Inside the parallel WhatsApp universes of new moms and new dads in the first three years
My moms' group is not a chat. It is a shared nervous system: the ever-breathing record of care, anxiety, invisible knowledge and unpaid labor. I spent a month reading it as data, then read two dads' groups the same way. What came back was not a mirror. It was a map of who is carrying the weight.
It is 11:56 PM on a Tuesday. A mother in Bengaluru has been awake monitoring her toddler's fever since 10:40 PM. Six women in a WhatsApp group have been awake with her, trading dosing protocols, cold compress techniques and warnings about febrile seizures from their own experience. At 11:56 PM she types:
"I don't think I'm sleeping tonight anyway 😅"
A mother, April 22, 11:56 PMShe is not awake because she cannot sleep. She is awake because there is still work to do. This is not exceptional. In 662 messages across 31 days, this group had messages every single day, from 5 AM to midnight. Not one day without activity. No days off.
Here is what one month of those messages actually showed.
The two peaks of messages being sent on the moms' WhatsApp group are 10 AM and 9 PM. The first is the first free moment after the morning rush: tiffin, playschool drop, the maid coming or not coming. The second is the first free moment after bedtime. Both get consumed. The mental load does not clock out when the child does.
Each wedge is one hour. The tallest is 10 AM, the first peak of the group's day.
Each wedge is one hour. Taller bars = more messages. The dashed ring marks 9 PM. Anything beyond it is the post-bedtime peak.
Every message in this dataset was categorized by topic. A message about a child's fever landing at 11 PM counts the same as one about a nanny recommendation at 10 AM. Below is where the conversation actually lives.
Percentages sum above 100% because some messages touch multiple categories. A smaller 11th category (second-hand economy: product sales, giveaways) accounted for ~4% and is not shown separately.
The group exists almost entirely to care for others.
The sliver (barely visible) is for the mothers themselves.
One mother asked about tampons for the first time. It was 8:29 PM on a Wednesday. Sandwiched between a thread about screen time and one about chocolate spread. She got an answer in four minutes. Then the group moved on. It was the only thread that month with no follow-up question.
This is a composite reconstruction of a real thread from the dataset, anonymized. The timestamps are real. The sequence is real.
Each dot below is one member of the group. The ones in red asked at least one question this month, generating threads that others could read, absorb and use.
In the paid economy, this would be called a platform. A knowledge service. A community product. Here it is called a WhatsApp group. It runs on maternal labor. It costs nothing to join. No one is compensated for producing it.
Some cognitive labor is a one-time project. You research it, decide, move on. But some labor recurs, not because something went wrong, but because it is never finished. House help coordination is the clearest example in this dataset.
Hover over any cell to see what kind of search it was. Every week, without exception, the group ran at least one.
There is no formal infrastructure for finding house help in India. No verified database, no reliable agency for most areas, no portable worker history. So the search runs through "someone please share a contact" messages. And those messages, for mothers, run through WhatsApp. The coordination cost falls entirely on them, every week, without exception. The urgent week-4 search, the 24-hour nanny, almost always coincides with the regular maid being away in her village or the child's fever spiking. Often both.
Look at the fever thread again. There is a framework² that names cognitive labor in four stages, and all four ran in those eighty minutes between the first Calpol question and the cold compress instruction. M. anticipated: she warned about febrile seizures because her own son had one at 19 months. P. identified the options: Calpol now, ibugesic if it crossed 102. The mother decided. Then she monitored, hourly, until 4 AM. Four stages, one Tuesday night, six unpaid women.
But the framework is missing the stage the chat actually exposes. It is the stage that runs in the silence between 12:04 AM and the 5 AM check. The fifth stage is hold. To hold the anxiety of not knowing if tonight's fever will spike.
Stages 1 to 4: Allison Daminger's cognitive labor framework (2019). Stage 5, Hold: the continuous layer this dataset exposes, the one no taxonomy has yet named.
To hold six months of painful pumping while still deciding whether to switch to formula. To hold the guilt of playschool separation, the weight of weaning a child who is not ready, the hormonal aftermath of a body that just grew a person.
"Pumping all day long is becoming painful. Help!"
April 28, 10:44 AM"I teared up reading this. I am so exhausted."
The response, April 28, 10:50 AMThe group holds all of it, for free, at 10 AM on a Tuesday, with no credit and no category in any economic framework. This is the function no taxonomy captures and no GDP counts.
Two dads' WhatsApp groups: one Bengaluru-specific (115 members), one pan-India (963 members). I read through them the same way: counted messages, mapped topics, looked at when things were sent. What came back was not a mirror. It was an absence.
Active from 5 AM to midnight. Messages every day. Two peaks: 10 AM and 9 PM.
Concentrated in work hours. The messages mostly stop when the office Slack does. 3× more members; less than a quarter of the messages.
Both clocks on the same scale. Bar height = messages relative to the moms' peak (76 in one hour).
Adjusting for group size makes the gap wider, not smaller. The national India Dads group (963 members) produces about 1 in 8 of the moms' rate per member per month. The Bengaluru dads' group, the closest geographic match, runs at about 1 in 71. Two different denominators. The same story.
The dads in these groups are already self-selected toward greater engagement. This likely understates the gap in the broader population.
The India Dads group does discuss parenting. Administrative tasks appear: Aadhaar linkage, NPS queries, insurance. Article links get shared. A playschool question or two. But entire categories that define the moms' group, the ones requiring real-time monitoring, anticipatory planning and emotional labor, are completely absent.
These are not trivial categories. Nutritional surveillance is anticipatory cognitive labor. Developmental monitoring is the sustained mental work of knowing what is normal, what is delayed and what requires action. House help coordination is a standing weekly load that keeps care arrangements from collapsing. Their absence from the dads' groups does not mean fathers are uninvolved. It means one parent is carrying the cognitive architecture of the household, and the other is not.
The Bengaluru dads' group had 11 messages across four months. But in those 11 messages, one thread stands out. A father writes that he feels disconnected from his child. He is not the primary caregiver. His child reaches for the mother first, always. He is asking the group how to break in.
The thread is warm. Other fathers respond with genuine empathy. And then, in the suggestions that follow, is the tell.
Look at what just happened. The father asked the group how to be closer to his child. The group's answer was what his wife had to remember to do differently. Even his bonding became her homework: anticipating when he needs space, monitoring whether she is creating it, holding the memory of when to step back and when to stay back. He gets to be present. She gets to be the one keeping track of his presence.
The mother must now anticipate a need (her husband's disconnection), identify the solution (a deliberate step back on her part), decide when to implement it and monitor whether the experiment is working. All four stages of the same framework, running in the background of what looks, from the outside, like a dad finally bonding with his kid.
"He wanted to be closer. The community supported him. And the path forward ran, once again, through her."
This is not anyone's fault. It maps to what researchers call maternal gatekeeping¹: the pattern where fathers' involvement gets mediated by whether mothers create and hold open the space for it. The thread just makes it unusually visible. Her new task has no WhatsApp group. Nobody asks her at midnight whether it's working.
The moms' group is not an anomaly. It is the norm, in urban India, and almost certainly well beyond it. An 18-hour coordination system, running on maternal labor, subsidized by the 80% who read but do not post, operating without compensation, recognition or a replacement when the primary contributors burn out.
The dads' groups are not the problem. Most fathers in them are trying. The man who posted about languishing is not a villain. He is the product of a system that never required him to carry the cognitive architecture of the household, and so he never learned how. The group's solution (have your wife create space) is well-intentioned, practically accurate and structurally devastating.
The school admission cycle, the vaccine-tracker app, the pediatrician's number, the maid's leave calendar, the milk-delivery WhatsApp, the building's society group: all of it lives in one phone, in one woman's notifications, with no escalation path.
A WhatsApp group is not the right unit of analysis for GDP³. But it is a very precise unit of analysis for where, exactly, the invisible labor of a household actually lives. This is where it lives. It never stops. It never sleeps.
¹ Fagan & Barnett 2003; Allen & Hawkins 1999.
² Allison Daminger, "The Cognitive Dimension of Household Labor," American Sociological Review, 2019.
³ For the foundational economic critique of why this work doesn't show up in national accounts, see Marilyn Waring, If Women Counted: A New Feminist Economics, 1988. India's first national Time Use Survey (NSO/MoSPI, 2019) documented a substantial gender gap in time spent on unpaid domestic and caregiving work.
This piece is built on three WhatsApp groups: one Bengaluru moms' group and two dads' groups (one Bengaluru-specific, one national). All three were accessed through members of the respective groups. No one is named. No message is quoted verbatim; all reconstructed threads are composites. Only aggregate patterns are reported. Counts are keyword-based estimates; some messages touch multiple categories.