You Need to Understand Their Problem Before You Fucking Market to Them

One way to fuck up marketing is to become so knowledgeable about what you sell that you forget what the customer doesn't know.

Spend ten years inside an industry and its terminology starts sounding ordinary. You learn the categories, distinctions, and acronyms. Everyone around you understands exactly what everyone else is talking about.

Then somebody says, “Great. Let's build the marketing.”

And the entire fucking company proceeds to market to itself.

There is one minor problem.

Your customer didn't attend the meetings.

Your Customer Does Not Know What You Know

Customers may not know your terminology, your service categories, or what kind of professional they need. They may not know that the thing that happened to them is something your industry can help address.

Why the fuck would they?

That may be precisely why they need a professional.

There is research behind the difficulty, although it does not establish the performance of your particular marketing campaign. In a 1989 experimental paper, Colin Camerer, George Loewenstein, and Martin Weber examined the “curse of knowledge”: better-informed participants had difficulty setting aside information when judging from a less-informed position. [1]

The practical question for a business is simple.

Can you explain your offer from the customer's starting point, or only from your own?

Google's SEO Starter Guide makes a related, narrower point: people familiar with a topic may search using different terms from people new to it. It also says its language-matching systems can connect related queries without requiring every exact phrase on the page. [2]

So this is not an instruction to stuff a hundred awkward search phrases into your website.

It is an instruction to understand the person trying to use it.

The Customer Starts With the Event

The orthopedic surgeon knows the procedures they can offer. The patient knows their fucking knee hurts.

The air-conditioning contractor can investigate whether a compressor has failed. The homeowner knows the house will not cool down.

The cybersecurity specialist understands identity and access controls. The business owner knows, “Someone keeps getting into our employee accounts.”

The attorney understands the possible legal issues. The prospective client knows, “My business partner took money out of the company and now he will not answer my calls.”

Those descriptions do not establish a diagnosis. They establish where the conversation starts.

The provider thinks through symptoms, possible causes, classification, options, and solutions. The customer may still be standing at: What the fuck just happened?

Then we build the marketing for somebody who already knows the answer.

In my work with attorneys, this distinction matters. An attorney might investigate potential professional malpractice. The customer says, “My lawyer fucked up my case.” An attorney might investigate possible insurance bad faith. The customer says, “My insurance company will not pay me.” An attorney might evaluate a premises-liability question. The customer says, “I fell at a store and got hurt.”

Or someone says, “My boss fired me after I complained.” That calls for questions. It does not, by itself, establish wrongful termination.

A disappointing outcome is not automatically malpractice. An unpaid insurance claim is not automatically bad faith. The facts and applicable law determine whether a claim exists.

The customer can accurately describe what happened without knowing its professional classification. If your marketing only recognizes the classified version, you are effectively saying: Please diagnose yourself before contacting the person you are trying to hire to evaluate the problem.

That is fucking backwards.

Search Language Reveals Different Information Needs

Consider these invented searches:

“contingency malpractice attorney near me”

“my attorney messed up my lawsuit what can I do”

“lawyer missed deadline on my case”

The first uses a service category and a payment arrangement. The others describe an event or concern.

None proves that the person has a viable claim, can become a client, or even understands the words they typed. People can copy terminology they just found elsewhere. But the wording suggests different information needs worth investigating.

Someone describing the event could be a suitable prospect who has not learned the vocabulary. Someone using the technical phrase could already be comparing providers.

Your customer should not need to become you before your marketing recognizes them. And you do not have to make an informed buyer sit through a beginner's lecture, either.

Credibility and Relevance Are Not Interchangeable

Imagine a generic billboard. Attorney's face. Firm logo. Telephone number. Enormous lettering about millions recovered.

That can communicate experience and historical results when the claim is accurate and properly presented. It can give someone a reason to remember the firm.

But picture a person whose car was hit yesterday. Their vehicle is fucked. Their neck hurts. The insurer called. They do not know what happens next or whether they need an attorney.

What does the recovery number help them understand at that moment?

It might establish credibility. It might leave their immediate question unanswered. The firm is communicating something it is proud of. The customer needs to understand whether the firm is relevant to what just happened.

Those jobs can work together. They are not interchangeable.

There is also a legal boundary: California's Rule 7.1 prohibits false or misleading lawyer communications, and its commentary explains how even truthful past results can create unjustified expectations. [3]

Plain language does not give you permission to promise a result you cannot support. It gives you a way to explain what you can actually do.

The Misunderstanding Can Continue Through Intake

Fine. Add some long-tail keywords. Done.

No.

The misunderstanding can continue through the entire acquisition process. What does the advertisement say? What does the landing page explain? What does the form ask? What does the person answering the phone expect the caller to know?

If the form requires an unfamiliar service category, some people may choose incorrectly or leave. If intake treats an uncertain description as evidence of poor fit, it may reject an inquiry before gathering the information needed to evaluate it.

Those are hypotheses to test in your business, not excuses to declare every rejected inquiry valuable. But if the whole process assumes customers already understand the professional classification, you have built an expert-facing acquisition system for non-experts.

The consequence can be expensive in either direction. Suitable people fail to recognize themselves in the offer. Unsuitable people misunderstand it and contact you anyway. Employees spend time repairing confusion the marketing created.

More inquiries alone will not tell you which problem you are solving.

A Service Funnel Is Not E-Commerce

Sometimes a digital action and a transaction are close together.

I see shoes. I like shoes. I click shoes. I buy shoes.

Beautiful.

When tracking works, the business can connect an order with a price and then account for delivery costs, returns, and other expenses. Even then, revenue is not profit and attribution is not proof that the ad caused the purchase.

Now compare a service requiring consultation, investigation, a proposal, and a decision before work begins. Someone searches, clicks, reads, contacts you, and explains the problem. Your business gathers information. Someone determines whether it can help. A contract may be signed. The work has to be delivered and payment collected.

For some services, final economic value takes months or years to establish. Installing a form does not remove those steps.

A law firm did not become e-commerce because somebody installed HubSpot.

An air-conditioning company can sell filters online and still need a different process for a major system replacement.

Google Ads distinguishes inquiries from qualified and converted leads and supports importing later-stage information. Those stages depend on your definitions and records. [4]

The technology between the customer and the business does not settle the economics of the transaction.

I Want the Fucking Phone to Ring

When I am marketing a service whose sale depends on a conversation, I want marketing to help produce that conversation.

In the legal businesses I am describing, the phone can be an important route. So yes: I want the fucking phone to ring.

But ringing is not the final outcome either.

A suitable inquiry needs a useful response. The business has to understand the problem, determine whether it can help, explain the options, and handle the next step competently.

Some customers prefer or need a written channel. A useful form, email, or chat can start the same process. The route should fit the customer and the service.

Watching 75% of a video or visiting three pages can be a useful signal. Neither automatically establishes that the customer understands the offer or that the business acquired a client.

The question is what happened next.

KevinBot and the Economics of Automation

Here is a hypothetical.

A business spends $500 generating an inquiry for a complex service. After the advertising, content, tracking, and persuasion, a human finally says, “I have a problem. Can somebody help me?”

The business responds, “Absolutely. Please explain your complicated situation to KevinBot.”

KevinBot is fictional. The decision is real enough to examine.

If the system asks clear questions, answers accurately within its role, and hands off appropriately, it may help. If it demands a diagnosis the customer does not have, you have automated the same misunderstanding.

Suppose automation saves a net $40 per inquiry compared with the staffed process, after accounting for the automation's costs. Suppose a completed, paid engagement contributes $5,000 after variable delivery costs, before acquisition and intake costs.

If the automated process reduces the close rate from 20% to 19%, expected contribution falls by $50 per inquiry:

One percentage point × $5,000 = $50.

You saved $40 and lost $50 of expected contribution.

Net result: $10 worse per inquiry under those assumptions. The original acquisition cost is the same in both versions, so it does not change that comparison.

These are invented numbers, not an AI performance benchmark. Hold lead mix, service quality, and other costs equal for this simple example.

If automation preserves the close rate, the saving may be real. If it answers inquiries you previously missed, it may improve the outcome. If it changes contract value, refunds, or delivery costs, include those too.

The cost of the interaction and the value of the interaction belong in the same decision.

AI Can Help Without Replacing the Conversation

This argument has limits. Some customers know exactly what they need. Some professional buyers expect precise terminology. Some services are purchased entirely online. Credentials and documented results can help people choose between qualified providers.

Do not erase useful information to make every page sound like an introductory conversation.

And AI can be valuable. Research by Erik Brynjolfsson, Danielle Li, and Lindsey Raymond found productivity gains from an AI assistant supporting human customer-service agents. That studied assistance in a particular support setting; it did not establish that an unattended bot improves every professional-service intake process. [5]

I use AI. I am interested in what it helps the business do.

Summarize information. Reduce repetitive work. Assist staff. Route inquiries. Make it easier for a person to get an accurate response.

The same questions apply to humans. An employee who misunderstands the customer does not become effective simply because they are not software.

The argument is about starting from the customer's understanding and measuring the result.

Test Whether People Understand the Offer

Ask intended customers to read the page and describe, in their own words, who it helps and what they would do next. Listen to the questions people ask before your team supplies the terminology.

Then compare suitable versions of the message: one organized around the professional service, another beginning with the customer's event or problem. Keep the offer and other conditions comparable, and give the test enough time and volume to mean something.

Follow the result beyond clicks.

Did people understand the service more accurately? Did suitable inquiries increase? Did unsuitable inquiries or handling costs increase? Did more paid work result at acceptable economics?

If customers already understand the technical version and it produces better outcomes, use it. If plain language creates more confusion, attracts mostly unsuitable inquiries, or reduces worthwhile business, revise the explanation.

If customers understand the offer but nobody answers them, wording may not be the constraint.

A thin sample or several simultaneous changes can leave the cause uncertain. Do not declare victory because the version you preferred got three more clicks.

Your customer research has to be allowed to disagree with your fucking theory.

Sequence the Expertise Correctly

Businesses naturally want to advertise what they are proud of. Attorneys advertise results. Contractors advertise equipment. Technology companies advertise features. Consultants advertise methodologies. Agencies advertise capabilities.

Everybody wants to talk about the solution. Of course they do. The solution is the part they spent years learning.

But the customer may still be standing at the event.

Something broke. Something hurts. Someone took their money. Something stopped working. Something happened that they do not understand.

Start there when that is where your customer is. Then move toward the questions, evaluation, options, and service that can actually help.

This does not mean dumb down your expertise.

It means sequence it correctly.

Before building the campaign, buying the billboard, writing the landing page, configuring the form, or deploying the chatbot, ask:

What does this problem look like to someone who does not know what I know?

The customer does not need to arrive fluent in your profession.

You need to understand their problem before you fucking market to them.