Short answer: AI automation is being sold aggressively right now, and not every provider can actually deliver a working, safe, owned system — some are reselling a generic tool with a markup, others are building something you can’t take with you if the relationship ends. These ten questions separate a genuine automation partner from a reseller, before you sign anything.

1. What exactly will this system do — and what won’t it do?

Ask for a plain-English description of the specific tasks the automation or AI agent will handle, and just as importantly, what it won’t touch. A provider who can’t draw that boundary clearly hasn’t scoped the project properly yet, which usually means scope creep — and cost creep — later. A specialist should be able to put that scope back to you in writing before any build starts, from a single discovery conversation about your actual tools and workflows, not a lengthy back-and-forth.

2. Is this a fixed automation, or an AI agent making judgement calls?

These solve different problems and carry different risks. A fixed automation (a workflow trigger between your CRM and your invoicing platform, for example) does the same thing every time and is easy to audit. An AI agent makes a first-pass judgement call on things that don’t fit a template — drafting a reply, triaging a request — which means it needs guardrails, not just a brief. Get a straight answer on which one you’re actually buying; a provider who’s vague about the distinction is a warning sign. (See our full breakdown: AI Agents vs Marketing Automation: What’s the Difference?)

3. Where does the system live, and who can see the data flowing through it?

Any automation touching customer data, leads, or financial records is a data-security question first and a marketing question second. Ask where it’s hosted, what security certifications the hosting meets, and whether your data is ever used to train anything outside your own account. Sharp Instincts’ own infrastructure runs on Amazon/Google Cloud servers hosted in Sydney, Australia, meeting ISO 27001/27017/27018, GDPR, PCI DSS Level 1, and IRAP standards, with real-time intrusion protection and daily backups — a reasonable baseline to compare any provider against.

4. Who owns the system once it’s built — can you take it with you?

If the relationship ends, do you keep the workflow, the automation logic, and your data — or does it all belong to the provider’s platform? This should be answered in writing before you start, not discovered on the way out. Sharp Instincts’ standing terms: a client’s work-in-progress materials are retained for three months after the relationship ends, available on request within that window, then deleted — automation documentation and logic handed over as part of offboarding, the same policy that applies across every Sharp Instincts engagement.

5. What happens when the automation gets something wrong?

No automation or AI agent is error-proof. Ask specifically what the human-in-the-loop safety net looks like — who gets notified, how fast, and what the rollback process is — rather than accepting “it’s very accurate” as the whole answer.

6. Are you locked into a contract, or can you leave if it’s not working?

Ask about minimum terms, notice periods, and exit costs up front. Sharp Instincts works without contracts, lock-in periods, or notice periods on its services generally, automation included — clients who’ve stayed 10+ years have done so because the work keeps earning it, not because they’re bound to.

7. How is pricing actually structured?

Automation and AI-agent pricing can be a flat build fee, an ongoing retainer, a usage-based model where cost scales with volume, or some mix of the three — and providers aren’t always upfront about which one you’re getting into. At Sharp Instincts, GEO and AI-automation work is already included within the existing Growth ($3,000–$5,000/month) and Authority ($5,000+/month) pricing tiers — if you’re spending $2,000/month or more, automation shouldn’t appear as a separate line item. On a lower-spend plan (Sharp Instincts’ Foundation tier, under $2,000/month), it’s billed as an additional cost on top of the existing SEO package. Either way, ask any provider directly which bracket your spend falls into before assuming automation is included or extra.

8. What’s the realistic timeline from signing to something actually running?

“A few weeks” and “a few months” are very different experiences when you’re waiting on a system to start saving you time. Ask for a specific first-working-version date, not just a project-kickoff date. Sharp Instincts’ standard onboarding across its services runs around two weeks from first contact to work starting, via a single shared-screen briefing call — an automation build’s actual first-working-version date depends on how many systems it touches, and a specialist should scope that specifically rather than assume it matches the general onboarding timeline.

9. Who’s your actual point of contact once it’s live?

Ask whether you get a dedicated account manager or get routed through a general support queue once the initial build is done. Sharp Instincts’ standard model is a single dedicated account manager as the point of contact, backed by a monthly reporting dashboard and a monthly catch-up call — worth confirming the same structure applies to an AI-automation engagement specifically.

10. Can you show me a real, live example — not just a demo?

This is the hardest question for the AI-automation space right now, and it’s fair to ask it plainly. Be wary of a provider who can only show a slideshow. In Sharp Instincts’ own case, directly: several client AI-automation projects are in progress but not yet public as case studies — real results are expected to be shareable by the end of 2026. What is live today is Sharp Instincts’ own internal AI tooling, built and used in its own operations — Sharp Instincts Sage and the Sharp Connect MCP Automations system — which a prospective client can ask to see directly as evidence of real, working build capability, even ahead of named client case studies.

Why this matters beyond the checklist

AI automation and AI agents are being sold by everyone from specialist automation shops to generalist marketing agencies to pure software platforms, and the quality gap between providers is wide. Treat the above as a floor, not a complete vetting process: a provider who answers all ten clearly and specifically is worth a real conversation; one who answers vaguely on more than two or three is worth a pause. For the deeper technical and strategic picture — what agentic automation actually is, deployment phases, and governance approach — see Sharp Instincts’ full guide: Business Automation & Agentic Automation: A Definitive Guide. If you’re still working out whether you need a fixed automation or a judgement-making AI agent in the first place, or you want the broader evaluation picture before narrowing to these ten questions, start with How to Choose an AI Automation Agency in Australia and AI Agents vs Marketing Automation: What’s the Difference?

Frequently asked questions

What’s the most important question to ask before hiring an AI automation partner?
Whether you’ll own the system and the data once it’s built — if a provider can’t answer that clearly in writing, the rest of the pricing and timeline conversation matters less than it seems to.

What’s the difference between an AI agent and a simple automation?
A fixed automation runs the same rule every time with no judgement involved; an AI agent makes a first-pass judgement call on things that don’t fit a template, like triaging or drafting a reply. They solve different problems and need different kinds of oversight.

Should I expect a contract or lock-in period for an AI automation build?
Not necessarily — some providers, including Sharp Instincts, work without contracts, minimum terms, or lock-in periods. Always confirm this specifically rather than assuming it applies industry-wide.

How long does it typically take to get an AI automation system running?
It varies by build complexity and how many systems it touches, so ask for a specific first-working-version date rather than just a start date — a specialist should be able to scope that once they understand your actual process.

Can I see real client results from an AI automation provider before I commit?
A credible provider should be able to show something real, even if formal case studies aren’t public yet. Sharp Instincts’ own internal AI tooling — Sharp Instincts Sage and Sharp Connect MCP Automations — is one example of a live, working system a prospective client can ask to see directly, while its own client AI-automation case studies are expected by the end of 2026.