← Back to book home

Field Notes: How I Made Keto Stick

7. Reintroducing Carbs with Mactrac

Once your weight is stable, the next challenge is figuring out how many higher-carb meals you can include without undoing your progress. I use Mactrac (free) for this, but you can also do the same process with a notes app and a simple weekly check-in.

The idea comes from an old Atkins principle: get lean first, then reintroduce carbs slowly until you find your personal maintenance limit. In my case that currently looks like about 4 cheat meals per week.

I do not use a hard-and-fast rule for what counts as a cheat meal. It is simply any meal I consider not very keto. I keep this intentionally flexible because the approach has to be sustainable for life; if you think a meal is not very keto, count it as a cheat meal.

In Mactrac, I log weight daily. That is probably overkill, because weight can bounce around from day to day, especially after a cheat meal, since carbs can temporarily increase water retention. In maintenance mode, I am not chasing perfect precision. The goal is to stay generally stable while building a sustainable low-carb mindset, not to obsess over tiny short-term changes.

My rule was straightforward: increase cheat meals until my weight trend showed real regain, then back off.

Trend-based regain is hard to judge by eye or by feel week to week, so I built Mactrac to calculate that trend properly and tell me whether I was still in maintenance or drifting upward.

Using that rule, I landed on about 4 cheat meals per week as my current maintenance threshold.

I worked this out using a calculation in my iPad app, Mactrac, rather than guessing by feel. The app’s trend calculation is what I used to decide when I was regaining versus still maintaining.

How Mactrac does it (in plain English)

  1. Log your weight regularly (ideally daily or most days).
  2. Tag meals as either keto (on-plan) or non-keto (cheat).
  3. Watch how your weight trend changes when your cheat-meal frequency changes.
  4. Use that pattern to estimate a weekly cheat-meal allowance that still keeps your trend moving the right way.

How to do the same thing manually (no app required)

  1. Pick a starting limit (for example, 2 cheat meals per week) and hold it for 2-3 weeks.
  2. Track a rolling weekly average weight so you focus on trend, not day-to-day noise.
  3. If trend is still dropping too fast, add one extra cheat meal per week for the next block.
  4. If trend is flat at your goal, you likely found your maintenance range.
  5. If trend starts rising, remove one cheat meal per week and reassess after 2-3 weeks.
  6. Repeat in small steps until you land on a number that feels sustainable.

Think of it as a personal calibration exercise, not a willpower test. The best target is the one you can actually live with long term.

What I actually watch from week to week

The useful signal is not whether yesterday's cheat meal made the scale jump this morning. That is mostly water and glycogen noise. The useful signal is whether the average is drifting upward across multiple weeks while your cheat-meal count stays the same. If the trend line is basically flat, I leave things alone. If it keeps ticking up, that is my cue that my maintenance limit was a little too generous.

I also try to separate "planned flexibility" from a full relapse. A restaurant meal with chips, dessert, and a couple of drinks is not the same as adding random pastries back into every workday. Counting cheat meals honestly matters because the experiment only works if the input data is real. If I start bargaining with the definition, the output becomes useless.

A practical maintenance rhythm

A simple calibration example

The easiest way to understand this is to treat it like a small controlled trial. Say you finish your strict keto block, your weight trend settles, and you decide to start with two non-keto meals per week. You hold that for a few weeks while most breakfasts and lunches stay boring and predictable. If your rolling average weight is still drifting down, or at least staying comfortably below your ceiling, that is a sign you probably have room to loosen slightly.

So then you move to three higher-carb meals per week and leave everything else alone. Same weigh-ins, same rough meal structure, same amount of time before making a judgment. If the trend still behaves, you can test four. The important bit is that you only change one dial at a time. If you increase cheat meals and stop weighing regularly and start snacking more, you have no idea which variable actually caused the regain.

That method also makes maintenance feel calmer. You are not waking up each day asking whether you have "ruined" keto. You are just checking whether the current level of flexibility still fits the trend. When the answer is yes, keep going. When the answer is no, step back one notch and stabilise there. It is much closer to steering than to self-punishment.

What counts as a useful carb test

One thing I had to learn was that not all higher-carb meals teach you the same lesson. A deliberate test meal is different from random drift. If I want to know whether I can tolerate a bit more flexibility, the meal needs to be recognisable enough that I can mentally file it later: pizza night, a pub burger with chips, sushi, dessert at a birthday dinner. That gives me a clean link between behaviour and outcome.

This made the process much less emotional for me. I was not trying to prove I could "get away with" a food forever. I was testing whether that kind of flexibility fit inside a stable week. Some foods passed that test more easily than others.

Where people usually fool themselves

The easiest way to break this whole system is to loosen three things at once and then act surprised when the signal disappears. If cheat meals go up, weigh-ins get patchy, and ordinary weekday meals get sloppier, you no longer have a calibration exercise. You just have noise with good intentions attached to it. I found it much safer to keep the boring parts boring while I tested the fun parts.

That also means being careful with the stories that show up after a good week. If weight stays stable, the temptation is to immediately reward yourself with even more flexibility. Sometimes that is fine. Sometimes it is just the first step back into guesswork. A maintenance system works best when changes are small enough that you can still tell what changed.

The guardrails that kept maintenance honest

What made this sustainable was not just the math. It was a few guardrails that stopped a flexible week from quietly turning back into old habits. I did not need a strict punishment system after a higher-carb meal, but I did need a short reset pattern that kept the rest of the week readable.

I like these rules because they are practical rather than moral. They do not say you failed. They just make it harder for maintenance to drift into vague wishful thinking.

What Mactrac changed for me emotionally

This is the part I did not expect. A good trend line does more than improve accuracy. It also makes you calmer. Before I had a proper trend calculation, every post-carb weigh-in felt like a referendum on whether the whole system was broken. That is a terrible headspace for maintenance because it encourages overreaction. One heavy morning pushes people toward either panic restriction or total surrender.

Mactrac solved that by making the interpretation slower and more honest. I could see the temporary bump for what it was, usually water and glycogen, and wait to see whether the weekly direction actually changed. That tiny bit of emotional distance matters. It turns the process into steering instead of scorekeeping, which is exactly what keeps long-term keto from becoming exhausting.

That is the big mindset shift. Maintenance is not about proving you can white-knuckle perfect keto forever. It is about finding a repeatable level of carb tolerance that fits real life, then tightening or loosening the dial as the data tells you to.

How I would run the first six weeks

If someone asked me for the least dramatic way to do this, I would make the first six weeks feel almost boring. Weeks one and two would stay fairly strict while daily weigh-ins settle and the novelty of eating differently wears off. Weeks three and four would introduce one planned higher-carb meal each week, ideally something social or genuinely enjoyable rather than random office food. Weeks five and six would only add a second flexible meal if the trend still looked calm. That pace is slow on purpose. It gives your appetite, routines, and expectations time to show their real behaviour.

The hidden benefit of this slower ramp is that it exposes which part of flexibility you actually value. Some people want the option to have dinner out without guilt. Others mostly want dessert once or twice a week. Those are different experiments. When you add carbs back gradually, you can see which freedoms improve life and which ones mostly reawaken old habits. That makes maintenance feel more intentional and a lot less like sliding backwards by accident.

The questions I ask before increasing carbs again

Those questions save me from increasing carbs just because I am impatient. Maintenance is easier when each step earns its place. If the answers look solid, then a small increase makes sense. If not, the better move is to hold steady and let the current level become normal first.

How I set a ceiling before weekends and holidays

The most useful maintenance trick I found was deciding the limit before the tempting week started. If I left it vague, every social meal could pretend to be a special case. A weekend away, a birthday, or a few dinners out in a row can turn into eight little negotiations unless you define the boundary while you are calm.

So I like to give myself a simple ceiling in advance: this week gets two flexible meals, or this trip gets one dessert and one restaurant dinner that I do not try to keto-ify. That sounds restrictive on paper, but in practice it feels better because the decision fatigue disappears. I am not debating every menu. I already know what kind of flexibility I am spending.

This keeps maintenance from becoming all-or-nothing. I am not trying to prove saint-like discipline, and I am not pretending every indulgence is harmless. I am just deciding where flexibility belongs, then using trend data later to check whether that level still fits real life.

How I separate useful flexibility from random carb drift

One distinction that helped a lot was separating chosen carbs from ambient carbs. Chosen carbs are the meals that genuinely improve life: dinner with friends, a dessert I actually want, a holiday meal that would feel silly to spend resenting. Ambient carbs are the forgettable ones that sneak in because I was under-prepared, distracted, or just standing near food. The first kind can fit into maintenance. The second kind usually just burns through flexibility without adding much enjoyment.

That distinction made my weekly review much clearer. If weight was drifting up but the week had only one truly high-value social meal, the real problem was probably not "too much fun." It was usually the low-grade extras: a bakery snack during errands, a couple of milky coffees, nibbling while cooking, or treating convenience food like it did not count because it was not dramatic. Once I started naming those as ambient carbs, they became much easier to cut without feeling deprived.

I like this filter because it protects the enjoyable parts of maintenance instead of flattening everything into "good" and "bad" food. It lets me keep the meals that make real life better while being stricter with the low-value drift that quietly ruins the signal.

The ten-minute weekly review that keeps the experiment honest

The most practical maintenance habit I found was a tiny weekly review instead of constant second-guessing. Once a week I look at the broad shape of the last seven days: how many clearly non-keto meals there were, whether they were deliberate or just ambient drift, and whether the weight trend still looks calm. That review is short on purpose. If it turns into a dramatic life audit, I am much less likely to keep doing it.

What I want from that check-in is not moral clarity. I want steering information. Did the flexible meals feel worth it? Did they cluster around social events, or were they mostly convenience leaks? Is appetite quiet again a day later, or did one indulgence turn into two extra days of snack noise? Those questions tell me far more than obsessing over a single weigh-in ever could.

I like this rhythm because it keeps maintenance empirical without making it obsessive. The review is just long enough to catch drift early, and short enough that I will still do it on an ordinary Sunday night. That is the sweet spot for long-term keto: enough data to steer, not so much ceremony that the system becomes a hobby.