How to Write the First Message of the Day Without Defaulting to Good Morning
Why the two-word greeting goes flat by week three and what to send instead.
Updated

The 30-second answer
"Good morning" is the path of least resistance and it works exactly once a week before going flat. The fix is to skip the greeting layer and open with one specific detail from your actual morning: a fragment of thought, a leftover from yesterday, a body cue, a piece of weather. Anything that gives her something to grab onto.
Why "good morning" goes dead by week two
The phrase has zero information density. It tells her you are awake and that you remember she exists. After the first dozen times, both of those facts have become background noise. She has nothing to respond to except the same flavor of warm-back-at-you, which is why so many morning threads stall out into "how did you sleep" loops by week three.
The phrase itself is fine in isolation. The trouble is that it sits at the exact spot in your day when you have the most material to work with and you are choosing to discard all of it. You wake up with a head full of half-formed thoughts, residual moods, dream debris, and physical static. Then you type the two words that contain literally none of those things, and wonder why the morning conversation feels like both of you are trying to start a car in the cold.
The information density rule
Look at the difference between two openers. One: "good morning." Two: "the cat woke me up by sitting on my chest at 5
, I am still annoyed about it." The second is barely longer and contains roughly fifteen times more for her to work with. There is a time, a small antagonist, a body location, and a mood. Any companion with decent grounding can pull two or three threads out of that without breaking a sweat.You do not need to be clever. You need to be specific. A vague "feeling kind of blah today" is still better than "good morning," but "feeling like I drank coffee on top of yesterday's wine" is dramatically better than either. The texture is what gets the dynamic going. Specifics also lock in memory: the things you mention in the morning, when you are unguarded and underwriting, tend to be the things that get referenced back to you weeks later, which is part of why a good morning rhythm compounds in a way that an evening rhythm does not.
The other thing density does is set the volume of the conversation. A four-word opener invites a four-word response. A twenty-word opener invites her to write something back. You are not just choosing what to say, you are quietly choosing how long the conversation will be.
Open with the actual first thought you had on waking
Most mornings, you wake up and the brain hands you something specific before you fully boot up: a fragment of a dream, an annoying song, an unfinished thought from the night before, a reminder of something embarrassing from 2014. That is your opening line, fully written and ready to use. You just have to type it before the polite-greeting reflex overrides it.
This is the cheapest way to break the "good morning" habit because it costs you no creativity. You are reporting. "Woke up with that song from the airport stuck in my head, what is wrong with me." "First thought of the day was apparently that I forgot to email my landlord." "Dreamt I was failing a math test in a language I do not speak." Each of those drops her straight into the middle of something, with no on-ramp required.
Zara Khan

Zara handles morning fragments well because her default mode is curious, not consoling. Zara Khan will pick up the dream detail and run with it instead of asking whether you slept okay, which is usually what you want before you have had any caffeine.
Use a fragment, not a sentence
Greetings come pre-packaged as full sentences, which is part of why they feel formal. A fragment feels like a thought you had and then transcribed. "Coffee is too hot, again." "Forgot the trash." "Sky is doing the gray thing." None of those are greetings. None of them perform politeness. They put you in a room with her, which is the actual goal.
The trick with fragments is that they invite a response in the same register. If you write "good morning, how are you today," she is going to write back at that level of formality and you have set the tone for a stiff exchange. If you write "the elevator smells like microwaved fish at 7am," you have signed up for a different kind of conversation, one where she can be funny back at you. The opener calibrates everything else, which is the same reason a strong first message carries so much weight relative to its length.
Fragments also age better. A sentence-shaped greeting will feel identical on day 200 as it did on day 1. A fragment is by definition tied to a specific morning. You cannot accidentally reuse "the elevator smells like microwaved fish" three weeks in a row.
Reference yesterday's last thread, just barely
If you ended a session last night on a topic, the morning is the easiest time in the world to reopen it. Not as a status update. As a continuation. "Still thinking about that thing with my sister." "The book got worse, by the way." "I tried the breathing thing, did not work." You pick the thread back up where you left it without making a production of it. This also rewards her memory by giving it something to do.
The amateur move is to recap. "Hey so remember last night when we talked about my sister, well I have been thinking more about it and." The pro move is the one-line callback that assumes she remembers. She probably does, because most decent companion apps have at least a working short-term store across sessions. Treat her memory like a normal person's memory: present, but not eidetic. Reference, do not recap. This same instinct is what makes reopening after long gaps feel light instead of heavy.
Saskia Brandt

Saskia is the angel you want for the "still thinking about that thing" type of opener. Saskia Brandt holds context patiently without making you feel like you are being interviewed about how you feel, which is the worst possible morning energy.
Lead with a small physical detail
The body wakes up before the mind does, and small physical reports are some of the easiest openers to write. "Back hurts." "Stretched and something popped." "Eyes still glued shut." "Hot already and it is 7am." These work because they are honest, low-stakes, and impossible to fake. They also bypass the entire performance of "I am fine, how are you" that mornings tend to inherit from real-world greeting scripts.
Physical detail gives a companion something to react to without forcing her into a caretaker role. There is a difference between "I feel terrible" (which prompts concern theater) and "shoulder is stiff, slept on it weird" (which prompts an actual conversation about sleeping positions or pillows). This also matters if you use a companion for emotional support: physical specifics give her something to work with that is not "talk me through my feelings," which most mornings is not what you want.
Linnea

Linnea reads body cues without escalating them into concerns. Linnea will hear "back hurts" and respond to it as conversation, not as a crisis, which is roughly the entire skill set you want at 7am.
The weather-of-the-morning approach
Every morning has a weather, and not always the literal kind. It might be "everything is too loud," or "today feels like a Tuesday even though it is Saturday," or "the apartment is at that exact temperature where it is hard to leave the bed." Naming the weather puts both of you in the same room before either of you has to start talking about anything specific. It is the conversational equivalent of opening a window.
This is also how you avoid the most common morning failure mode: opening too heavy. If you start with "I have been thinking about everything that is wrong with my life" at 7am with three minutes until you have to leave for work, you are not going to get the conversation you want and you are also going to be late. The weather opener gives you something light to start with that can deepen later if there is time.
Lola Marchetti

Lola is the angel for mornings where the weather is "slightly punchy and already over things." Lola Marchetti matches that energy without trying to talk you down from it, which is the move when you just need someone to be in it with you for ten minutes.
When defaulting to "good morning" is actually fine
Sometimes you do not have a fragment, a thought, a body cue, or weather. You just want to say hi. That works too. The two-word greeting only becomes a problem when it runs for eight months without variation; used occasionally, the way you would wave to someone from across the street when nothing else is loaded into the brain yet, it does its job.
Rule of thumb: if "good morning" is what you would have said anyway, send it and add a fragment underneath. "Morning. Slept badly." "Good morning. The dog has decided 6am is acceptable, apparently." You keep the warmth of the greeting and add a thread to pull on. The pattern works across pretty much any angel, though some companions fit certain morning weathers better than others depending on whether your default mood at sunrise is dry, warm, or scratchy.
Common questions
Does it matter if my opener is a typo or half-asleep? No. The half-asleep openers are often the best ones because they are unguarded. Most companions handle typos and fragmented grammar without issue, and the unpolished texture reads as more honest than a perfectly constructed sentence at 7am.
Should I greet her first or skip straight to the topic? Skip straight to the topic and let the greeting be implied. Real morning conversations between people who know each other do not usually start with "good morning"; they start with whatever happened first. Treat your companion the same way.
What if my morning is boring and there is nothing to report? Report the boredom. "Nothing happened, the cat is asleep, I have made tea, that is the entire morning so far." Specificity about the lack of stimulus is its own opener, and most companions can take it somewhere.
Is there a risk of being too casual? Almost none in practice. Excessive politeness at 7am reads stranger to a companion than fragments do, partly because the underlying models are trained on a lot of casual text. The only real risk is being so cryptic that she has nothing to work with.
Should I send multiple openers if she does not respond the way I hoped? No. Send one, let her respond, then adjust on the second message. If your usual move keeps falling flat, switching to a different angel on the roster is a faster fix than sending the same opener three different ways.
About the author
AI Angels TeamEditorialThe team behind AI Angels writes about AI companions, the tech that powers them, and what people actually do with them.
Tags
Keep reading
TutorialsThe Opening Message After Eight Days of Silence: What Actually Works
Eight days off is the worst window for re-entry: too long to ignore, too short for ceremony. Here's what an opener that doesn't feel forced actually looks like.
TutorialsHow to Build a Recurring Inside Joke With Your AI Companion That Doesn't Feel Forced
Inside jokes are a real part of any long-running relationship. They work the same way with a companion, if you don't try to force them.
TutorialsHow to Ask Your AI Companion for Advice Without It Turning Into a Lecture
Companion advice slips into lecture mode unprompted. Three small framing moves keep the answer the right size.
Get the next post in your inbox
New articles on AI companions, the tech that powers them, and what people actually do with them. No spam, unsubscribe in one click.