There was a time when building a technology company sounded expensive before you had even explained the idea.
Developers.
Designers.
Servers.
Agencies.
Marketing teams.
Months of development before a customer could touch anything.
And, somewhere near the beginning of the conversation:
How much are we raising?
That world has not disappeared.
Some businesses genuinely need significant capital, specialist engineers, infrastructure and large teams.
But something else is happening alongside it.
A founder with a laptop, a collection of AI tools, a no-code platform and enough curiosity can now build and test things that would once have required considerably more money and people.
I find that exciting.
I also think it creates a new danger.
When building becomes easier, it becomes very easy to build things nobody actually needs.
That is why my approach to this new generation of entrepreneurship increasingly comes down to two things:
a cold head and a warm heart.
Cold enough to question the numbers.
Warm enough to understand the people behind them.
Bootstrap no longer necessarily means basic
The word bootstrapped used to carry a certain image.
Small.
Underfunded.
A temporary stage before the "real" investment arrived.
I am not sure that description works anymore.
AI, no-code tools, cloud infrastructure, automated marketing, payment platforms and an enormous ecosystem of software have changed what a small team can accomplish.
The evidence is starting to show it.
Startup teams are becoming leaner.
Carta reported that the median seed-stage team on its platform was only four people in 2025.
Stripe Atlas has found startups staying lean for longer, while more founders are reaching their first paying customers quickly.
That does not mean four people can build everything.
It means headcount is no longer automatically evidence of capability.
For a founder, that changes the question.
Instead of:
Who do we need to hire before we can start?
you can increasingly ask:
What can we prove before we hire?
I prefer that question.
I learned some of this before there was AI
Technology makes the current version faster, but the underlying lesson is not new to me.
Before building digital products, I had already spent years operating hospitality businesses.
Hospitality is a very effective education in entrepreneurship because reality arrives every day.
Customers walk through the door, or they don't.
Food gets wasted.
Staff need to be paid.
Suppliers need answers.
Rent does not care how exciting your concept sounded in the planning stage.
A beautiful idea can still have terrible margins.
A busy restaurant can still lose money.
A promotion can bring people through the door and still fail commercially.
You learn very quickly that enthusiasm and viability are not the same thing.
That experience has stayed with me.
Technology may change how cheaply we can experiment.
It does not remove the need to ask whether the experiment makes sense.
The cold head
By cold head, I don't mean cynical.
I mean disciplined.
The founder needs to be willing to ask questions that are considerably less exciting than:
Wouldn't it be amazing if...?
Questions like:
Who is this actually for?
What problem does it solve?
How are they solving that problem today?
Would anybody pay to change?
How often does the problem occur?
What does acquiring a customer cost?
What happens if our assumptions are wrong?
What is the cheapest way to find out?
And perhaps the most uncomfortable:
Do we need to build this at all?
That last question matters much more now.
AI can produce an interface in minutes.
No-code tools can turn a description into a functioning prototype.
You can generate copy, workflows, images, data structures, automation and product concepts at extraordinary speed.
The danger is mistaking the ability to build something for evidence that it deserves to exist.
The danger is mistaking the ability to build something for evidence that it deserves to exist.
Faster building should mean faster learning
The biggest opportunity I see in this new environment is not simply:
We can launch faster.
It is:
We can be wrong more cheaply.
That sounds negative.
I think it is one of the most positive things that has happened to entrepreneurship.
Imagine spending twelve months and a significant amount of money building a product, only to discover customers do not care.
Now compare that with spending a few weeks producing a useful prototype, putting it in front of real people and discovering the same thing.
The failure still happened.
But the second founder bought the lesson for much less.
That is what a good minimum viable product should do.
It should not prove how much you can build with limited resources.
It should help you discover whether continuing to build is justified.
But then comes the warm heart
This is where I think some technology-first founders get into trouble.
We love optimisation.
Conversion.
Automation.
Funnels.
Retention.
CAC.
LTV.
Activation.
Churn.
All useful.
But behind every one of those metrics is a person doing something.
A conversion is someone deciding to trust you.
Churn is someone deciding you are no longer worth paying for.
A support ticket is somebody whose day has just become slightly more difficult because your product did not behave as expected.
A lead is a person, not a row inside a CRM.
You can become so focused on optimising the system that you stop understanding the human experience inside it.
That is why the founder also needs a warm heart.
Not as a slogan.
As a way of working.
Talk to people before you automate them
This is particularly relevant now because almost everything can be automated.
You can automate outreach.
Onboarding.
Support.
Follow-ups.
Lead qualification.
Surveys.
Reviews.
Content.
Reporting.
Even parts of sales conversations.
But automation works best once you understand what is happening.
If ten customers are asking the same question, perhaps automate the answer.
If you have never spoken to those customers, automating the conversation may simply help you misunderstand them at greater scale.
Early-stage founders sometimes need to do things that would be inefficient in a mature company.
Call the customer personally.
Watch somebody try to use the product.
Read every support message.
Ask why they cancelled.
Ask why they bought.
Notice where they hesitate.
Listen when they describe your product using completely different language from the words on your website.
That information is gold.
And it is very easy to automate yourself away from it.
A warm heart does not mean saying yes to everything
Customer empathy is sometimes misunderstood as customer obedience.
Those are not the same thing.
Customers can tell you what frustrates them.
They can describe what they are trying to accomplish.
They can show you where the experience breaks.
That does not mean every requested feature should be built.
If five customers ask for five completely different things and you build all five, eventually you no longer have a product.
You have a collection of favours.
The warm heart listens.
The cold head decides.
You need both.
Cheap technology can create expensive distraction
There is another strange consequence of today's tools.
Ideas have become cheap.
Very cheap.
Spend an hour with generative AI and you can leave with:
twenty product concepts,
ten possible customer segments,
a brand name,
a logo,
a landing page,
a pricing strategy,
a launch plan,
and enough enthusiasm to start three companies before lunch.
I know the temptation.
When tools make creation enjoyable, starting something new can feel more rewarding than fixing the boring thing that already exists.
Founders have always suffered from shiny-object syndrome.
Now the shiny objects can build themselves.
That makes focus more valuable.
The question is no longer:
Can I make this?
Often, you can.
The better question is:
Is this the best thing I could be spending the next six months learning about?
Do more with less, but don't turn it into a religion
I like lean businesses.
I like finding ways to test before spending heavily.
I like tools that allow small teams to compete with larger ones.
But there is a danger in turning "doing more with less" into another startup vanity metric.
Fewer employees is not automatically better.
Bootstrapped is not morally superior to funded.
Using AI is not automatically innovative.
There comes a point where refusing to hire becomes a bottleneck.
There are businesses where capital creates a genuine advantage.
There are problems that require deep technical expertise.
And there are moments when moving slowly because you refuse to spend money is more expensive than raising or investing properly.
Being lean should be a strategic choice.
Not an identity.
Revenue is a wonderfully clarifying form of feedback
Likes feel good.
Waitlists feel encouraging.
Website traffic is interesting.
Compliments from friends are dangerous.
A paying customer is different.
Money is not the only measure of whether something has value, particularly for community, public or purpose-led projects.
But for a commercial startup, someone paying for what you built is an unusually useful piece of evidence.
They had alternatives.
Including doing nothing.
And they chose yours.
Recent startup data suggests founders are reaching that moment faster. Stripe Atlas reported that 20% of startups formed through its platform in 2025 landed a first paying customer within 30 days, more than double the proportion in 2020.
I find that direction healthy.
Not because every company should monetise immediately.
Because it moves the conversation away from theatre.
The company exists to create value for somebody.
Find out whether it does.
The founder's job may be changing
When I think about what AI and no-code mean for entrepreneurship, I don't think the main story is that founders no longer need technical people.
Great engineers remain enormously valuable.
I think the more interesting change is that founders can personally participate in far more of the early discovery process.
You can prototype.
Research.
Build a landing page.
Test positioning.
Create workflows.
Analyse feedback.
Automate repetitive tasks.
Put a simple version in front of customers.
That means the distance between:
I have an idea
and
Here, try this
has become dramatically shorter.
Which also means founders have fewer excuses for avoiding customers.
Cold head. Warm heart.
The businesses I find most interesting in this new era are not necessarily the ones using the most AI.
They are the ones using new technology to stay close to the problem while removing unnecessary cost around solving it.
They move quickly without becoming careless.
They measure without forgetting what the numbers represent.
They automate without disappearing from the customer relationship.
They know when not to hire.
And they know when being too lean has become a weakness.
They are willing to abandon an idea they love when the evidence says it is wrong.
And willing to continue with one nobody else understands yet when customers tell them it is right.
That requires two different instincts.
A cold head asks:
Does this make sense?
A warm heart asks:
Does this actually help somebody?
If I had to reduce a lot of what I have learned from entrepreneurship, across physical businesses and now digital ones, to two questions, those might be enough.
Technology will keep reducing the cost of building.
AI will keep making small teams more capable.
Another platform will arrive tomorrow that lets us create something today that required a specialist yesterday.
Good.
Use it.
But easier building does not remove the difficult parts of entrepreneurship.
Choosing.
Listening.
Focusing.
Selling.
Taking responsibility.
Knowing when to stop.
Knowing when to continue.
And understanding another person well enough to build something they genuinely want.
Perhaps that is the real opportunity of the bootstrap tech era.
We can spend less energy proving that we are capable of building things.
And more energy making sure we are building things worth having.