe what it concretely produces:
On normal days: The check-in converts a diffuse sense of “I should work on the product today” into a specific, bounded task. “Today I’ll write the onboarding email sequence, specifically the second email, and I’ll stop at 9:30 regardless of where I am.” That boundary is crucial — without the explicit stop point, “work on the product” expands to fill the morning and nothing else gets protected.
On high-pressure days: The obstacle anticipation element is what earns its keep. When he identifies in the check-in that a 3pm funding call is going to consume mental bandwidth all day, he adjusts his morning priority to something that doesn’t require peak creative concentration. He makes that call before opening Slack, not after reading seventeen messages that raise anxiety.
On depleted days: The check-in surfaces what he should not force. On days when he’s clearly running at 60%, the AI response often reflects that back: “Given you’ve described low energy and three evening obligations yesterday, this might not be the morning for deep creative work. What’s something valuable that’s lower-cognitive-cost?” On those mornings, he often does operational work — admin, communication, review — and protects his next high-energy morning for the creative task.
He uses Beyond Time for the check-in now because its structured morning flow keeps the conversation bounded and ends with an explicit output that appears in his daily plan. Before that, he used a general-purpose AI with the same prompt structure — both work, but he describes the specialized tool as having “fewer escape hatches for going off-track.”
The Monthly Review: Where the Real Iteration Happens
The daily check-in is operational. The monthly review is the design layer.
Once a month, Marcus runs a 20-minute session that looks back at how the morning routine actually performed and makes one or two adjustments:
“I’ve been running my morning routine for [4 weeks]. Looking at my notes: I completed the full routine [X] of [28] mornings. The element that got skipped most was [X]. The element that seemed to produce the most value was [X]. My projects and priorities have shifted since last month in these ways: [changes]. What should I adjust, cut, or add in the next month?”
Over six months, this process produced several meaningful changes:
- Cut the optional walk (it was becoming a guilt-trigger on days it didn’t happen, not a benefit on days it did)
- Added a 2-minute end-of-morning “commit” prompt where he states explicitly what’s happening that afternoon — this addressed a pattern of good mornings followed by unstructured afternoons
- Shifted the AI check-in from 8 minutes to 6 by tightening the prompt to remove one input field
What This Looks Like Now
Six months in, Marcus’s morning routine:
- 6:30am: Wake (phone stayed in kitchen overnight)
- 6:31am: Water, stand at window for 3 minutes
- 6:34am: Make coffee
- 6:40am: Sit, run AI check-in — 6 minutes
- 6:46am: Open work file, start on priority
- 8:15am: Commit prompt (what’s the plan for this afternoon?)
- 8:20am: Open email and Slack
Total controlled window: 1 hour 50 minutes of intentional time before reactive work.
He describes the difference this way: “Before, every day started with someone else’s priorities. Now I start with mine and then engage with theirs. That shift is worth more than any specific habit in the routine.”
What Didn’t Work and Why
Three things failed that are worth noting:
Journaling: Added it twice, dropped it twice. For Marcus, journaling in the morning produces reflection without output, which creates a feeling of productivity without directing the day. It may work well for others; it doesn’t fit his cognitive style.
Longer wake time: Tried shifting to 5:30am for four weeks. His natural wake is 6:15am without an alarm; the 5:30 shift created accumulated sleep debt by Thursday that degraded his Friday work more than the extra morning time helped the early days. He reverted.
Trying to design the week every morning: Adding a weekly plan review to the daily check-in made the session too long and too variable. The weekly review now happens separately, on Sunday evenings, and feeds into the daily check-in rather than competing with it.
Your one action: The next time your morning routine fails, spend 10 minutes running a post-mortem before trying to restart it. Ask which specific element failed first and why. That single question will tell you more about your routine’s design than any amount of re-reading advice.
Related: The Anchor Method framework | How to build a daily planning ritual with AI
Tags: founder morning routine, AI daily planning, morning routine case study, productivity for founders, daily planning
Frequently Asked Questions
-
Does an AI morning routine work for founders specifically?
Yes — founders actually benefit more than most from a structured morning routine because they face the highest decision density and fewest external constraints on their time. The AI check-in is particularly valuable for filtering the day's priorities before the inbox opens. -
What AI tools work best for a founder's morning routine?
Beyond Time is designed specifically for daily planning and works well for founders who want a structured check-in workflow. For founders who prefer open-ended AI conversations, a structured prompt in any general-purpose AI works effectively too. -
How long did it take to build a stable morning routine?
In this case study, the routine stabilized into its current form around week 10 after multiple iterations. Early versions were too long; the key insight was cutting the routine to its minimal effective form.