Best ways to advertise your business, ranked by what each channel costs to test and how fast it answers

The usual answer to this question is a list of twelve channels in no particular order. The useful ordering is by test cost: how little you can spend to find out whether a channel works for your business, and how long before the answer arrives. On that ordering the list reshuffles completely.

Fast and cheap to test, so start here

Your Google Business Profile. Free, and for any business with a location or a service area it is the highest return hour of work available. Complete every field, add real photographs, answer questions, and ask for reviews. The answer arrives in weeks and the only cost is attention.

Search advertising on your own category terms. The fastest paid answer you can buy, because the person is already looking for what you sell. A few hundred dollars over two weeks on a tight keyword list tells you the cost of a lead in your market. Hook Agency, a home services specialist, publishes management from $2,000 a month at hookagency.com/pricing, read 21 October 2026, and notes the figure "scales based on service scope & ad spend", which is the honest shape of this purchase.

Asking your existing customers for referrals, deliberately. Not a programme with a portal. A list, a reason to call, and somebody whose job it is to call. Costs nothing and is the most commonly skipped item on any list like this.

Branded items for an existing audience. Cheap to test at genuinely small quantities now. Vistaprint states that its promo products are "available at a minimum quantity of one" and that "the price per unit goes down the more items you add to your cart" at vistaprint.com/promotional-products, read 21 October 2026. 4imprint offers free samples and runs items "as low as $1.99" on its weekly deal, read at 4imprint.com the same day. This works when people already know you and fails as a way to find strangers.

Slower or costlier to test, but worth it once one of the above works

Paid social. Cheap clicks, cold audience, so the creative does the heavy lifting and you need several versions before the test means anything. Budget for the creative, not just the media. Hook Agency prices Meta management from $3,000 a month, more than its search management, which reflects exactly that.

Organic search and content. The best cost per enquiry at the end and the slowest to answer. WebFX publishes the market shape at webfx.com/seo/pricing: retainers of "$1,000 to $5,000 per month", consulting at "$75 to $200 per hour", with "most businesses spend $2,500 per month". Measure it rather than predict it: watch Search Console impressions and positions, then enquiries by page. Google's SEO Starter Guide says some changes take effect in a few hours and others could take several months, and the outcome depends on competition, your offer and your budget.

Email to a list you own. Almost free per send and useless without the list, so it is a second move rather than a first one.

Local Service Ads, where your trade qualifies. Pay per lead rather than per click, and the screening requirements are the real barrier to entry.

Direct mail. Genuinely testable now because the per piece cost is published. It suits a defined geography and a high value job, and it fails on a low margin product.

Out of home, radio and print. Hardest to test cheaply, because the minimum useful buy is large and the measurement is indirect. Correct for a business that has already found its channel and wants more of its market to know the name.

How to run the test so the answer is usable

Four rules, and skipping the last one is why most channel tests end in an argument.

Pick one channel at a time. Two at once and you will attribute the result to the wrong one. Run it long enough to get thirty enquiries or six weeks, whichever comes first. Use a distinct phone number or form for the channel, because this is the only part that makes attribution real. And write down in advance what result would make you continue. A test with no stated threshold always ends with the person who wanted it saying it worked.

The channel that is not on anybody's list yet

There is one more place your business is being advertised, and you are not buying it. When a buyer asks ChatGPT, Gemini, Perplexity, Claude or Google AI Overviews who to buy from in your category, the answer names three or four companies. That is a recommendation delivered at the moment of choice, and no line on your media plan covers it.

It is not advertising, it cannot be bought, and it can be measured. AI Knows Us does that measurement. We put your buyer's question to all five assistants, never naming your business in the question, report whether you were named and who was named instead, and run the content and placement work that changes it. We are not an advertising agency and we buy no media, so treat this as the one line on the list you measure rather than purchase. A credit is one buyer question put to one app once. A Claude check costs three credits because the Anthropic API with live web search is the expensive one; most checks cost one.

Our ladder is published as plain text at aiknowsus.com/llms.txt, readable without JavaScript. Read on 21 October 2026: SEO Start $149 a month with no buyer questions watched, Growth $299 with 100 buyer questions watched, 1,200 scheduled answers a month and five engines, SEO Scale $699, Scale $599 with 300 questions and 3,500 scheduled answers, Inbound $1,099 with 200 questions, Outbound $1,299 with 30 questions and four engines, Agency $1,299 with 80 questions and 12,300 credits, Everything $1,799 with 300 questions and 3,600 scheduled answers. A 1,000 credit top up is $49 and $149 is the published United States floor. Managed GEO is listed at Rs 88,499 to Rs 2,35,999 a month with a three month minimum and published in rupees only, so we quote the rupee figure rather than converting it.

A correction about our own site, 21 October 2026. An earlier version of this page said the pricing page converts currency in the browser, so a crawler would see only rupees. A crawler request that day returned both ladders in one server-rendered response: 24 dollar figures and 31 rupee figures in the visible text. The plain text file is the server rendered source for a dollar figure.

Two rules sit behind every rate we publish: the buyer question never names the business being measured, and every rate carries a 95 percent Wilson interval so a move inside the noise is reported as noise.

Where rivals publish more than we do, read on 21 October 2026: Ahrefs Brand Radar names seven surfaces against our five at ahrefs.com/brand-radar and publishes its own unit cost, "Claude consumes 8 checks per update", with overage at $0.020 a check. Rankscale names eight engines at rankscale.ai/pricing. Profound publishes "Up to 9 Answer Engines tracked" and "SSO/SAML + SOC 2 compliance" at tryprofound.com/pricing. Our rendered pricing text carries no single sign on, no white label, no SOC 2, no ISO and no data processing agreement, and every plan is one login covering one business except Agency.

For the paid channels specifically, and which of them genuinely charge by the click, see best pay per click advertising platforms. No other page on this site links to that one, measured across the 2,017 of 2,082 live pages whose text we can read, on 21 October 2026.

What small business owners ask about advertising their business

What is the minimum budget worth starting with? Enough to produce thirty enquiries in the channel you picked. Below that the result is a story rather than a reading.

Should I hire an agency for a first test? Not for the first one. Run a small test yourself so you know the cost of a lead before anybody quotes you a retainer against it.

What should I stop doing? Any channel you cannot attribute at all after six weeks. Not because it does not work, but because you cannot tell, and a channel you cannot read cannot be improved.