Most HubSpot projects fail in discovery — the wrong questions get asked, vague answers get accepted, and the HubSpot architecture inherits every gap. klaryn.ai runs the HubSpot interview the way a senior RevOps architect would: one question at a time, and no answer accepted at face value.
Around 40–60 exchanges, each one building on the last. The HubSpot interview never dumps a questionnaire on you, and it never lets a vague answer slide.
Every turn opens with “what I understood” — a running read-back of your business — then asks exactly one next question. Multiple-choice when the answer has discrete choices, so most turns are one click.
Say “we use spreadsheets for deals” and the follow-up asks which columns, who updates them, and what breaks. That depth is the difference between a generic CRM template and a HubSpot architecture that fits.
The interviewer checks tier gates, feature limits and product release notes against a full index of the HubSpot documentation while you talk — so what it recommends is what HubSpot actually ships today.
As soon as a concrete advantage appears — time saved, a leak closed, a manual step removed — it lands behind the insights icon. Tap it to open. Fit is honest: Strong, Solid, Partial, or Not yet clear.
I don’t have this information parks the question and moves on. Before the ICP, every parked item comes back — answer now, or finish later. The workspace keeps an N open badge so nothing is lost.
Discovery ends with a 14-section Ideal Customer Profile, written from what you actually said. Yours free, whatever you decide next — the strategic foundation the HubSpot blueprint is built on.
Every HubSpot discovery opens with a short intake — who this is for, how technical to be, which HubSpot price list to quote, how many people will sit in the portal, how much AI you actually want, and the monthly budget the licensing plan has to fit.
Set the numbers yourself, answer a few sizing questions, or leave it to the interviewer. What you pick is written into the session, so the chat, the ICP, the blueprint and the runbook all follow it.
Own company, or an agency running discovery for a customer. The interviewer addresses “your business” or “your client's business” from the first question, and never asks whose it is again.
Complete beginner, used HubSpot before, or RevOps pro — jargon density follows you. Pick the language once: the interview, ICP, blueprint and runbook are all produced in it.
Country sets the regional price list — standard global, or HubSpot's new seats-and-credits packaging. An existing portal takes the new list off the table; nothing is quoted against the wrong card.
Headcount, then seats: core for everyone in HubSpot day-to-day, plus sales, service or Front Office where the edition allows. The licensing plan is built from real roles, not a round number.
Monthly HubSpot budget, marketing contacts, email volume, CRM records. Known figures go in now; otherwise a few ranges, or “you decide” — and the first-year TCO is designed inside that appetite.
Breeze credits are metered. Say none-to-start, light assistants, a couple of agents, or heavy across sales, service and content — so the plan does not prescribe AI the budget and readiness cannot support.
You are not filling a form in the dark. As you answer, the session researches your stack, notes where HubSpot would actually help, and lets you skip or sample anything you would rather not share yet.
A live panel accumulates concrete advantages for this business, plus an honest HubSpot-fit read. Not a scorecard of the interview — the value of the hour, surfaced as it emerges.
Name a tool and it is researched against the HubSpot Marketplace. Cards show Native, Data Sync, Middleware or Custom, certified status, and a marketplace link — so the architecture is grounded in what actually exists.
Revenue, margins, volumes, pricing and salaries are optional. Use an example figure and the interview substitutes a labelled sample you can replace after download. The rest of the interview is still about your real business.
I don't have this information parks the question and moves on. Parked items come back before the ICP — answer them now, or finish discovery later. The workspace keeps an "N open" badge so nothing is lost.
Most discovery interviews extract. This one also returns. As soon as a concrete advantage appears — time saved, a leak closed, a manual step removed — it lands behind the insights icon, tied to what you just said.
Fit is honest, not a sales pitch: Strong, Solid, Partial, or Not yet clear. Everything behind that icon feeds your free ICP.
After intake, the first turn is a choice, not a questionnaire. Talk, share what you already have, or let the interviewer read your public site first.
One question at a time, with multiple-choice when the answer has discrete choices. No HubSpot knowledge required — the interview asks about your business.
PDF, CSV, TXT or MD — process notes, CRM exports, an existing ICP. The interview maps each file to the phases it informs, confirms before using it, and skips the opening menu.
The interviewer reads the public site once, then pressure-tests what it found with you. It is briefing, not assumed fact — you still confirm what is true.
The order matters: the interview earns the right to ask HubSpot-technical questions by understanding your business first. Architecture questions asked cold produce generic answers — the same questions asked after the ICP produce your HubSpot architecture.
Progress is tracked live in a sidebar, phase by phase, and you can pause and resume any time — the session checkpoints itself.
Start free →When you can't answer something, tap I don't have this information. The interview parks it as an open item and moves on — a missing volume figure or a legacy system detail never stalls the hour.
Before your ICP is written, every parked item comes back. Answer them now, one at a time, or Finish discovery later. The workspace shows an N open badge and a resume path so you can pick the thread back up. Unresolved items are carried into the blueprint as clearly-labelled assumptions, never silent guesses.
HubSpot discovery ends with a document, not a transcript. The structure is fixed — no section is dropped because an answer was thin. Every section is written from what you actually said, and it is the strategic foundation the HubSpot blueprint is built on: Green/Yellow/Red → scoring, buying triggers → lifecycle, exclusions → suppression. Useful on its own even if you never buy anything.
An archetype label your team can remember, and a single sentence that states who the ideal customer is, what triggers them, and why they are a great fit. This is the line a CEO, a new hire, and a HubSpot scoring rule can all share.
Feeds blueprint §0 Engagement Context · §1 Executive Summary
One paragraph in the customer's voice — who they are, what they are trying to accomplish, why they need this now, and what makes them a great fit. Written in their language, not product-marketing language, so HubSpot nurture copy and Breeze context start from how they actually talk.
Feeds blueprint §7 nurture messaging · §20 Breeze context seeded from this engagement
A firmographic table: industry and vertical, employee count, revenue range, geography, tech stack, budget range, and business stage. These are attributes a salesperson can verify before a conversation — not budget or urgency, which only surface mid-call. They set the data-model anchor (company-centric vs contact-centric) and the geography that drives HubSpot routing, localisation and compliance.
Feeds blueprint §3 Data Model · §2 licensing geography · §9 Integrations starting point
The primary pains in their words, the events that make them start looking, and the emotional blocker — the underlying fear that makes them hesitate. Buying triggers become HubSpot lifecycle entry criteria; the emotional blocker becomes the nurture sequence that actually lands, instead of a generic drip.
Feeds blueprint §5 Lifecycle entry criteria · §7 nurture automation
How they research, who sits on the buying committee (titles, roles, vetoes), how they measure success, the minimum involvement required to succeed, and the content lane they actually trust. Decision-committee structure drives HubSpot association labels and contact roles; the content lane drives forms, subscription types and what you gate.
Feeds blueprint §3 associations and contact roles · §8 Marketing Assets
The three outcomes they will pay for — ranked, specific, and rooted in product success rather than who was easiest to close. These become the six-month success criteria the HubSpot blueprint has to trace back to, so the architecture is measured against what this customer actually values.
Feeds blueprint §1 six-month success criteria · §11 reporting that proves them
A table of observable signals, what each one means, and the recommended action. Growing headcount, a churned anchor customer, a renewal season they cannot report on — each signal maps to a HubSpot engagement score, a saved view, or a sales play, so the team does not treat every form fill as equal.
Feeds blueprint §4 Engagement scoring · sales View-Gate saved views
Accounts classified by firmographic fit a salesperson can verify in the first five minutes. Green is who you built this for — proactive outbound, all channels. Yellow is where you experiment — engage inbound, do not pursue cold. Red means do not sell to them, even at an attractive deal size. An ICP without a Red tier is a wishlist; this one is explicit. Fit scoring in HubSpot is graded from Green-tier attributes, not a generic "filled in a form."
Feeds blueprint §4 Fit scoring and MQL threshold · routing priority
Who is out, named, plus the churn pattern to watch — the most common reason a good-looking prospect does not work out. Rooted in highest-LTV customers, not fastest-closing ones. Exclusions become HubSpot disqualification properties and suppression lists; the churn pattern becomes negative scoring and the service design that stops the leak repeating.
Feeds blueprint §4 negative scores · §8 suppression lists · §6 service / CS pipeline
Who else they consider, who those alternatives actually serve, and why this ICP prefers you. Written as a positioning table, not a feature dump — so sales knows when to lean in, when to walk, and which objection is really a Red-tier mismatch dressed up as "we're looking at X."
Feeds sales enablement and outbound messaging; not a HubSpot object on its own
What actually informed this ICP, with facts you stated kept separate from what was inferred. Every inferred field is listed — never silently merged into client-stated facts. Recommended next steps name the gaps that still need a customer interview or a source-system check before they become HubSpot baselines.
Feeds labelled assumptions in the HubSpot blueprint — never silent guesses
What Marketing, Sales and Customer Success each do with this document on Monday. Targeting, messaging and content tied to the emotional trigger and content lane. Outbound prioritisation, Green/Yellow/Red disqualification, and intent-signal responses. Onboarding triage by tier, expansion triggers, and churn prevention tied to the named churn pattern — so the ICP is an operating system, not a slide.
Feeds blueprint §7 workflows · §16 enablement · how each team uses HubSpot day one
A named owner, method and cadence for keeping the ICP honest: quarterly review with sales and CS, bi-annual customer survey, annual LTV cohort re-analysis. Any figure you chose not to disclose is listed under sample data used — labelled as a placeholder to replace after download, never treated as a real baseline. Unresolved discovery items land here as things to confirm, with a source, not as invented numbers.
Feeds blueprint §18 post-launch milestones · sample figures stay labelled, never silently adopted
Version, date, named owner, and next review date. Someone on your team owns this document after the interview ends — so the profile that drives HubSpot scoring and routing has a human responsible for the next quarterly check, not a file that goes stale in a folder.
Feeds blueprint §14 RACI — who owns the ICP after go-live
Your ICP and the full HubSpot discovery transcript feed the blueprint, then the runbook. Same HubSpot engagement, four deliverables, in this order.
The Ironwood Field Services ICP, unedited — every section from summary through validation schedule. Use the contents list to jump; scroll to read the whole document.
Loading the full document…
Same generator as a live engagement. Figures are discovery inputs, labelled for confirmation.
The AI-led HubSpot discovery interview and your Ideal Customer Profile are free — no credit card, no sales call. Pay only if you want the HubSpot blueprint, or Blueprint + Runbook together. The runbook is not sold on its own.
