Inbound marketing agency: what the label means in practice, what a retainer produces each month, and what it costs

An inbound marketing agency is hired to make a company findable and persuasive to buyers already looking, then to capture and nurture them until sales takes over. The label now covers three different purchases, and a large share of the firms using it are platform implementation partners, because the word was popularised by HubSpot and the ecosystem around that software adopted it. So the first question for any firm on your list is plain: are you selling me content and search visibility, a software implementation, or a lead nurturing operation? All three are legitimate and priced differently.

If you are still deciding whether this approach suits your market at all, the comparison with buying attention directly is set out in inbound and outbound marketing, the real difference. This page assumes you have decided and are now buying.

The three purchases

• Demand creation. Research, publishing, search visibility, lead magnets, conversion paths. The original meaning of the term and the one most buyers have in mind, judged on qualified enquiries from organic sources.
• Platform implementation and operations. Setting up and running the software: forms, lists, workflows, lifecycle stages, lead scoring, pipelines, attribution reporting, the handoff rules between marketing and sales. Judged on whether the reports are trusted.
• Nurture and sales enablement. Email sequences, collateral, follow up cadences and the content answering late stage objections. Judged on enquiry to opportunity conversion.

Plenty of firms sell all three, which is fine as long as the scope says how the fee divides and who owns each outcome. The failure case is one undifferentiated retainer where implementation eats two quarters and the publishing never reaches a volume that could produce anything.

What a US retainer costs, and what it should produce

These figures are my own working estimate of the US market, from scopes and proposals rather than any survey, and they are set against output on purpose, because here a price with no volume attached tells you nothing.

Monthly fee	Realistic monthly output	What is usually missing at this level
2,500 to 5,000 dollars	Two to four substantial articles, basic on page work, one landing page a quarter, a monthly report	Original research, design resource, sales alignment, link acquisition
5,000 to 10,000 dollars	Four to eight pieces including one larger asset, conversion paths built and tested, workflow upkeep, quarterly planning	Dedicated strategy time, video, outreach and digital PR
10,000 to 20,000 dollars	A named strategist, eight or more pieces, original research once or twice a year, lifecycle workflows, reporting on your CRM data, sales material	Little, if the scope is written properly
20,000 dollars and above	Multiple product lines or regions, a pod of named people, demand and brand running together	Nothing, though insist the team is named with its time allocation

The project and unit prices alongside a retainer are worth knowing too. Platform onboarding runs 5,000 to 30,000 dollars as a one off, by how many objects, integrations and legacy lists have to move. A strategy sprint producing positioning, a topic map and a measurement plan runs 7,500 to 20,000 dollars. Individual articles bought without a retainer run 300 to 1,500 dollars. A substantial pillar asset, such as a research report, a calculator or a buyer's guide, runs 2,500 to 9,000 dollars. A freelance inbound specialist charges 90 to 200 dollars an hour.

The software licence is a separate number and not a small one. Automation platforms price on contact tiers and feature levels, from a few hundred dollars a month at entry level to several thousand for a mid market suite, with vendor onboarding sometimes charged on top. Ask every agency to quote fee and licence apart, and ask what referral commission or partner margin it takes from the platform it recommends. Many take one. The disclosure matters more than the commission.

The condition that decides whether this works

Inbound pays back when search demand already exists for the problem you solve and the buyer researches before committing. Two checks before you sign.

• Count the demand. Pull search volumes for the ten questions your buyers actually ask. A combined monthly total in the low hundreds means publishing will not build a pipeline alone, and the money belongs in outbound to a named list.
• Count your capacity to feed it. Good inbound needs a subject expert for an hour or two a week, somebody who can approve copy inside a day, and access to customers and sales calls. An agency starved of those produces material that reads like everybody else's.

On timing, be realistic in the contract rather than hopeful. New pages in a competitive category commonly take two to three quarters to reach a stable position, so a six month contract measures setup and a twelve month one measures the programme. Ask for leading indicators from month two: pages published against plan, pages gaining impressions, conversion rate on the paths built, and enquiries from organic sources however small the number.

Five things to ask the firm that pitches you

• Who writes, and will I speak to them before signing? The strategist in the pitch is rarely the writer on the account. Ask for a named writer and a sample from your industry.
• Show me a programme you took from nothing to a measurable pipeline, with dates. Look for a timeline and a traffic to enquiry conversion figure, not a traffic chart.
• How will you get our subject expert's knowledge onto the page? Recorded interviews, call recordings, access to sales. A firm with no process here writes from competitor research.
• What do you do about links and mentions? Publishing alone rarely wins a competitive query. Refuse anything that involves paying for links.

What you keep when it ends

• Copyright in everything produced, assigned to you in writing, drafts and underlying research included. Beware a licence to use rather than ownership.
• Working files, not only published pages: design sources, spreadsheets, interview recordings and transcripts, plus the topic map and measurement plan as documents the next supplier can pick up.
• The platform account in your company name, billed by you, agency as users, with workflows, lists, lifecycle definitions and reports documented in plain language.
• Disclosure handled. Case studies, customer quotes and content produced with an incentive attached are endorsements under US advertising rules, so a material connection has to be disclosed and a claim has to be supportable. Agree who holds the evidence file, because the duty rests with your business rather than the agency.

When something other than an inbound agency is the answer

• Your market is a list of three hundred named accounts. Outbound and account based work reaches them faster and cheaper, with publishing supporting the conversation rather than starting it.
• You already have a writer and a plan. Buy specialist help for the parts you lack, usually technical search work or design, rather than a retainer duplicating your own people.
• The software is the blocker. Buy the implementation as a project, get it working, then decide about content separately.
• You need revenue this quarter. Inbound compounds rather than sprints. Paid media, partnerships and direct outreach are the honest answers to a quarterly number, with the inbound programme started in parallel so next year differs.

Buying an inbound programme: the recurring questions

Where does an inbound agency end and a content agency begin?

They overlap heavily. A content agency sells production: research, writing, design, video. An inbound agency adds the capture and nurture machinery around it, which usually means software configuration and lifecycle reporting. Without a need for the machinery, you are paying a premium for it.

Do we have to use HubSpot?

No. The method predates and outlives any one platform, and plenty of firms work in other automation tools or in a simpler stack of a content management system, an email platform and a CRM. A firm that cannot operate outside one piece of software is telling you where its expertise sits, which is fine as long as you are also buying that software deliberately.

Should we commit to twelve months at the start?

Six months with a clear ninety day review, then a twelve month term once you have seen the work. Agencies push for twelve up front because the model needs runway, which is a fair argument. Meet it with a break clause tied to named deliverables rather than to outcomes nobody controls in month four.

Is four pieces a month really enough?

Fewer than most proposals suggest, done properly. Four strong pieces answering a real buying question, each with a reason for somebody to link to it, beat twelve written to a volume target. Ask for a publishing plan built from your buyers' questions rather than a keyword export sorted by volume.

Being findable now includes being quotable

The premise of inbound is to be present at the moment somebody goes looking, and that moment has partly moved. A buyer researching a purchase now asks ChatGPT, Gemini, Perplexity, Claude or Google AI Overviews to lay out the options and pick one, then reads a single composed paragraph instead of clicking through four articles. A good inbound library is the raw material those paragraphs are built from, and that is the opportunity: pages answering a buying question plainly, with specifics and numbers, are what a model can lift and attribute. The awkward part is that nobody tells you when it happens. Analytics shows the visit, never the sentence that caused it, and a library can be feeding answers that name a competitor. Whether yours is the thing being quoted is a question AI Knows Us answers directly, page by page: aiknowsus.com.