However, I wasn't just a buyer. I was a seller as well. I had sold products before right on instagram. As I safeguarded myself against consumerism to the extent I could, I couldn't help but wonder that there would be more people like me. People who are annoyed with being sold to on instagram constantly.

For my own brand that I had started recently, for which we were targeting parents, I spent a lot of time chasing non scalable channels initially. How can I sell to people one on one? Where do my customers have high attention span? Probably at a place where they are not seeing barrage of ads.

High dependency on Meta + Google doesn't help in building long term audience. The recall has become very low. CAC's are rising.

I have first hand been bitten by over dependency on Meta and heavy revenue fluctuations with a small Meta update. That gap creates a very specific opportunity: a marketplace for micro ad spaces - small, high-intent placements that are currently unmonetized, under-monetized, or sold manually. Think: - receipts, bills, packaging inserts - QR posters at checkouts - apartment elevators and lobbies - hyperlocal community boards / RWAs - gyms, cafés, clinics, salons - micro-influencer “spaces” that don’t justify a full sponsorship - niche newsletters / WhatsApp communities - college fests, events, in-venue screens Basically anyone to be able to sell their digital or physical space and brands getting high intent spaces where recall is higher. I talked to 30+ brands and ad real estate owners and have synthesised my insights below. --- ## The core insight: micro inventory exists everywhere - it’s just not structured Micro ad spaces already exist in two forms: 1. Implicit spaces Places where attention naturally lands (checkout moment, delivery moment, waiting rooms), but nobody has productized it. 2. Manual sales spaces Someone is selling it informally (posters, inserts, local screens) - but the process is manual. That’s why this market is still inefficient: - supply is fragmented and unstandardized - buyers can’t discover inventory reliably - measurement is weak - payments and proof-of-run are painful - repeated buying is rare because coordination cost is too high --- ## Why now: three forces are making micro inventory more valuable ### 1) CPM inflation + fatigue on digital Performance ads are crowded. CAC is unstable. Brands are hunting for incremental, less competed channels. ### 2) Offline is becoming trackable again QRs, UPI flows, coupon codes, lightweight attribution, and geo-lift tests make “micro” placements measurable enough to justify repeat spend. ### 3) Consumer attention moved to “moments,” not platforms The highest intent moments aren’t always inside Instagram or Google. They’re: - when you’ve already purchased - when you’re waiting - when you’re in a routine location - when you’re in a community context --- ## The real market structure: it’s multi-sided and incentive-misaligned It's a structure with 4 stakeholders. ### 1) Inventory owners (supply) Gyms, cafés, clinics, RWAs, delivery brands, printers, newsletters, communities, event organizers. What they want: easy money, zero headache, predictable payouts.

What they fear: brand risk, operational burden, “who will execute?” ### 2) Advertisers (demand) D2C brands, local businesses, coaching centers, fintech, food brands, health/wellness, edtech — anyone who wants discovery or trials. **What they want:** outcomes + clarity. **What they fear:** wasting money, fake reporting, “unscalable channel”. ### 3) Agencies / freelancers They’ll either amplify you or block you. **What they want:** margin + control + reliability. **What they fear:** platforms that kill their economics or add churn. ### 4) Consumers They’re the reason micro works - but also the reason it can backfire. **What they want:** relevance, not spam. Micro inventory must feel like “helpful discovery,” not “cheap clutter.” --- ## The product truth: Supply side is harder to build If you build this like a simple two-sided marketplace, you’ll stall. The real bottleneck is meaningful incentives for people in exchange of their ad spaces. The reason why these spaces have higher recall is because the owner has maintained a high trust environment. If you run ads- that gets contaminated. The better the ad space- the more restrictions to run ads there. This is why people have to come up with hacks to post in high intent reddit communities. --- ## What “micro ad spaces” actually sell: conversions The best micro inventory has one of these properties: ### High intent Checkout counters, delivery boxes, clinic waiting rooms. ### High repetition Gyms, elevators, cafeterias, commute routes. ### High trust Communities, newsletters, local groups. ### High targeting A very specific demographic that’s hard to isolate on mainstream platforms. This is why micro can outperform: Not because it reaches more people — because it reaches the _right_ people in the _right moment_. --- ## The wedge: don’t start with “all micro inventory.” Pick one repeatable lane. All these spaces are very different structurally. A strong wedge looks like: - **one inventory type** (e.g., packaging inserts, clinic waiting rooms, elevator screens) - **one buyer segment** (e.g., D2C wellness, local services, coaching) - **one measurable action** (QR scan → WhatsApp → coupon/lead) Once you own reliability in a lane, expansion is natural. --- ## Pricing + trust: the only two things that matter early ### Pricing needs to feel stupidly simple Most micro inventory fails because pricing feels arbitrary. Early on, the best structure is: - fixed packages (X locations, Y days, Z impressions proxy) - transparent add-ons (printing, installation, creative help) - “guarantees” framed as operational SLAs (proof-of-run, makegoods) ### Trust needs to be built Trust in micro ads comes from: - verified inventory - standardized proof (geo-tagged photos, timestamps, QR logs) - dispute handling - repeatability (same workflow every time) If buyers don’t trust proof-of-run, they won’t repeat. If suppliers don’t trust payouts, they won’t stay. --- ## What success looks like (and what won’t work) ### This will work if you become: - a **distribution primitive** for a category (e.g., “the easiest way for D2C to run insert campaigns”) - the **operating system** for micro inventory owners to monetize repeatedly - a **reliable execution layer** that agencies can plug into ### This won’t work if you become: - a directory of random ad spaces - a marketplace that pushes execution onto sellers - a platform with weak measurement and unclear outcomes --- ## The big bet Micro inventory is one of those markets that looks small from the outside because it’s fragmented. But fragmentation is exactly why it’s interesting. If you can: - standardize supply, - make buying repeatable, - and wrap it with enough measurement to justify re-spend, …you unlock a new category of ad spend that currently leaks into: WhatsApp deals, one-off sponsorships, offline posters, and “random experiments” that never scale.

Why does “work management” still feel painful even after Asana, Jira, Notion, Slack, and a dozen other tools?

Oyester started as an attempt to answer that. We explored an AI-native work management system that could reduce manual task upkeep, improve visibility, and remove the constant overhead of coordination.

We’re not building Oyester anymore — but the research was valuable enough that it’s worth documenting. This is a synthesis of the patterns that showed up repeatedly across user calls, not a list of feature ideas.

The core problem wasn’t task management. It was “execution friction.”

Most teams we spoke to didn’t say they lacked tools.

They said something closer to:

• “We have tools, but no one keeps them updated.”

• “We don’t know what’s actually happening until it’s too late.”

• “Half of management is chasing updates.”

• “The cost of coordination keeps increasing as the team grows.”

So the pain wasn’t creating tasks.

It was everything around tasks:

• capturing work from messy channels

• translating it into ownership + deadlines

• ensuring follow-through

• preventing work from disappearing

• maintaining visibility without constant manual updates

We started calling this execution friction — the invisible tax teams pay to keep work moving.

Insight #1: Work entered the system from everywhere — and died in the gaps

One recurring theme: teams didn’t have a “source of truth.”

Work showed up through:

• WhatsApp threads

• Slack pings

• meeting notes

• random voice calls

• hallway conversations (or their remote equivalents)

• emails, forwards, screenshots

Then it either:

• never got logged,

• got logged but never updated,

• or got lost between tools (“it’s in Slack”, “it’s in Notion”, “it’s in someone’s head”).

The result was a phenomenon we heard in many forms:

the “black hole” problem — tasks disappear between intent and execution.

Insight #2: Managers weren’t asking for more data — they were asking for fewer surprises

“Visibility” is often described like a reporting problem. In reality, it surfaced as an emotional one.

Managers repeatedly described:

• anxiety about not knowing what’s going on

• fear of missing something critical

• frustration at finding out late

• the need to “check in” constantly just to feel safe

This wasn’t because they wanted to micromanage.

It was because they were operating with incomplete signals.

We started framing it as manager blindness:
when the system doesn’t naturally surface risk until it becomes a fire.

Insight #3: Status updates were not “work.” But they consumed an absurd amount of work.

Almost every team had a form of “update ritual”:

• daily standups

• weekly reviews

• Friday status docs

• sprint planning

• manager 1:1s

• client update threads

People didn’t hate updates because they were lazy.

They hated updates because:

• writing them felt like redoing work

• tools demanded structured input that didn’t match how work happened

• maintaining tools became a second job

• and the “admin layer” scaled faster than the team did

This was the admin tax — coordination overhead that compounds with growth.

Insight #4: Tool fatigue was real — and it wasn’t about too many tools, it was about too many obligations

Many teams were already using:

• Slack/WhatsApp for communication

• Asana/Jira/ClickUp for tracking

• Notion/Confluence for documentation

• Google Docs/Sheets for planning

• Calendar + email for scheduling

The complaint wasn’t “too many tools exist.”

It was:

• “Every tool demands upkeep.”

• “Every tool creates a new place to check.”

• “We spend time maintaining systems instead of doing the work.”

So adoption failure was predictable:

• leaders tried to implement process,

• the team complied for a week,

• upkeep dropped,

• the tool became stale,

• and everyone returned to chat + memory.

This wasn’t a product problem. It was a behavior + incentives problem.

Insight #5: The best teams didn’t have better tools — they had better “default flows”

What separated high-performing teams wasn’t “what tool they used.”

It was whether the system made the right thing the easiest thing.

Examples of “better defaults” we observed:

• clear ownership norms (“every task has an owner, always”)

• explicit priority visibility (“what matters this week is obvious”)

• lightweight check-in loops that didn’t feel performative

• a culture where task capture happened naturally (not as an extra step)

So the real wedge wasn’t building a better Asana.

It was reducing friction so the system stayed true without effort.

Insight #6: AI skepticism wasn’t about capability — it was about trust and control

When we brought up automation, responses split into two camps:

1. Excited: “If it can remove admin, I’m in.”

2. Suspicious: “If it creates wrong tasks or misses context, it’s worse than useless.”

The skepticism was less about “AI can’t do it.”

It was about:

• who is accountable when AI is wrong

• whether the AI understands context, nuance, and constraints

• whether it creates work instead of reducing it

Teams didn’t want AI making decisions in the dark.

They wanted AI that:

• captures context,

• suggests actions,

• and stays auditable.

What this changed in how we thought about the product

Oyester began as “AI-native work management.”

Over time, the research pulled us toward a more specific framing:

teams didn’t need a tool to track work — they needed a system that prevented work from slipping.

The value wasn’t a prettier task list.

The value was:

• catching work that never got captured

• surfacing risk early

• reducing coordination load

• and making execution behavior automatic by default

The bigger takeaway

Work management is one of those spaces where feature checklists are misleading.

Most tools already do the “management” part.

The market gap lived in the unglamorous layer between:
“we agreed to do this” and “it actually got done.”

That layer is where:

• coordination breaks,

• accountability becomes personal,

• and time gets burned silently.

Even though we stopped building Oyester, the research convinced us of one enduring truth:

execution isn’t a planning problem — it’s a systems problem. —

New posts coming regularly. Subscribe to the RSS feed or check back here.

Did school realise at that point how these things compound? It's not a co incidence that it were the same kids being given these opportunities in 4th and 12th grade. Mostly because they were better in written exams in 4th grade. How can someone be good at written exams in 4th grade? For me, I was never good at it. Including now.

School was a great. I feel fortunate to have met people that I did. You cannot trash on 14 years of your life. That's more than half my life. So I really don't know if it was or was not great- but I will always have to believe that it was.

I started exploring education space out of strong personal interest. I was very clear I wanted to build a new age school. Although my school made sure (JK no blames) that I wasn't permitted to dream that high.

I went back to basics and tried to understand the ecosystem the way it actually functions day-to-day — across parents, teachers, schools, and students. I talked to 50+ stakeholders across India. What follows is a crisp synthesis of the top-level insights I kept hearing.

Approach: talk to every stakeholder in the loop

It's a multi-stakeholder system where the person paying isn't the person using the services. The results are long term- in terms of job and what the person would end up becoming. Yes- no matter what the parents say a lot of schooling is still correlated with the results it gives in 12th grade.

The ecosystem, in one line: everyone cares but in different directions

A pattern that showed up repeatedly was this:

• Parents want outcomes, safety, and social validation.
• Teachers want manageable classrooms and to focus on the "top slice."
• Schools want measurable progress + reputation protection.
• Kids want autonomy, meaning, and relief from pressure - but also get swayed by trends.

This creates a system where everyone is "trying"

1) Parents: more progressive - but only until grades drop

One of the most interesting things: parents are open to kids exploring interests and passions. But this openness has a boundary.

I've always believed that even though my parents had expectations, they rarely said it out loud. If you would ask them- they would have said that me and my brother were open to pick our careers. Which in all honesty- we were. So were all my cousins. What morning discussions would include at all these houses were even though stories of what other kids have achieved.

What parents say vs what parents do

• Parents like the idea of passion and holistic growth.
• Parents also care deeply about status - and the moment academics drop, they start panicking
• Most are willing to spend on both academics and co-curriculars, but co-curricular tends to become "second priority" the minute academics feel threatened.
• In 2 very distinct cases they were approving of the child's career- where they have given up on child's academics or where the child is great at academics and hence has earnt the right to pick

Parents are deeply involved emotionally but not always proactive

• They listen to their kids.
• They are protective about parenting style and school choices.
• But many don't intervene before things go wrong (the child lacking in studies) - they react after the alarm.

In practice, this means parent behavior is often late-stage: action happens when marks fall and not when motivation falls.

2) Teachers: group classes are built for the top 5%

A teacher insight I heard repeatedly (sometimes explicitly, sometimes implied) is that the classroom group setting does not work for most students.

The classroom reality

• Group classes disproportionately serve the top performers.
• The remaining majority starts lagging behind.
• Once a student falls behind, recovery is hard without extra support - and that's where tuition fills the gap.

The vicious loop: tuition + disengagement

A painful loop shows up:

• Students go to tuition because school isn't enough.
• Once tuition becomes the "real learning," school feels pointless.
• When students disengage, teachers get less rewarded for effort.
• Teachers then focus even more on the few students who respond.

It's an incentive + bandwidth problem.

Teachers and AI: cautious acceptance, real fear

Another tension: teachers don't "hate tech" - but many fear AI reduces thinking.

• There's skepticism that AI will "do the work" for children.
• Some schools/teachers are okay with AI if it's positioned as scaffolding
• The bigger issue is how to integrate without breaking learning integrity.

3) Schools: "what gets measured gets managed" (and that's both good and bad)

Schools sit at the center of the system - and they operate on hard constraints:

• reputation
• academic outcomes
• parent expectations
• compliance and operational feasibility

Parents are "inactive" - until they're not

Schools often perceive parents as inactive in the learning process (day-to-day), and believe that parents are not playing their role

Schools label kids as "fragile"

A phrase that came up often (in different forms) was that kids today are seen as:

• more emotionally sensitive
• less resilient
• more psychologically demanding

And whether or not that's fair, it changes what schools are being forced to build: counselling, emotional support, career guidance - all of which are now becoming mainstream rather than "nice to have."

Marks still dominate the system

Even when schools talk about holistic development, the operational reality is:

• "numbers" are still the cleanest way to prove progress
• hard data is what parents respond to
• schools need measurable structures (submissions, engagement scores, deadlines, etc.)

So the system still leans heavily toward what can be tracked - even if it's not the full picture.

A notable emerging theme: policy-driven FOMO

Hands-on learning, mentorship models, "try things early" initiatives - these are increasingly being pulled into schools, sometimes because:

• parents want them
• colleges signal they matter
• other schools are doing it (and competition forces adoption)

4) Kids: more self-directed, more distracted, more emotionally loaded

Kids were the most nuanced stakeholder - because their behavior looks contradictory.

More career-aware (but trend-driven)

Kids today are:

• more aware of careers early
• more exposed to non-traditional success stories
• heavily influenced by social media narratives

This creates two realities at once:

• some kids genuinely want to explore and build
• others adopt "I don't need to study" as a coping strategy - even when they know it's not fully true

Self-learning is real - but not always academic

Kids are learning online constantly:

• skills, hobbies, knowledge
• but often outside the academic structure

So learning isn't dead. It's just moving to environments that feel faster, more rewarding, and less judgmental.

Interest hinges on teachers (more than we expected)

This came through strongly:

• a good teacher can make a subject lovable
• a bad teacher can make a strong student average
• interest is not fixed - it's highly context-dependent

Long school routines + emotional needs are increasing friction

Long school days + pressure + structured routines + constant evaluation means kids often carry:

• stress
• identity anxiety ("am I good enough?")
• emotional needs that schools/parents weren't trained to handle historically

What this means for anyone building in Indian education

Based on what I heard, a few opportunity directions feel especially real:

1) Products that align stakeholders

If the parent/school/teacher loop isn't addressed, adoption becomes fragile.

2) Tools that support teachers at scale (without adding admin work)

Teachers don't need another dashboard. They need systems that reduce load and help them intervene earlier.

3) Measurable holistic development (beyond marks) - but still "trackable"

Schools want holistic frameworks, but they need them in measurable formats that parents can trust.

4) Career guidance and counselling are no longer niche

This is becoming a core layer of schooling - especially as emotional load rises and career paths diversify.

5) Structured "hands-on" learning that isn't performative

Mentorship, projects, try-before-you-choose models - these are moving mainstream, but need better scaffolding and credible measurement.

Closing thought

Doing this research was a good reminder that education in India isn't a "market." It's a social system. People's hopes, anxieties, identities, and status are all entangled with learning.

And that's why the bar is higher here:
If you want to build something meaningful in this space, you can't just build for engagement.