Gilles Crofils

Gilles Crofils

Hands-On Chief Technology Officer

Tech leader who transforms ambitious ideas into sustainable businesses. Successfully led digital transformations for global companies while building ventures that prioritize human connection over pure tech.1974 Birth.
1984 Delved into coding.
1999 Failed my First Startup in Science Popularization.
2010 Co-founded an IT Services Company in Paris/Beijing.
2017 Led a Transformation Plan for SwitchUp in Berlin.
November 2025 Launched Nook.coach. Where conversations shape healthier habits

Your lunch is too fast to count in your brain

Abstract:

The article explains why a rushed, distracted “desk lunch” squeezed into a 6‑minute gap between meetings often doesn’t feel like it “counts,” even if you ate enough—because the day treats eating like clearing notifications (container by the keyboard, three bites during a “quick question,” Slack ping mid‑chew, then the lid closes and you’re back in the doc). It argues this isn’t a willpower or meal-prep problem but a signal-quality problem driven by two mechanisms: satiety lag (fullness signals arrive after you’ve already finished) and meal memory (when attention is split, the brain barely logs that a meal happened), which can lead to a noisy afternoon of phantom hunger, unplanned snacks, extra kitchen trips “just to check,” coffee #2/#3 feeling mandatory, and dinner turning into recovery mode—an effect supported by research on distracted eating and a vivid amnesia study where people who couldn’t remember eating accepted multiple meals close together. To break the loop without moralizing or heavy tracking, it proposes a small 5‑day experiment: an “8‑minute landing protocol” with a 2‑minute screen-free ramp-in, a 4‑minute deliberately slower core, and a 2‑minute ramp-out with a clear ending marker (plus “degraded” versions for brutally packed calendars and a tiny reboot for interruptions), alongside a minimalist debug log that checks whether you feel steady, snacky, or sleepy 60–120 minutes later and counts a few mid‑afternoon behaviors (unplanned snacks, kitchen trips, purchases). The piece also distinguishes this from stress-muted midday appetite, notes caffeine and sleep can amplify the cycle, and adds safety cautions for eating-disorder risk, medical nutrition needs, and GI or hypoglycemia issues, framing the whole problem as a fixable system with bad logging rather than a personal flaw.

Lunch that doesn’t register

If your lunch happens in the 6-minute gap between 2 meetings, it’s not really a meal. It’s more like clearing notifications. Container open next to the keyboard. 3 bites while answering a “quick question.” Slack ping mid-chew. Then the lid closes and you’re back in the doc like nothing happened.

And later, somehow, you’re hungry again.

I’ve done this for years at a desk—Beijing, then Berlin, now Lisbon—often past midnight.

This is for that exact loop. Not for perfect meal prep. Not for macro math. And not for advice that assumes you have a calm 45-minute lunch break and a separate personality for weekdays.

The goal: fewer 15:00–17:00 “episodes” without changing what you eat. Make lunch land so the rest of the day gets quieter. Less random kitchen drifting. Fewer unplanned snacks that don’t even feel like a decision. Coffee #2 (or #3) feeling optional again. Dinner not turning into recovery mode.

We’ll walk through:

  • Why desk lunch often fails to “count” in the brain, even when you ate enough
  • The 2 main mechanisms behind it: satiety lag (your signals arrive late) and meal memory (your brain barely logs the meal if attention is split)
  • A practical cause-and-effect map of the desk day loop, based on what’s measurable in real work life (calendar saturation, eating while working)
  • A small experiment called the 8-minute landing protocol, plus degraded versions for days when your calendar laughs at you
  • A tiny debug log that doesn’t turn your life into a tracking app, just enough to see whether anything is improving

No moralizing. No “just slow down” hand-wave. More like fixing a system with bad logging and too many interrupts. Because if lunch gets treated like background noise, your appetite will keep filing tickets all afternoon.

The high speed lunch low signal meal

Lunch that feels like clearing notifications

There’s no clear ending, just an empty box that gets closed like a laptop lid.

This isn’t you being weird. It’s the environment. Microsoft’s Work Trend Index has been flagging calendar saturation and constant time pressure. And US time-use diaries (ATUS, Bureau of Labor Statistics) show eating often isn’t a single protected activity on workdays. It overlaps with other tasks. So even when you eat a real lunch, the context can make it register more like scattered snacking than a finished meal.

So if this isn’t “you chose the wrong lunch,” what is it?

  • A sudden “urgent snack” feeling around 15:00–17:00
  • Coffee #2 (or #3) stops feeling like a choice
  • More kitchen trips than expected, sometimes without a clear plan
  • Extra eating episodes that don’t register as meals
  • Dinner quietly turns into recovery mode

These are useful because you can notice them without turning life into a tracking app. Counting “episodes,” coffee timing, and kitchen trips is often enough to see the pattern.

This isn’t about perfect ingredients or the right macro split. It’s about signal quality. Speed plus fragmentation, with no clean boundary. Plenty of experts say “slow down,” but they rarely give something that fits inside an actual workday.

Why the meal does not land

2 mechanisms your desk day skips

Speed is one half of the problem. The other half is attention. Meaning what your brain stores as “i ate” versus “something happened near my keyboard.”

Mechanism 1 is satiety lag. Your gut and brain communicate, but the messages arrive late, like notifications on a bad connection. If lunch is finished before the feedback loop catches up, you can end the meal feeling oddly unchanged.

In studies where people eat the same food at different speeds, slower eating tends to increase satiety (Kokkinos et al., 2010)—meaning you’re less likely to get that “urgent snack” feeling at 16:00 even if lunch was the same size. No universal “eat in X minutes” rule needed. The point is simple: it’s possible to out-run your own fullness signals.

Mechanism 2 is meal memory. Distraction doesn’t just change the meal. It changes what happens after. In the Robinson et al. (2013) meta-analysis, distracted eating is reliably linked with more eating later on. If your brain can’t really remember lunch, it keeps looking for “proof” later.

A dramatic example makes the point: in Rozin et al. (1998), amnesic patients who couldn’t remember eating were willing to accept multiple meals close together. Desk work is not amnesia, obviously. The takeaway is that memory is one input in appetite regulation.

Put both mechanisms into a meeting-packed day. Weak satiety signal (because speed outruns feedback) plus weak confirmation (because attention is split) creates a messy result. Then the system tries to patch it with extra snacks, extra coffee, and extra kitchen laps. You’re not even that hungry in the stomach, but the brain keeps sending “go check the kitchen” anyway. Wired-tired.

Caffeine can also muddy the logs. It may suppress hunger briefly for some people, but it does not reliably reduce total intake (Schubert et al., 2014). And late caffeine can hurt sleep (Drake et al., 2013), which can make the next day’s appetite regulation harder.

A desk day cause and effect map

The loop in 4 steps

In a meeting-dense day, fast lunch is a rational response to restart cost. Step away, lose context, then pay the tax to “get back in.” With calendar saturation being a measured pattern (Microsoft Work Trend Index), minimizing offline time becomes a survival move, not a character flaw. Time-use diaries show the same thing from another angle: eating often overlaps with work (ATUS). So the fix has to work inside overlap, not pretend overlap doesn’t exist.

When the system doesn’t clearly confirm a meal, it starts sending you to look for more input. Think of lunch like a transaction. If it doesn’t close, the “did we actually do that?” flag stays open.

  • Speed shortens oral processing time, worsening the timing mismatch
  • Discontinuity splits attention, so the brain stores less of a clear “meal happened” record
  • Result: lunch gets encoded as small packets, not a completed event

Then come the patches, and they make the rest of the day noisier.

  • More eating episodes because “real lunch” quietly turns into 3–6 micro-events
  • Extra unplanned snacks, the kind you didn’t decide on in the morning
  • More kitchen trips “just to check”
  • Desk-side grazing like candy bowls or leftover meeting pastries

These are decent diagnostics because they don’t require calorie math. Planned vs unplanned snacks and “episode count” already capture a lot.

There’s also a sleep spillover. Late caffeine can steal sleep (Drake et al., 2013). Short sleep is linked with higher intake and more snacking in controlled settings (Markwald et al., 2013; St-Onge et al., 2012). So the day ends noisy, the night gets lighter, and the next day starts with a fussier appetite system. Not moral failure. Just a feedback loop with bad logging.

How to tell this apart from nearby problems

A nearby pattern looks similar but feels different at noon: stress-muted appetite. One quick diagnostic question helps.

Did food sound appealing at noon?

  • If appetite is basically offline at midday, that can fit the stress story. Stress effects vary a lot person to person.
  • In the “ate but it didn’t land” story, you do eat, but later the system behaves as if it didn’t happen.

Liquids aren’t the villain. Speed often is. Thicker, slower liquids can sometimes land better than rushed solids.

Safety note so this doesn’t read like a universal work hack.

  • If there’s current or past eating disorder risk, focusing on meal pace can backfire.
  • If there’s malnutrition risk or medically required high intake, “slow down and stop when full” messaging can be harmful.
  • If there are GI motility issues (for example gastroparesis) or hypoglycemia treatment situations, timing and speed can be medically sensitive.

The 8 minute landing protocol

The protocol in one block

Treat this like a 5-workday experiment on 1 meal—usually lunch—not like becoming a new personality. No meal prep. The only variables are pace and a clear boundary so the meal has a chance to commit in your brain.

2-minute ramp-in

  • Sit if possible. If not, at least stop moving and face one direction.
  • First bite screen-free. No inbox, no Slack, no “quick reply while chewing.”
  • Mute inputs for 2 minutes. Phone face down, laptop pushed back 10 cm.

This is mostly about attention. Distracted eating is linked with more eating later (Robinson et al., 2013). Protecting the first bite increases the odds your brain logs “a meal is happening.”

4-minute slow core

Slightly too slow for a workday, but not ceremonial.

  • Smaller bites
  • Put the fork down sometimes
  • Let there be tiny pauses where nothing happens (yes, uncomfortable)

You don’t need rules like “30 chews.” You need a pace that gives signals time to arrive. Slower eating of the same meal tends to increase subjective satiety (Kokkinos et al., 2010). Two desk-legal cues: take a sip of water halfway, and don’t start the next bite until the previous one is fully done.

2-minute ramp-out

Utensils down, close the container, wipe hands, drink water. Then reduce re-entry friction with a 5-second note like “Next: review doc.” This is an end marker. Without it, lunch can stay like an open thread.

If your calendar laughs at 8 minutes, use degraded modes instead of quitting.

Degraded modes for real calendars

If you only have 6 minutes:

1) 90 seconds screen-free start

2) Eat normally for the middle

3) Clean stop (close container + water + “Next: ___” note)

If camera-on meetings make eating weird, apply this to breakfast or your first chewable intake of the day. Not just coffee. First time you eat something you have to chew.

Interruptions will still happen. Use the tiny reboot: 1 breath, screen away for 10 seconds, take 1 bite, then continue. It’s not elegant. It’s Tuesday.

If symptoms are severe or changing, treat it as health, not productivity. New or major fatigue, significant appetite or weight changes, or worsening GI symptoms are reasons to stop experimenting and get evaluated.

The 2 minute debug log

A 2 question check that fits a workday

Before lunch, do a 3-second baseline. Rate urgency to eat as low / medium / high. Not as a feelings exercise. More like a status ping.

Then check the outcome 60–120 minutes after lunch. Label your state as:

  • Steady (you can work, lunch is in the past)
  • Snacky (you start scanning for “something small”)
  • Sleepy (you want a nap, you’re reaching for caffeine, you’re reading the same sentence 4 times)

If you want one objective-ish metric, count behaviors in the 15:00–17:00 window for 5 workdays. And if you already wear something like a Polar H10 or even a basic Decathlon watch, don’t chase calories—just notice whether the post-lunch dip lines up with your snacky window.

  • Unplanned snacks (anything you didn’t bring or decide on in the morning)
  • Kitchen trips (even “just checking”)
  • Workplace purchases (receipt-level data is imperfect, but often less memory-dependent than intake guesses)

If the protocol is working, success looks boring.

  • Fewer phantom hunger spikes mid-afternoon
  • Fewer just-in-case snacks
  • Coffee #2 feels optional again
  • Dinner stops being recovery mode

If nothing changes, it doesn’t mean you failed. If episode count doesn’t drop after 5 days, test only lunch size next—keep the same 8-minute boundary.

If it gets worse, or focusing on pace triggers rigidity or anxiety, stop. A systems fix is only useful if it reduces noise without creating new problems. If it creates new problems, it’s the wrong lever, and clinician support beats debugging alone.

If lunch keeps happening in the 6-minute crack between meetings, it’s not a character flaw. It’s a system issue. The meal is fast, interrupted, and doesn’t get a clean “saved successfully” message in your brain. Then satiety arrives late, meal memory is fuzzy, and the afternoon gets noisy. More kitchen laps. More “something small.” Coffee stops feeling optional. Dinner becomes damage control.

The fix here isn’t perfect food or tracking everything. It’s signal quality. The 8-minute landing protocol (or the degraded versions) gives lunch a start, a slower middle, and a clear end, plus a tiny debug log to see if anything changes.

On days when lunch actually lands, the afternoon is quieter. Less scanning, less caffeine bargaining, less dinner damage control.

You might be interested by these articles:


25 Years in IT: A Journey of Expertise

2025-

Nook
(Lisbon/Remote)

Product Lead
Building the future of health coaching. Leading product development and go-to-market strategy for a platform that makes personal wellness accessible through natural dialogue.
Making health coaching feel like talking to a friend who actually gets you.

2024-

My Own Adventures
(Lisbon/Remote)

AI Enthusiast & Explorer
As Head of My Own Adventures, I’ve delved into AI, not just as a hobby but as a full-blown quest. I’ve led ambitious personal projects, challenged the frontiers of my own curiosity, and explored the vast realms of machine learning. No deadlines or stress—just the occasional existential crisis about AI taking over the world.

2017 - 2023

SwitchUp
(Berlin/Remote)

Hands-On Chief Technology Officer
For this rapidly growing startup, established in 2014 and focused on developing a smart assistant for managing energy subscription plans, I led a transformative initiative to shift from a monolithic Rails application to a scalable, high-load architecture based on microservices.
More...

2010 - 2017

Second Bureau
(Beijing/Paris)

CTO / Managing Director Asia
I played a pivotal role as a CTO and Managing director of this IT Services company, where we specialized in assisting local, state-owned, and international companies in crafting and implementing their digital marketing strategies. I hired and managed a team of 17 engineers.
More...

SwitchUp Logo

SwitchUp
SwitchUp is dedicated to creating a smart assistant designed to oversee customer energy contracts, consistently searching the market for better offers.

In 2017, I joined the company to lead a transformation plan towards a scalable solution. Since then, the company has grown to manage 200,000 regular customers, with the capacity to optimize up to 30,000 plans each month.Role:
In my role as Hands-On CTO, I:
- Architected a future-proof microservices-based solution.
- Developed and championed a multi-year roadmap for tech development.
- Built and managed a high-performing engineering team.
- Contributed directly to maintaining and evolving the legacy system for optimal performance.
Challenges:
Balancing short-term needs with long-term vision was crucial for this rapidly scaling business. Resource constraints demanded strategic prioritization. Addressing urgent requirements like launching new collaborations quickly could compromise long-term architectural stability and scalability, potentially hindering future integration and codebase sustainability.
Technologies:
Proficient in Ruby (versions 2 and 3), Ruby on Rails (versions 4 to 7), AWS, Heroku, Redis, Tailwind CSS, JWT, and implementing microservices architectures.

Arik Meyer's Endorsement of Gilles Crofils
Second Bureau Logo

Second Bureau
Second Bureau was a French company that I founded with a partner experienced in the e-retail.
Rooted in agile methods, we assisted our clients in making or optimizing their internet presence - e-commerce, m-commerce and social marketing. Our multicultural teams located in Beijing and Paris supported French companies in their ventures into the Chinese market

Cancel

Thank you !

Disclaimer: AI-Generated Content for Experimental Purposes Only

Please be aware that the articles published on this blog are created using artificial intelligence technologies, specifically OpenAI, Gemini and MistralAI, and are meant purely for experimental purposes.These articles do not represent my personal opinions, beliefs, or viewpoints, nor do they reflect the perspectives of any individuals involved in the creation or management of this blog.

The content produced by the AI is a result of machine learning algorithms and is not based on personal experiences, human insights, or the latest real-world information. It is important for readers to understand that the AI-generated content may not accurately represent facts, current events, or realistic scenarios.The purpose of this AI-generated content is to explore the capabilities and limitations of machine learning in content creation. It should not be used as a source for factual information or as a basis for forming opinions on any subject matter. We encourage readers to seek information from reliable, human-authored sources for any important or decision-influencing purposes.Use of this AI-generated content is at your own risk, and the platform assumes no responsibility for any misconceptions, errors, or reliance on the information provided herein.

Alt Text

Body