GTM signal glossary

A short research document on one company, written for the people who will sell to it: what the company does, why now is a good moment, likely pain points, the key people and a first angle for outreach. A good dossier fits on one screen and shows the source of each claim.

Finding out enough about a company to contact it with something specific to say. It covers the company's situation, recent events, the people involved and how your product fits. It is worth doing in depth only for accounts that fit your ICP and show a reason to talk now.

Ranking companies by how likely they are to buy, so the team works the best ones first. A score usually combines fit (how closely a company matches the ICP) with timing (recent signals). A useful score is explainable: a seller should be able to see why an account ranks where it does.

Selling to a defined set of target companies as units, coordinating outreach to several people at each, rather than processing individual leads. It is the sales side of account-based marketing. It works best with a tight ICP and a reason to approach each account now.

Software that does part of a sales development representative's work with language models: finding accounts, researching them, writing outreach and sometimes handling replies. Products range from assistants that draft for a person to approve, to agents that send on their own. The useful questions are what it sends without a person, and whether you can check what it wrote and why.

A description of a person who takes part in a purchase: their role, goals, problems and how they judge a vendor. An ICP says which companies to target; personas say whom to talk to inside them and what each cares about. One ICP usually has several personas, such as the economic buyer, the user and the technical evaluator.

The people at a company who take part in a purchase decision: the economic buyer who signs, the champion who pushes for it, the users, and the people who can block it, such as security or finance. B2B purchases rarely rest on one person, so outreach that reaches only one of them often stalls.

An event at a specific company that changes the odds it buys something soon, and that you can point to. Funding rounds, hiring in a function, a new leader, a product launch, an acquisition or a move into a new market are common examples.

Two tests separate a signal from noise: the event has to plausibly lead to a purchase you can serve, and you have to be able to show where you saw it and that it was about that company. In face-to-face selling the phrase also means cues from a prospect in conversation, such as asking about price or implementation; in account-based prospecting it means company events.

Adding missing details to a record, such as a work email, title or phone number for a person, or size and industry for a company, usually from data providers. Enrichment results should say how they were checked: a guessed address and a verified one are not the same thing.

Whether your emails reach the inbox rather than spam. It depends on domain authentication (SPF, DKIM and DMARC), the sending domain's reputation, bounce and complaint rates, sending volume and content. Outbound teams usually send from a separate domain so a mistake does not damage the main one.

Descriptive attributes of a company: industry, employee count, revenue range, funding stage, location, ownership and age. They are the first filter in most ICPs because they are easy to find and change slowly. Because they change slowly, they say who could buy, not when.

Signs of interest you observe directly: visits to your pricing page, demo requests, sign-ups, replies, content downloads. It is the most reliable kind of intent because you can see it yourself, but it only covers companies that already found you.

Building the systems behind a sales and marketing motion: data pipelines, enrichment, scoring, routing and automated outreach, usually with a mix of tools and code. A GTM engineer turns a repeatable sales process into software so a small team can run it at a larger scale.

Information that tells a go-to-market team whom to sell to, when and how: company and contact data, buying signals, intent, research and scoring. The term covers both the data and the tools that turn it into a ranked list and a first message.

A description of the kind of company that gets the most value from your product and is most likely to buy it. A useful ICP covers firmographics (industry, size, stage, location), technographics, the team that owns the problem, the events that make a company ready, and explicit exclusions. It describes companies; a buyer persona describes the people inside them.

Data that suggests a company is researching a purchase. It usually comes in two kinds: first-party intent, which is activity on your own properties, and third-party intent, which is aggregated research activity elsewhere on the web. Intent data says someone may be looking; a buying signal such as a round or a hire says something has changed. The two are often used together.

Ranking individual people, rather than companies, by how ready they are for a sales conversation, typically from their fit and their engagement with your marketing. Account scoring asks which companies to work; lead scoring asks which people at them are ready.

Companies that resemble the ones that already bought from you. Ranking prospects by similarity to real customers is a check on an ICP written from intuition: it shows which traits actually predicted a purchase.

Writing each outbound message for its recipient: a reason to write now, a problem they are likely to have, a relevant example. Real personalisation rests on something true and checkable about the company; inserting a name and a company into a template is not personalisation.

A planned series of touches to a prospect over days or weeks: a first email, follow-ups, sometimes a call or a social message. Good sequences stop the moment the prospect replies and give each follow-up something new to say.

How recent a signal is, and so how much it should still count. A round announced last week says more about this quarter's plans than one announced two years ago. Different signals fade at different speeds: a new leader's review window lasts months, a spike of interest may last days.

Choosing which accounts to contact, when, and with what message, from recent events at those accounts rather than from a static list. The signal decides the order of the work and gives the first message its reason. It replaces volume with timing: fewer messages, each with a reason the recipient can recognise.

The technology a company uses: its CRM, website platform, analytics, cloud provider and so on, usually inferred from its website and job posts. Useful when your product replaces, integrates with or depends on a specific tool. A change in the stack can be a signal in its own right.

Research activity observed outside your own properties, typically aggregated from publisher networks and review sites and attributed to companies rather than people. It covers companies that have not found you yet, at the cost of being indirect: a topic spike says someone at a company read about a subject, not who or why.

In outbound, the full list of companies that match your ICP, whether or not they are ready to buy. Building the TAM once and watching it for signals is more efficient than building a new list for every campaign.

An event that creates a reason to contact a company now rather than later. Most buying signals are trigger events; the term stresses timing. A trigger is strongest when it touches the problem you solve: a new head of sales is a trigger for a sales tool, much less so for an accounting one.

A person whose contact details, typically a work email address, have been checked to be deliverable, rather than guessed from a naming pattern. Sending to unverified addresses raises bounce rates, which harms the sending domain's reputation.