Customer Questions and AI Search: What Plumbing Companies Should Understand

There is a great deal of noise about AI and search at the moment, much of it selling
something. So it is worth starting with what the primary documentation actually
says.

Google’s published guidance on succeeding in AI experiences is unusually blunt:
there is no special markup for AI features, and no separate optimisation process. The
advice given is the same advice as for Search — unique, satisfying, people-first
content, technically accessible, with the page’s meaning clear.

The best AI-search strategy starts with making the business easier to
understand.

Why questions became more valuable

When search returns a list of links, the value of answering a question is that
somebody clicks through to read your answer. When a system answers directly, the value
changes shape: your explanation may inform an answer, and your company may be named as
somebody who handles that problem in that area.

Either way, the raw material is the same. Something on the public web has to actually
contain the answer, clearly, attached to a company whose identity, services and service
area are unambiguous.

Most plumbing companies are sitting on an enormous supply of this raw material and
have published none of it.

The questions worth writing down

Not invented keyword questions. The real ones, in the words customers use.

• “Why does my water heater keep running out of hot water?”
• “Is it worth repairing this or should I replace it?”
• “How do I know if I have a slab leak?”
• “Why does my drain keep backing up even after it’s cleared?”
• “What causes the white scale on my fixtures, and does a softener fix
  it?”
• “How long should this take and what will you need access to?”

Your technicians answer these constantly. The answers are already good, already
tested against real customers, and almost never written down.

What a good answer looks like

• Answer first. Lead with the answer, then explain. A reader
  shouldn’t wade through three paragraphs of preamble.
• Use the customer’s words. They say “the toilet keeps
  running,” not “flapper valve failure.” Use theirs first, yours
  second.
• Be specific about uncertainty. “Usually X, but if Y is also
  happening it is more likely Z” is more credible and more useful than false
  confidence.
• Don’t hide the answer to force a call. Content that withholds
  the useful part performs badly with readers, and it is the opposite of what the helpful
  content guidance describes.
• Connect it onward. A reader who now understands the problem should
  have an obvious route to the relevant service.

Where structured data actually fits

Schema.org vocabulary lets a page state what it is and what it describes, in a form
machines parse without guessing. That is genuinely useful, and it is worth doing
properly.

It is also worth being precise about its limits. Google states plainly that
structured data does not guarantee a rich result. And FAQ structured data in particular
has had its display eligibility narrowed substantially — which is a good
illustration of why building a strategy on a specific result feature is fragile, while
building it on being genuinely clear is not.

Two rules that keep you out of trouble: mark up what is actually visible on the page,
and never mark up something the page doesn’t contain.

The unglamorous prerequisites

None of this works if the content can’t be reached. Pages have to be crawlable
and indexable. Important text has to be in the HTML rather than locked behind an
interaction. Your business name, service area and services need to be consistent between
your site and your Business Profile, because inconsistency creates exactly the ambiguity
you are trying to remove.

Where to start

1. Ask your technicians for the ten questions they answer most often. Record the
   answers as they say them.
2. Publish those answers where they belong — usually attached to the relevant
   service, not dumped in a single FAQ page.
3. Make sure each one routes to the service that solves it.
4. Check the basics: is it indexable, is the business information consistent, does the
   page say plainly who you are and where you work?

The FLOWLINE view

The companies most likely to be described accurately by an answer engine are the ones
whose public information is accurate, specific and consistent. That is not a trick, and
it is not new. It is the same discipline that made a website useful before any of this,
which is the most reassuring thing about it.