B2B Inbound Marketing

Getting Found Instead of Buying Attention

Inbound marketing describes a programme built so that buyers find you while they are looking, rather than being interrupted while they are not. The term is widely credited to HubSpot's co-founders, Brian Halligan and Dharmesh Shah, who used it in the mid-2000s to name a specific contrast: a company blog, search visibility and social distribution, set against cold calling, rented lists, print advertising and trade-show badge scans.

What it originally described was narrow and concrete — publish, be indexed, be found, convert. It was also a category-creation exercise by a company selling the software to do it. That does not make the model wrong; it explains why the label expanded until it described almost everything a marketing department does.

The useful definition is narrow again: inbound is the part of your programme where the buyer initiates contact, having found material you published, at a moment they chose. Everything else — advertising, outbound sequences, events, the trade press — is a different mechanism with different economics.

Why the Model Suits Business Buying, Up to a Point

Cycles are long and involve several people, so material that can be read at any hour and forwarded to a colleague does work a sales conversation cannot. Buyers self-educate before they want to be sold to, and a supplier visible during that education is in the consideration set by default. Procurement and security review raise factual questions a website answers better than a call.

The limit is arithmetic. Ehrenberg-Bass's Professor John Dawes proposed the 95-5 heuristic in a May 2021 paper published through the LinkedIn B2B Institute, derived from interpurchase intervals rather than from a survey of buyers: because purchase cycles are long, the large majority of category buyers are out of market at any given time. Dawes explicitly calls 95% a heuristic rather than a precise figure, and the true share depends on your category's replacement cycle.

That is the central caveat about inbound: a programme built only to be found serves whoever happens to be in market this quarter.

The Mechanics, End to End

An inbound programme is a chain, and only as strong as its weakest link.

A published asset that answers a real question better than the alternatives — a page, a guide, a tool, a recorded session.
Discovery, mostly through organic search, plus referrals, social distribution and AI assistants, which are identifiable by referrer and now sit in GA4's default AI Assistants channel group.
A page that delivers on the promise, loads acceptably, and is readable on a phone.
A conversion route proportionate to the visitor's stage: a short form, a calculator, a recorded demo, or a phone number.
Follow-up that continues the specific conversation the visitor started rather than dropping them into a newsletter.
A handoff with an agreed definition of who sales contacts, how quickly, and what happens to everyone else.

The tooling is unremarkable: a CMS, marketing automation such as HubSpot, Marketo or Salesforce Account Engagement, a CRM, GA4 and Search Console. None of it fixes a weak asset or an unagreed handoff, which is where programmes break.

Capturing Demand Versus Creating It

This is the distinction that calling everything inbound obscures.

Capturing demand means being present for a search that already happens. Keyword research returns volume, and your job is to be the best answer. Creating demand means persuading people a problem is worth solving before they have named it. Keyword research returns almost nothing, because nobody searches for a category they do not know exists.

If you sell managed print services, buyers search. If you sell a genuinely novel category, or a service for a problem organisations tolerate silently, they do not. Inbound cannot capture demand that has not formed. There the money belongs in mechanisms that create demand: outbound to a defined account list, trade press and events, conference speaking, paid social aimed at a job function.

Most companies need both in a deliberate ratio. The damage the word inbound does is to let a team spend an entire budget harvesting a small in-market share and call it a growth strategy.

Lead Scoring and Its Reliable Failure Modes

Lead scoring assigns points for attributes and behaviours, then routes anything above a threshold to sales. In principle it prioritises attention. In practice it fails in predictable ways.

It scores activity, not authority. An intern downloading six PDFs outranks the operations director who read one page and forwarded it.
It rewards the wrong populations. Job applicants, students, competitors and your own agency all generate behaviour.
It collapses fit and intent into one number. A perfect-fit account with no activity and a poor-fit visitor with plenty land on the same score.
It is fitted on samples too small to fit anything. With a few dozen closed-won deals a year, the weights are guesses expressed to two decimal places.

What works better is duller: firmographic fit on one axis, a short list of behaviours only a buyer performs on the other — pricing page, implementation documentation, a security questionnaire, a demo request — plus a weekly conversation with sales about named records.

Common Mistakes

Treating inbound as a substitute for sales. It changes who starts the conversation, not whether someone has to have it. Programmes without follow-up discipline produce contact records, not revenue.
Publishing to a funnel stage rather than a question. Buyers do not experience stages; Gartner's model is a set of jobs revisited in loops.
Demanding budget and timeline on a first form. A form asking for company size, budget and timeline before the visitor has read anything filters out the researcher who would have become your champion.
Nurture that ignores what the person did. A sequence about a topic the recipient never touched trains them to ignore your name.
Reporting sessions to the board. Traffic is not an outcome. A rise in unqualified visits looks like progress and costs the sales team time.
Killing the programme before the payback window closes. The most expensive mistake here, and the easiest to prevent by agreeing the horizon in writing before you start.

Measurement, and What to Tell the Finance Director

Report enquiries by source with a stated definition of source, qualified enquiries as judged by sales rather than by score, pipeline created and its age, and non-branded impressions and clicks in Search Console. Ignore raw traffic, bounce rate, and last-click attribution, which in long cycles credits whichever channel happened to be last.

The traps are specific. Deals close a quarter or more after the touch that earned them, and volumes are small enough that ordinary variance looks like a trend. GA4's data thresholds are system-defined and unpublished, so the small segments B2B reporting depends on quietly disappear.

To a CFO, describe inbound honestly: an asset build with a lag, not a purchase of leads. Pages must be indexed, earn position and accumulate links before the sales cycle adds its own quarters. Agree a review horizon longer than that cycle, report leading indicators in the interim, and say plainly that paid search buys the same enquiry sooner at a higher marginal cost — a trade, not a defeat.

When Inbound Is the Wrong Investment, and How to Start

Inbound is a poor first investment in three situations. When nobody searches for what you do, because the category is new or the problem is unnamed. When your entire market is a hundred named accounts, because you can reach every relevant person directly for less than the cost of ranking. And when the business needs revenue this quarter, because inbound does not work on that timescale and pretending otherwise gets it cancelled at month four.

Where it does apply, start narrow. Pick the three commercial questions that precede a purchase in your category and answer them better than anything currently ranking. Make the conversion route proportionate — one short form, not a qualification interview. Agree with sales, in writing, what happens to an enquiry within one working day. Then leave the pages alone long enough to work, and add breadth only once the core produces enquiries you would be happy to hand a rep.

Frequently Asked Questions

What is the difference between inbound and outbound marketing?

Inbound means the buyer initiates contact, after finding something you published. Outbound means you initiate, to a list you selected — cold email, calls, direct mail, advertising into an account list. The practical difference is control versus cost: outbound gives you timing and targeting but you pay for every touch, while inbound costs less per enquiry once it works and you cannot choose when it happens.

Outbound creates demand in categories buyers do not search for; inbound captures it once they do.

Does inbound marketing work if we only have a few hundred potential customers?

Rarely as a primary channel. With a small, nameable market, direct contact is cheaper and faster than earning visibility for queries a handful of people ever type.

Content still matters, but its job changes: it exists so that when a buyer you contacted goes looking for evidence, what they find is specific and forwardable inside their organisation. That is content in support of outbound, not an inbound programme.

Is a marketing qualified lead a useful metric?

Only if the definition is agreed with sales, written down and stable. An MQL is a threshold your own team chose, which means it can always be met by loosening it — something that happens whenever volume targets are missed, and the reason MQL counts often rise while pipeline does not.

Report it beside two harder numbers: how many sales accepted, and how many became qualified opportunities. If acceptance falls while MQLs rise, the definition has drifted.

Do we need marketing automation software to run inbound marketing?

No. You need a way to publish, a way to be contacted, a record of who contacted you, and a person who follows up; a CRM and an email tool cover that. Automation earns its licence when enquiry volume exceeds what a person can track, or when follow-up needs to branch on behaviour.

Buying the platform first is a common and expensive sequencing error. The licence provides workflow, not judgement.