When Trying to Do Everything Backfires

Autistic people differ, but I have repeatedly seen one pattern in myself and others: when something matters deeply, we can keep developing it until the initiative attempts to contain the whole ideal situation we have imagined.

Every addition may appear reasonable. Information needs a community. The community needs events. Events need publicity. Publicity needs advocacy. Advocacy needs an organisation, funding, services and political influence. Each step follows logically from the previous one, yet the completed system may no longer make clear what it is primarily for.

Every addition also consumes time, money, attention and future options. An unclear initiative is difficult to test, so it can continue consuming resources without producing clear evidence that it works. By the time failure becomes undeniable, the creator may have lost not only the project but also savings, career development and the ability to make a better second attempt.

The danger is not caring too much. The danger is allowing everything that ought to exist to become the specification for what we must build now. I discuss execution more generally in Pragmatic Planning. This article addresses a narrower problem: how to decide what an initiative fundamentally is, limit what we risk on it and preserve a route to recovery if its first form fails.

Separate the purpose from the machinery

An initiative has four different layers:

Purpose: Why do I care?
Primary outcome: What one specific change am I mainly trying to create, and for whom?
Vehicle: What am I using to produce that change: a website, book, service, cafe, group, event or organisation?
Supporting features: What helps the vehicle deliver the primary outcome?

These layers easily collapse into one another. A website, community or organisation begins as a means, then preserving and expanding it becomes the work. We start measuring whether the machinery is busy rather than whether the intended person received the intended benefit.

“I want to run a cafe”, “I want to build a community” and “I want to become an advocate” describe vehicles or roles. They do not yet say what will be different for another person because the initiative exists.

For a defined person or group in a defined situation, this initiative will create one specific benefit they can recognise through one main offer.

That sentence is deliberately restrictive. It does not describe everything I value. It governs this initiative. It may change when evidence changes, but every major feature should have to justify itself against it.

One initiative needs one primary promise

Having one primary promise does not mean doing only one activity. A restaurant may cook, take bookings, train staff, market itself and manage suppliers. These activities form one coherent system because they support the same customer promise.

Clarity requires a hierarchy:

one main person or group being served;
one main benefit being offered;
one intelligible participation and payment model; and
one main result by which the initiative should be judged.

Primary does not mean exclusive. Secondary users and benefits may exist. The primary promise is the tie-breaker when two good aims compete for design, money or attention. Some hybrids genuinely work, but one promise must govern them. A newcomer should not have to reconstruct the hierarchy from rules, explanations and exceptions. Clarity exists in the other person’s mind, not merely in mine.

This is also risk control. When people do not know whether they are buying food, admission, education, membership or moral participation, disappointing results tell us little. A clear initiative exposes itself to an uncomfortable but useful test: will the intended person make the intended commitment for the intended benefit?

Do not let immediate complaints consume the strategic purpose

My purpose was not to create another platform whose main activity was repeating well-known autism problems. I wanted to add something less readily available: future planning, overlooked connections, emerging risks and possibilities that people had not yet recognised.

Yet I repeatedly found discussions pulled towards immediate problems and personal complaints, by autistic and non-autistic partners alike. The concerns could be genuine without being the work this initiative was created to do. When every concern became a new priority, the strategic contribution lost its place.

I call this reactive drift: strategic work gradually displaced by whatever is most immediate, familiar or insistently raised. Nobody needs to take over deliberately. The initiative can remain entirely about autism while losing the particular value it was meant to add.

People facing a problem today can explain why it needs attention. People who will face an avoidable problem years from now are not present to insist that we prepare. Unless we protect that work, the future repeatedly loses to whoever is speaking now. Future planning is not something to begin only after every present problem has disappeared; part of its purpose is to prevent those problems from recurring or becoming harder to solve.

Listening does not require every conversation to redefine the mission. An immediate problem may need direct assistance, referral or action through an existing organisation. A recurring pattern may reveal a strategic question. These are different responses, not automatic new commitments for the same team.

Ask not only, “Is this a real problem?” but also, “What useful contribution can this initiative make that is not already being made?” The answer may be better implementation of an established solution, a new connection or preparation for an emerging change. Novelty alone does not establish usefulness, and evidence can justify changing direction.

Keep support conversations separate from project decisions, reserve time for strategic work, and require a proposed change of scope to identify what it would displace. Turn complaints into questions that can be investigated without promising to take over every case. Protect the purpose, not a favourite untested idea. Understand the route to implementation as well as the idea itself.

A request for more activity is also a request for somebody’s preparation, judgement and follow-up. Criticism can be valid without the critic volunteering, but a desired expansion does not commit another person’s labour or make a retired founder the fallback. When Inclusion Quietly Excludes examines how to preserve useful criticism while keeping the work within an explicitly accepted scope.

When a cafe is also an attraction

A public Singapore example illustrates this tension. In 2021, The Straits Times profiled Shinjuku Kissa Tetsudo, a cafe created from its founder’s love of trains. He was publicly reported as having Asperger’s syndrome. The concept combined Japanese coffee and desserts with an interactive model-railway experience. A detailed contemporary review described ticket-style seat reservations, fixed seating periods and separate charges for some seats, track use and train rental. Another rail publication said that the pricing structure could initially appear daunting, while a food review praised the ambience, coffee and desserts.

In April 2022, AsiaOne reported that the cafe would close on 18 May 2022, seven months after its official opening. The founder said the business was struggling to bring in revenue because some people viewed the main display from outside without entering. The report also described a privacy screen and a later S$10 exhibition fee.

This example offers a useful question: was the customer primarily buying food in a railway setting, or admission to a railway attraction that also served food? Several models could have been coherent:

Cafe first: Customers mainly buy food and drinks. Ordinary seating is part of that purchase, while the railway display provides atmosphere. Any premium railway activity is clearly optional and secondary.
Attraction first: Visitors buy a timed ticket for a defined model-railway experience. Refreshments support the visit but are not the main transaction.
Bundled experience: One clearly stated price or minimum spend includes a defined period, seat, refreshment and railway experience.

A hybrid can work, but one coherent model must govern it. The problem arises when two attractive ideas create competing expectations, separate payment decisions and different assumptions about what the customer will receive.

Count the full opportunity cost

I see a 24-year-old with very limited formal qualifications and work experience committing most of his savings into a new, fixed-site business that closed within seven months. Betting heavily before confirming sustained demand is a concentrated and difficult-to-reverse risk. The more severe the consequences of being wrong, the stronger the evidence should be before taking that risk. An initiative should earn the right to consume more of our future by producing evidence from real behaviour.

Money already spent is visible. Opportunity cost is often invisible because it consists of the strongest realistic path we did not take. The founder’s capital might have supported further education, housing, diversified investments or a smaller train-related initiative.

Do not compare the initiative with doing nothing. Compare it with the best credible use of the same years, money, attention and reputation.

Direct resources: Cash, debt, unpaid labour, equipment, rent and continuing operating costs.
Career capital: Qualifications, technical depth, increasingly difficult assignments, references, seniority and professional relationships that might otherwise have accumulated.
Financial compounding: Income, CPF savings and investment capital that might have grown over time, while recognising that returns are never guaranteed.
Foreclosed options: Study, housing, health needs, caregiving or a different venture that becomes harder after resources are depleted.
Recovery burden: The additional time needed to regain income, rebuild current skills, repay obligations and become ready for another attempt.

My own opportunity cost was not merely the time spent on autism work. I lost continuity in an IT career that I could have deliberately built for about two decades. During that period, I might have accumulated formal qualifications, deeper technical experience, harder projects, professional references, industry relationships, stronger earnings and savings with more time to compound.

Decide whether the risk is justified

The relevant question is not merely, “Could this initiative succeed?” Almost anything could succeed under favourable conditions. We must also ask what happens if it fails completely, takes three times longer than expected or remains only marginally successful. Use a risk gate before making an irreversible commitment:

State the one governing outcome. Name the person, the change and the main offer. If this cannot be expressed clearly, the initiative is not ready to consume major resources.
Identify the assumption most capable of killing the model. For a cafe, this may be whether enough strangers will complete the intended purchase at a sustainable price. For a community, it may be whether suitable members return and contribute without constant pressure from the founder.
Compare three paths. Write down the likely five- and ten-year consequences of a full launch, a small pilot and not launching. Include employment, education, savings, health and relationships rather than comparing only projected revenue.
Set an affordable loss and recovery floor. Entrepreneurship research calls this the affordable-loss principle: decide what can be lost without destroying the ability to continue. Keep daily living, debts and emergencies outside the risk budget. Singapore’s MoneySense guide uses at least three to six months of expenses as a general emergency-fund rule of thumb.
Pilot the real transaction. Test whether unfamiliar people will make the exact payment, commitment or behavioural change on which the model depends. Compliments, publicity and support for the mission are not substitutes. Enterprise Singapore similarly treats pilots and test-bedding as ways to demonstrate viability and establish a track record before scaling.
Preserve an independent path. Continue building employability, qualifications, income and relationships unless strong evidence justifies reducing them. A worthwhile mission should not require every part of life to depend on one unproven vehicle.
Pre-commit the scale and stop decisions. Define what evidence permits expansion, what losses trigger redesign or closure, who may call a halt and when the review will occur. Do not quietly raise the limit because much has already been spent.

The first version should be small enough that failure teaches more than it destroys.

Recovery belongs in the original plan

Recovery is not an admission that we expect to fail. It prevents failure of one vehicle from becoming failure of the founder’s whole life.

Stop further loss. End avoidable commitments, cancel automatic renewals and resist adding money merely to justify what has already been spent.
Stabilise ordinary life. Restore income, housing, healthcare, sleep, food and essential responsibilities before trying to recover the project’s prestige.
Salvage what remains useful. List equipment, intellectual property, research, relationships, audience trust, data and transferable skills. Sell, reuse or transfer them where appropriate.
Separate the purpose from the failed vehicle. A closed cafe does not make trains worthless. A dissolved group does not make peer support worthless. A failed campaign does not prove that the underlying need is imaginary.
Record what reality disproved. Write down the assumptions, warning signs, customer behaviour and missed stop conditions while the evidence is fresh. Avoid blaming one group when the design gave reasonable people conflicting expectations.
Rebuild the recovery floor. Return to work or study, update skills and replenish essential reserves before starting another high-risk version.
Restart only with a materially different test. Use a smaller vehicle, lower fixed costs or a clearer offer. Recreating the same system to prove that critics were wrong is not recovery.

A useful Singapore contrast, although not autistic-led, comes from Dignity Kitchen. Its founder told CNA that he initially asked workers to wear badges identifying their disabilities because he expected customers to be more understanding. Customers instead stayed away. The purpose remained employment and training, but the badges were removed and revenue improved. This does not prove that founders should endure unlimited losses; it shows that loyalty to the purpose may require abandoning a cherished feature. Do not confuse recovering the mission with recovering the original initiative.

Why some autistic people may need explicit safeguards

The problem may arise from strengths used without stopping rules.

We see dependencies. If one part will not work without another, adding the missing part feels rational. Soon we are building an ecosystem rather than delivering the original benefit.

We can explain every connection. Because the whole structure is clear inside our own minds, we may assume that newcomers see the same hierarchy. They often do not.

We dislike arbitrary incompleteness. Removing a useful feature may feel dishonest, wasteful or uncaring. Yet an initiative is not a complete model of the world. It is a deliberately limited piece of work within it.

Our favourite feature can become sacred. It may carry our identity, special interest, personal history or hope of finally being understood. We then defend it even when it creates a second audience, transaction or purpose.

The ideal future can feel more concrete than the ordinary alternative. We can vividly picture the completed centre, recognised movement or successful business while failing to picture the quiet accumulation of experience, savings and stability on another path.

Criticism can sound like rejection of the purpose. Someone may only be saying, “I cannot tell what I am being offered” or “This loss would be difficult to recover from”, while we hear, “What you care about has no value”. Those are different messages.

None of this means autistic people are unsuited to creating organisations or businesses. The ability to see systems and pursue a neglected idea can be a major advantage. The safeguard is to make the hierarchy, risk budget and stopping rules explicit, then give them authority over our attachment to individual parts.

A practical clarity test

Can an unfamiliar person identify the offer? Show them the page, menu or invitation without explaining it. Ask who it serves, what they receive, what they must pay or do and what happens next.
Does every feature serve the same promise? Label each feature core, support or separate. A feature is not core merely because removing it feels painful.
How many transactions are hidden inside the initiative? Count every fee, booking route, membership, eligibility rule, queue and behavioural expectation. Several different participation logics may indicate several initiatives sharing one name.
What is the strongest realistic alternative? Describe where the same resources could place you in five and ten years. Do not compare the initiative with an empty life.
What evidence permits more commitment? Specify the behaviour that must occur before spending more money, reducing employment or expanding the organisation.
What must remain intact? State the minimum income, emergency reserve, housing security, education, health and relationships that the initiative may not consume.
What ends this version? Define both success and stopping conditions before fixation, sunk cost and public identity make retreat harder.

Once the initiative is operating, assess whether the intended people receive the intended benefit rather than how busy the machinery becomes. I discuss this more broadly in How to assess the quality of autism work.

My favourite feature deserves the strongest challenge, not the weakest one. A trusted person can perform it. An AI tool can also act as an adversarial reader, provided I treat its answer as material to examine rather than as an authority. Replace each bracketed part of the prompt below with your own details:

My initiative’s primary user is [describe them]. Its one governing outcome is [state the change]. Its main offer is [state the offer, price or commitment]. My available resources and responsibilities are [describe them]. The minimum life position I must preserve is [describe it]. Review this proposed feature or commitment: [describe it]. Identify whether it strengthens the same promise, creates a second promise, adds hidden decisions or consumes resources disproportionate to the evidence. State the strongest steel man case against keeping it, the smallest real-world test, the stop conditions and the recovery plan if the test fails.

What changed for me

My own mistake was not a lack of effort or sincerity. It was treating the public form of the work as though it were the purpose.

I wanted to share my path so that other autistic people can choose to follow what I experienced; that can be done easily using my website. Instead, I gradually treated talks, media appearances, books, communities, services, newspaper letters and possible political involvement as prerequisites. Some of this work helped people and much of it taught me. Taken together, however, these activities created a second goal: sustaining my public role and the expanding apparatus around it.

The means displaced the end. I did not fail because every activity was worthless. I failed to distinguish the mission from the machinery, and I allowed the machinery to consume resources that the mission did not require.

Recovery did not return the lost years or restore the career position I might have reached. It meant stopping the loss from expanding, salvaging what remained useful and choosing a smaller vehicle that served the original purpose without consuming the foundations of my life.

Choose work that does not wait for community backing

For autistic people working without dependable caregiver, organisational or other support, my advice is to start with a contribution that can be completed and made available without first getting a community to cooperate.

A website is one example. I can write, revise and publish an article without recruiting a committee, filling a workshop or obtaining an organisation’s endorsement. Readers can use it individually and at different times. Publishing does not guarantee readership, impact or income; it gives me control over completing and offering the work.

This is a different starting point from forming a non-profit or an autism-focused business whose operation depends on continuing participation, referrals, funding or a team that has not yet committed. Before building around those dependencies, I would want evidence that the necessary people were willing and able to do their part.

The distinction is not that autistic people should avoid leadership. It is between cooperation already available and cooperation the plan merely assumes. A supported organisation can do work a website cannot; a lone founder cannot substitute indefinitely for absent support. Choose the form of contribution according to the backing actually available, so that collaboration can expand the work rather than be the condition for anything useful to exist.

Why authorship fits me better than professional practice

Lack of backing was not the only problem. I wanted to help people, but wanting to help did not mean that professional coaching or counselling was the right role for me. I found the conventional frameworks, protocols and expectations stifling. My interest was in questioning assumptions, exploring new possibilities and sharing what I had learned. Even with more support, an ongoing professional practice might still have been a poor fit. The prospect of licensing and further professional regulation made that route even less attractive to me.

This was a question of fit, not a claim that responsibility was unnecessary. I was better suited to developing ideas as a self-help author than organising my contribution around a professional practice. As an author, I can explain an idea, its basis and its limitations, and let readers consider whether it is useful to them. I do not need to turn every insight into a service or every reader into a client.

I also considered the fees I encountered too high for many of the people I wanted to reach. I believed useful self-help should be free or inexpensive. People with substantial needs may have little money to spare, and some may not yet recognise how an unfamiliar form of support could help them. Asking for a substantial financial commitment before they can understand or experience its value creates another barrier.

Charging little for individual support would not, by itself, solve the problem. I would still have to provide the time and carry the responsibilities of each relationship. A service that remained affordable only because I continually absorbed the cost could become unsustainable for me.

Writing offers a different arrangement: the same explanation can remain available to many readers without a separate appointment for each one. Free core material and reasonably priced books can coexist. Access to useful self-help should not depend on being able to purchase expensive individual attention.

Writing also lets me make lessons available without having to persuade particular people to accept them, as I explain in my experience of giving and seeking feedback. Readers can consider the material when they are ready. This does not guarantee that they will use it, and written self-help does not replace every need for individual assistance. It defines a contribution I can sustain without becoming responsible for every unmet need.

Takeaways

The ideal situation will always be larger than one initiative. We do not have to embody the whole solution before making a real contribution.

Decide which one change is yours to make. Choose one main vehicle. Let every feature serve that change or move elsewhere. Test the real transaction before building permanent machinery. Risk only what leaves you able to live, work, learn and recover. Stop when the evidence or the risk budget says to stop.

When Trying to Do Everything Backfires is not how much the initiative contains, how intensely we care or how much we sacrificed. It is whether the people we intended to serve can recognise and receive the benefit we set out to provide, without the attempt unnecessarily destroying our capacity to continue with life afterwards.