Planning in the fog: what it feels like to build a company in 2026

/ Office Space Guides / By
Team workspace behind glass partition with lounge seating

By Tisha, Facilities Manager, eCradle Coworking Spaces

It is a little after eleven at night, and the founder in the cabin at the end of the floor has three versions of the same spreadsheet open.

In the first, the company is 22 people by next September. In the second, it is 14, because two of the roles she budgeted for can probably be handled by tools her team is already testing. In the third, it is 35, because a large customer in the US might sign, and if they do, everything speeds up at once.

She has to pick one by Friday. Her investors want a plan. Her team wants to know whether to start interviewing. And honestly, she does not know which version is true. Nobody does.

We see some version of this scene almost every week. We are not founders, and we are not going to pretend we have the answers. But we sit close enough to watch the questions pile up. This piece is an attempt to describe them honestly, because recognising the problem clearly is the first part of handling it.

The clock got shorter

For most of the last decade, the deal between founders and investors was roughly this: grow fast, and profitability can come later. That deal has quietly been rewritten.

Indian tech startups raised about $10.5 billion in 2025, roughly 17% less than the year before, according to Tracxn. That is still a lot of money, and India still ranked third in the world. But the money now comes with different questions. Late-stage rounds have shrunk the most. More than a third of Indian startups chose to extend their runway and work towards profitability in 2025 instead of raising again, according to Inc42. One investor summed up 2026 bluntly: profitability and unit economics are no longer goals to work towards. They are what it takes to get capital at all.

What this means for a founder is not just “spend less.” It means every decision is now judged on a shorter clock. A hire is no longer an investment you explain in two years. It is a cost that shows up in this quarter’s numbers. An office is no longer just where the team sits. It is a line in a board deck that someone will ask about.

At the same time, nobody wants to be the founder who played it safe and missed the moment. In AI especially, early-stage valuations are running hot. Many investors openly describe Indian AI valuations as frothy, and the Reserve Bank of India has warned about the risk. So founders are pushed in two directions at once: move fast enough to catch the wave, and stay careful enough to survive if it breaks. There is no comfortable middle.

The world outside will not sit still

Even a founder who gets everything inside the company right is exposed to a world that has become hard to read.

In 2025 alone, the United States raised tariffs on many Indian goods to as much as 50%. It also announced a $100,000 fee on new H-1B visa applications, a programme that Indians accounted for 71% of in 2024. Whatever happens to these policies next, they made one thing clear. Assumptions that held for twenty years, about where clients are, how talent moves and what trade costs, can change in a single announcement.

Many Bengaluru startups sell to customers abroad and pay for cloud and AI models in dollars while earning in rupees. For them, a currency swing, a policy change in Washington or a budget freeze at a US client is not background news. It decides whether next quarter’s hiring plan is real.

The hard part is not any single shock. It is that founders now have to plan as though the next one is coming, without knowing what it will be.

AI: tailwind, headwind, or both?

Then there is the question that sits underneath every other question this year: what does AI actually do to my business?

The honest answer, for most founders, is that they do not yet know. And the evidence does not settle it.

On one side, recent research suggests that companies building AI into their products run with noticeably smaller teams than similar companies without it, at similar valuations. Founders hear stories every week about tiny teams doing what used to take fifty people.

On the other side, analysis of Indian startups points the other way: AI-era companies have raised more money, but so far they don’t show better revenue per employee than the generation before them. An MIT report in 2025 found that around 95% of corporate generative-AI pilots showed no measurable effect on profit. And a careful experiment by the research group METR found that experienced developers using AI coding tools took about 19% longer on real tasks. Even after the study, they still believed the tools had made them about 20% faster.

That last finding deserves a second look. It is not saying AI is useless. Tools are improving fast, and the researchers were careful to say the result should not be stretched too far. What it shows is something more uncomfortable: our sense of whether AI is helping can be confidently wrong. For a founder, that is the real problem. It is not about choosing between “AI will save us” and “AI is hype.” It is about how to make decisions when even your own experience might mislead you.

And it is not only about your own team. The same tools that might make your company leaner might let a competitor with five people undercut you, or let your customer build in-house what they used to buy from you. Founders tell us the fear is not being replaced by AI. It is being surprised by it, finding out six months too late that the ground moved.

How do you stay on top of something that changes every few weeks? Most founders we know have settled on small, practical habits rather than grand strategies: one person on the team who owns “what changed this month,” a few real experiments on real work instead of demos, and a rule to measure results rather than trust impressions. None of it is a framework. All of it takes time, which is the one thing founders have the least of.

The hiring question nobody can answer

All of this comes to a head in one very practical decision: who do I hire?

A few years ago the answer was simple, if expensive. You needed more engineers, so you hired more engineers. Today a founder building the same product might reasonably consider very different teams:

  • Ten developers, the traditional way.
  • One sharp analyst with some technical background, using AI tools to do work that used to need several people.
  • A senior project manager or tech lead, with three or four developers who are trained to work closely with AI.
  • A small core team, plus contractors and specialists brought in for a few months at a time.

Each of these could be right. Each could be a costly mistake. And the market is sending mixed signals. Staffing firm Xpheno reported entry-level tech openings down sharply year on year, while other data showed fresher hiring rising in some months, driven mostly by AI-related roles and internships. The skills employers ask for are shifting from degrees towards proof of work and adaptability. That makes sense, but it is hard to plan a hiring pipeline around.

The deeper difficulty is that a hiring decision used to be reversible only at great cost. Get it wrong, and you either carry a team you do not need or scramble for one you do. In a year when even the shape of the team is uncertain, every fixed commitment built around a headcount number becomes a bet on a forecast that might not survive the year.

Forecasting when the inputs keep changing

Put these together and you get the founder’s real problem in 2026. It is not any one of these pressures. It is that they all feed into the same few numbers: revenue, headcount and burn. And every one of those inputs is moving.

Growth projections used to be built on reasonably stable assumptions about hiring, cost per employee and the market. Now the cost of a unit of work might halve because of a new tool, or a key customer’s budget might vanish because of a tariff. The person you planned to hire in June might not be the person you need by then.

Founders are not short of advice about this. What they are short of is room for error: time, cash and attention. The founders who seem calmest are not the ones with the best forecasts. They are the ones who have reduced the number of decisions that lock them in.

One part you can take off the table

We are not going to pretend an office solves any of this. It does not tell you whether to hire the analyst or the ten developers, or whether the US deal will sign.

But it is one of the few parts of the problem that does not have to be uncertain, and that is the part we can speak to.

When the team size is a guess, the office should not be a bet on that guess. You do not need a three-to-five year lease with a large deposit, a fit-out that eats months and lakhs of rupees, and a building you are responsible for. You can start with the seats you need today and grow into the space as hires land.

When the investors ask about burn, the office can be one number: a fixed cost per seat that covers rent, electricity, water, power backup, internet, housekeeping and coffee. No surprise invoices, no generator diesel to track, no capital spent on furniture that could have gone into product.

And when the founder is already stretched across strategy, hiring, fundraising and a technology that changes every month, the office can simply work. The AC, the UPS, the cleaning and the 9 pm power cut are someone else’s job. The space is open 24×7, weekends included, because the work does not keep office hours either.

It is a small piece of a large, uncertain picture. But in a year when almost nothing can be planned with confidence, taking even one variable off the table is worth something.

The founder with three spreadsheets will still have to choose one on Friday. At least she will not need a fourth one for the office.

eCradle Coworking Spaces is in HSR Layout Sector 6, Bengaluru, next to Reliance Digital. It offers a ready private office for up to 30 people and dedicated desks, all-inclusive and with no lock-in. ecradle.in

Sources