Be the cleaning company the assistant recommends.

They are real and still early, which is exactly how we treat them. OpenAI's advertising business is approaching a $1 billion run rate with more than a billion weekly users on the platform, and the formats and targeting controls are still changing month to month. We run a funded test with written stop conditions rather than moving your main budget, and we report what actually happened. If it is not earning its place for your market, you will hear that from us first.

No. Prompt lists are a research tool, not a product, and we use them the way an agency uses a keyword tool: to learn how buyers describe a job and to test what assistants currently say about cleaning companies in your area. What you pay for is the work that follows — correcting what the assistant can verify about your business, publishing answers worth quoting, buying placement carefully, and measuring the contracts that result.

Absolutely not, and anyone saying otherwise is selling novelty. House cleaning has the lowest cost per lead of any home services category at roughly $12-$30, which is a very hard benchmark to beat, and 52% of consumers still start a local search on Google. We treat assistant advertising as an additional surface funded from a test budget, with your proven channels left intact and managed by the same team so nothing is double-counted.

Verifiability. These systems shortlist businesses whose details they can confirm: a correctly categorised profile with real service areas, consistent details across the web, service pages stating what a clean includes and roughly what it costs, insurance and vetting stated plainly, and a steady flow of genuine ratings — 85% of consumers say positive reviews make them more likely to use a business, and assistants weigh the same signals. Most of our first month goes into exactly this.

It is often stronger on the commercial side, because facility managers research in writing. They describe a building, ask what a nightly contract should cost per square foot, and ask for companies that carry the right insurance and can pass a vendor check. That is a long conversation an assistant is good at summarising, and a company with published scope, certifications and references gets named. We split residential and commercial into separate plans, since the buyer and the wording differ completely.

A fixed monthly management fee scoped to the work, plus a test media budget you set and can stop at any point. We recommend starting small enough that a poor result is a lesson rather than a problem, and we write the stop conditions into the plan before anything runs. Most of the early value comes from the groundwork rather than the media spend, which is why we sequence it that way rather than opening an ad account in week one.

The same way we measure anything: by the contract at the end of it. Calls, forms and your scheduling system are connected so an enquiry is traced to a first job and then to recurring work, since 30-40% of first-clean leads convert to recurring service and that conversion is where the business value sits. We compare cost per booked route against your existing channels every month, and we show the comparison even when it is unflattering to the new surface.

Small and local is an advantage here, because assistants answer specific questions and specificity is what you have. A company that clearly states it covers four suburbs, does move-out cleans with a landlord checklist, is insured and bonded, and prices a three-bedroom home in a stated range is far easier to recommend than a national brand with a generic page. The work scales down comfortably; the fee should too, and we will tell you if the smallest useful version of this is not worth doing yet.

The risk of being described inaccurately exists whether or not you engage with the surface, and engaging is how you correct it. We record what assistants say about you at the start, fix the underlying sources they are reading from — profile, website, directories, reviews — and re-test on a schedule so drift is caught. Everything we publish is checked with you first, so the version of your company in circulation is the one you would say out loud.

The baseline lands in the first fortnight and usually changes the conversation on its own, because seeing which three competitors get named is clarifying. Groundwork fixes take four to six weeks to be reflected in what assistants say, since they re-read sources on their own schedule. The ad test produces readable numbers within a few weeks of starting. A fair verdict on the whole thing takes about a quarter.

Our writers draft it and your supervisors correct it, because the detail that makes a page quotable — what is included in a standard clean, what products you use around pets, how a key handover works — lives with the people doing the job, not with a copywriter. Nothing goes live without someone in your business signing it off, and the finished pages serve your website and your quotes as much as they serve any assistant.

Yes, and the numbers get more honest when one team holds them. Our paid search, SEO and analytics specialists work from the same plan, so a booked contract is attributed once instead of claimed by three suppliers. It is equally fine to hire us only for this test while somebody else keeps the rest — we document everything and leave every account open in your name.