1.How clear is leadership on where AI actually creates value in your go-to-market?

Clarity means a specific, defensible view — not enthusiasm.

No shared view. AI is talked about, but nobody has said where it should matter in GTM.A general belief that AI matters, expressed mostly as efficiency or cost language.A few named areas of interest, but they change with whatever was read or demoed last.A written point of view on two or three GTM areas where AI should create value, and why.An explicit thesis tied to the current GTM constraint, reviewed as evidence changes.

2.Who owns AI in GTM, and how are competing ideas prioritised?

Nobody owns it. Whatever happens is individual initiative.Informally owned by whoever is most interested, alongside their day job.A named owner exists, but prioritisation is first-come or loudest-voice.A named owner and a simple prioritisation rule (impact, effort, risk) applied to a visible list.Clear ownership with budget and a standing review that starts and stops work deliberately.

3.How far has AI moved into the day-to-day GTM work your team actually does?

Essentially not used in the flow of work.Occasional personal use in a side browser tab, invisible to the rest of the team.Used regularly by some people for drafting and research, but outside the systems of record.Embedded in several defined GTM steps with a known place in the process.Embedded across the core motion, including hand-offs, with humans reviewing at defined points.

4.Are AI-assisted GTM workflows reusable, or rebuilt by each person each time?

Reusable means someone else can run it and get comparable output.

Nothing is reusable. Every attempt starts from scratch.A few personal prompts kept in private notes.A shared prompt library exists, but no defined inputs, outputs or quality bar.Documented workflows with named inputs, expected outputs and an owner.Versioned workflows with defined interfaces, review steps and deliberate retirement of old ones.

5.Can AI tools reach the context they need — product, ICP, pricing, past deals, content?

No. Context has to be pasted in manually every time.Only whatever a person remembers to paste, with no consistency.Some curated reference material exists, but it drifts out of date.A maintained context set (positioning, ICP, objections, product facts) is available to the tools that need it.Governed, current context is available through defined integrations, with owners for each source.

6.How reliable is your customer and account data for decisions and automation?

Unreliable. People do not trust the reports enough to act on them.Directionally usable for pipeline totals, but not trusted at account level.Reasonable in the CRM, with known gaps in sources and lifecycle stages.Consistent definitions across the core systems, with clear ownership of key objects.Trusted, monitored data with defined quality checks that fire before decisions are made.

7.How capable is the GTM team at getting reliable results from AI tools?

Little practical capability. Most people have not tried seriously.One or two enthusiasts; everyone else is watching.Broad basic usage, but quality depends heavily on who is doing it.Most of the team can reliably run defined AI-assisted workflows to a known standard.The team improves the workflows themselves and coaches others, with capability built into roles.

8.How consistent is adoption across marketing, sales, and the adjacent functions?

Isolated to one person or none.One function is experimenting; others are unaware.Two functions use AI, but in disconnected ways with no shared hand-offs.Marketing and sales use shared AI-assisted steps with agreed hand-offs.GTM, product, ops and support work from a common operating model and shared workflows.

9.What guardrails exist for AI use — data handling, approved tools, disclosure?

None. People use whatever tool they like with whatever data they have.Verbal caution only, with no written rules and no approved-tool list.Some written rules exist, but they are not consistently known or followed.Clear written rules on approved tools, data that must not be entered, and where disclosure is required.Guardrails are embedded in the workflows themselves and reviewed on a schedule.

10.How do you manage the risk of wrong, unsupported or off-brand AI output?

No review step. Whatever comes out may go out.Ad hoc spot checks when someone happens to notice a problem.Review happens for external-facing work, but the standard is informal.Defined review points, a written quality bar, and a named accountable reviewer per workflow.Review is tiered by risk, failures are logged, and recurring failure modes change the workflow.

11.How do you measure whether AI use is changing GTM outcomes?

Not measured at all.Anecdotes and a general sense that it saves time.Activity and output volume are tracked, but not the outcome that was supposed to move.Defined before/after measures on specific workflows, including quality, not just volume.Impact is measured against the constraint it was meant to relieve, with a stated falsification test.

12.How quickly do you learn from what works and change what you do next?

No feedback loop. Things are tried and quietly abandoned.Occasional retrospective conversation, with no change recorded.Learning happens inside individuals, not in the process.A regular review where workflows are kept, changed or stopped based on evidence.A short, deliberate cycle where evidence routinely reallocates effort and retires weak workflows.

AI GTM Readiness Assessment

Optional context

Strategy & leadership

Workflow integration

Data & systems

People & operating model

Governance & trust

Measurement & learning