Why You Should Aim for Zero Support for Your Products

A company announces "24/7 support" as a feature. The room claps.

No one asks the obvious question: why does the product need 24/7 support in the first place?

Jeff Bezos said it best: "The best customer service is if the customer doesn't need to call you, doesn't need to talk to you. It just works." Support is not a feature. It is a symptom.

The goal is not world class support. The goal is a product so simple and obvious that support is rarely needed.

Every Support Ticket Is a Bug Report in Disguise

Users do not file tickets for fun. They file them because something in the product failed them. An unclear label. A hidden button. A confusing flow. A step that assumed knowledge they did not have.

Every ticket is a user saying, out loud, "I got stuck here."

That is the most honest feedback channel you will ever get. Better than surveys. Better than NPS. Better than reviews. It is real people describing the exact moment your product failed them.

Most companies read tickets to close them. The better move is to read tickets to kill them at the source. Every repeating ticket is a UX decision you have not made yet.

I covered this in Stop Making Your Users Think: every moment of confusion is a failure. Tickets are the receipts.

The Hidden Cost of "Good Support"

Support is not free. And it is not cheap.

A single support ticket costs anywhere from $2 to $60 depending on the industry. SaaS tickets average $25 to $35. Banks and fintechs hit $50+ on complex cases.

The channel matters enormously. Self-service contacts cost $1.84 versus $13.50 for agent-assisted, per Gartner benchmarks. That is a 7.3x multiplier every time a user cannot figure something out on their own.

SaaS companies spend around 8% of ARR on support and success. For a $50M ARR company, that is $4M a year, mostly paying humans to explain things the product should have explained itself.

A support team is a tax on bad UX. And the tax compounds.

Why Teams Over-Invest in Support Instead of Simplicity

Hiring is easier than simplifying.

When a product is confusing, the default move is to add people, not remove friction. Add a support team. Add a CSM. Add onboarding calls. Add training videos. Every layer feels productive. None of them fix the product.

Simplification is hard because it requires throwing things away. Features. Options. Clever ideas someone fought for in a meeting. Steve Jobs put it plainly: "Simple can be harder than complex. You have to work hard to get your thinking clean to make it simple."

Builders also underestimate how confusing their own product is. I wrote about this in the curse of knowledge. Once you know something, you lose the ability to imagine not knowing it. A support team is often there to translate between the team's mental model and the user's.

That translation is not free. And it is not a moat. It is a cost center you hired to avoid a harder conversation about your product.

What Zero-Support Products Do Differently

The products that barely need support share a few traits.

The first interaction teaches the product. Google Search has no tutorial. WhatsApp has no onboarding video. Stripe Checkout does not need a walkthrough. Users open them and get it. The product is the manual.

Conventions over cleverness. Users arrive with expectations shaped by every other product they have used. Zero-support products meet those expectations instead of inventing new ones. Every deviation from a convention is a future support ticket.

Errors are actionable, not cryptic. A good error message tells the user what happened, why it happened, and what to do next. A bad one says "Something went wrong" and sends them to a help desk.

Scope is visible. Users can tell what the product does and does not do without trying. Unclear boundaries create tickets. "Can it do X?" is a support question that should have been answered by the interface.

Status is always obvious. Is it working? Is it saved? Is it sent? Every unanswered question becomes a ticket. Zero-support products make state visible so users never have to ask.

The WhatsApp Lesson

By 2016, WhatsApp had over a billion users. It was served by only 50 engineers. Almost none of its small team worked in customer support.

Their internal mantra was "just enough engineering". Pick the simplest approach. Use the fewest components. Ship only the features the core mission needs. Send messages. That was it. No feeds. No stickers marketplaces. No onboarding flows. You opened the app and used it.

WhatsApp did not scale support. They scaled not needing support.

Compare that to enterprise SaaS, where implementation consultants, customer success managers, and dedicated account teams ship as part of the product. The service layer exists because the product cannot stand on its own. The bigger the support org, the louder the signal about the product underneath it.

That is not always bad. But it is always worth asking: is the service making up for something the product should be doing?

Signals Your Product Needs Less Support, Not More Support Agents

Some patterns give it away.

• The same five questions show up in tickets every week. That is not a support issue. That is five UX bugs you have not fixed.
• Onboarding calls are "mandatory." If a user cannot start without a human, the first screen is wrong.
• A Customer Success Manager is standard. Sometimes necessary. Often a polite admission that users cannot succeed on their own.
• Docs are longer than the product is deep. If the documentation is doing the heavy lifting, the interface is not.
• Support satisfaction is high but support volume is rising. You are getting better at handling confusion instead of removing it.

The tell is simple: are tickets going down over time, or up?

If support is growing faster than the product is improving, the support team is covering for the product. That is expensive, and it never stops being expensive.

How to Drive Toward Zero Support

Five moves that actually work.

Read tickets like a product manager, not a support lead. Tag every ticket with the UX failure behind it. Fix the top five each month. Watch the same tickets stop appearing.

Make "tickets per active user" a product metric. Not CSAT. Not NPS. Ticket volume per thousand active users, trending down over time. That is the only number that tells you the product is getting easier.

Treat docs as a fallback, not a feature. If your docs are carrying the product, rewrite the product. A help center is a graveyard of UX decisions you did not want to make.

Ship before the tutorial. If new users need a video to start, the first screen is wrong. Make the right action obvious. The video is a crutch. The crutch becomes permanent.

Watch three users a month complete the main task with zero help. Steve Krug's rule, repeated in Stop Making Your Users Think. You do not need a lab. You need three people and one morning. Every session will reveal tickets you would have gotten later.

When Support Still Matters

Zero support is a direction, not a literal goal.

Some support is legitimate. Regulated industries where a human must confirm the decision. High-stakes enterprise where the relationship is the product. Edge cases that genuinely cannot be automated. Recovery paths after something unusual goes wrong.

The test is not whether support exists. The test is what the support is for.

Support that helps users recover from rare events is a feature. Support that explains basic flows is a bug. One serves the customer. The other covers for the product.

If you cannot tell which kind you have, look at your ticket tags. If most of them map to the core user journey, you are paying humans to explain your product. That is the expensive kind.

Conclusion

Every support rep you hire to explain a screen is a permanent cost you are paying for a one-time design decision you never made.

Flip it. Treat every ticket as a brief against the product. Treat your support queue as your most honest backlog. Treat "how do I do X?" as the most important question in your company, and answer it in the interface, not on a help desk.

The best support team is the one you did not need to grow.

Aim for zero support. You will never get there. That is the point.