Build the systems that bring your go-to-market plans to life.

What has to be decided

• A written definition of a good-fit account that sales and marketing both sign.
• Tiers, the reason for each, and how an account moves between them.
• What is left out (customers, other territories, duplicates, competitors) and who can change it.

Where it goes wrong

A list built once in a spreadsheet, stale by next quarter, owned by nobody.

What runs underneath

• Company records as the unit of the operation, matched on domain, with the fit definition held as properties rather than in a slide.
• Tier and its reason stored on the record, with history, so a change can be explained.
• Exclusion rules that run on every refresh: a new customer leaves the universe without anyone remembering to remove it.

What has to be decided

• Which roles make up the buying group for what you sell.
• How a person is tied to an account, and what happens when they move on.
• Which gaps are worth filling, and from which sources.

Where it goes wrong

Contacts float free of their accounts: a campaign reaches three people at one company and nobody at the next.

What runs underneath

• Contact-to-company associations carrying a role, not a free-text job title.
• Matching and deduplication on every import and integration, so people join the right account once.
• Enrichment that fills named gaps only, with the fields it may overwrite stated and its source kept per field.

Roles in this example: Economic buyer, Champion, Technical evaluator, User lead.

What has to be decided

• Which signals count for you: a hire, a funding round, a technology change, a visit, an enquiry.
• What a signal must show before anyone acts: act, hold, or no meaningful signal.
• Where signals are kept, with source and date, so a decision can be checked later.

Where it goes wrong

Every signal becomes a task, so sales stops reading them.

What runs underneath

• Signals stored against the account with their source, date and how each point is known: observed, or inferred.
• Collection limited to permitted sources, on a schedule, read only.
• An evaluation step that keeps evidence apart from inference, and a person who decides what follows.

What has to be decided

• Which plays exist: sales-led, a marketing programme, or wait.
• Who is eligible for each, and on what basis: consent, an existing relationship, an open deal, a recent touch.
• Which channels each play may use, and how often.

Where it goes wrong

An audience built from a filter nobody can explain, sent to people who opted out or are halfway through a negotiation.

What runs underneath

• Segments defined as rules on properties, versioned, not hand-picked lists.
• Eligibility and suppression checked at the moment of sending, not only when the list was built.
• Audiences synced to the tools that act on them, with a record of what went where.

What has to be decided

• One owner per account, and what marketing may still do there.
• The point at which a marketing account is handed to sales, and how sales accepts it or sends it back.
• Contact limits shared by both teams, so an account is not approached from four directions in one week.

Where it goes wrong

Sales and marketing reach the same buyer on the same day, with different messages.

What runs underneath

• Owner and play on the account record, visible to both teams.
• Handover rules with an acceptance step, a reason when one is returned, and a time limit.
• One contact log across tools, so a frequency limit can actually be enforced.

What has to be decided

• Which outcomes are recorded, and what each one changes: an opt-out, a bounce, a reply, a meeting, silence.
• How the operation is reviewed: which figures, how often, and by whom.
• What counts as evidence of what worked, and what does not.

Where it goes wrong

Results live in each tool’s own report, so nobody can say which accounts moved, or why.

What runs underneath

• Outcomes written back to the account and contact with the stage they evidence: sent is not delivered, and delivered is not read.
• Opt-outs and bounces flow straight into suppression and data repair.
• Reports that read the same fields the operation writes, so the review and the work agree.