Tools Overview
What It's For
The tools surface holds guided workflows for the work that lands between sales and delivery: scoping a portal, pricing a build, drafting a proposal, pushing tasks into a project plan. Most tools are wizards: you supply the facts only you know, the wizard does the arithmetic and the writing, and you decide what comes out before it goes anywhere. The HubSpot Builder is different: it lets an AI assistant change a registered client portal directly, as described below.
Which Tool Do I Want
| The job you have | The tool |
|---|---|
| A partner sent a half-formed brief and I need a proposal | Proposal Generator |
| Price moving a client off a legacy CRM, marketing tool, help desk or other platform | Migration Quoting |
| Price onboarding a client onto HubSpot hubs | HubSpot Onboarding |
| Price building out a HubSpot portal in detail | HubSpot Implementation |
| Price connecting HubSpot to another system | Integration Wizard |
| Find out what is wrong with a HubSpot portal | Portal Audit |
| Grade a live website and hand a client a fix list | Website Audit |
| Show a client how they rank and whether AI engines recommend them | SEO / GEO / AEO Audit |
| Price a HubSpot CMS website build or migration | Website Build |
| Turn audit findings into a scoped remediation estimate | Audit Remediation |
| Turn a signed SOW into a project plan with tasks | ClickUp Plan Builder |
| Connect a client's HubSpot portal so an AI assistant can build in it | HubSpot Builder |
| Change the defaults every wizard starts from, edit prompts, or see AI cost | Tools Settings |
The tools feed each other. A portal audit hands its findings to the Remediation Wizard. Every wizard's saved estimate can be loaded straight into the ClickUp Plan Builder without retyping anything.
What the AI Does and Does Not Do
Across every tool the division is the same, and it is worth internalizing because it explains most of what you will see.
Arithmetic is never the model's job. Hours, multipliers, overhead, totals, scores, and timelines are computed in ordinary code from a catalog and the numbers you enter. Run the same inputs twice and you get the same figures, every time.
Judgment is the model's job. Which questions still need answering. What a failed check means for the business. How to describe a scope in prose a client will read. Which of two integration approaches fits. That work is genuinely a model's, and it is grounded in data it was not allowed to invent.
You are the gate, with one exception. Every wizard ends in a document or a plan you can edit, regenerate, or throw away, and nothing from a wizard reaches a client without a person deciding it should. Where a wizard does write somewhere, it says so: the ClickUp Plan Builder creates real tasks when you push, and the Proposal Generator's Send to PandaDoc creates a draft document in PandaDoc that you open and send from there. The exception is the HubSpot Builder: it writes directly into a client's HubSpot portal the moment the assistant makes a request, with no review step and no forced draft state.
A model's number becomes an input to the arithmetic in only three places, and each is fenced. The Proposal Generator asks a model for a low and a high base-hour figure per line, because a pasted brief has no catalog to price from; you review and edit every line on the Estimation step. The Website Build asks a model to estimate custom features you described in plain English, then makes you review and confirm every one before it counts. The Audit Remediation asks a model to estimate findings its catalog could not match, labels each such line "AI" so you can see which numbers came from where, and holds those hours between a floor and a ceiling that an administrator sets.
The Estimate Spine
Five tools price work the same way: the Migration, Onboarding, Implementation, Integration, and Website Build wizards. Understanding the spine once means you understand all five.
Base hours come from a catalog, not from a model. Each tool carries a catalog of the work it knows how to price: more than 160 configurable items for a HubSpot implementation, a feature list per source platform for a migration, a deliverable list per hub and tier for onboarding. You pick what is in scope, you enter the counts, and the base hours follow.
Skill tier prices those hours; it never changes them. Every line item carries a tier, and every tier a billing multiplier that applies only when a quote includes Tier Time (see below). Base hours are the effort figure whoever does the work:
| Tier | What it means | Multiplier |
|---|---|---|
| T1, Standard | Routine configuration any HubSpot admin could do | 1.0x |
| T2, Advanced | Needs HubSpot expertise or cross-system knowledge | 1.5x |
| T3, Expert | Needs developer skills: custom code, APIs, deep integration | 1.75x |
You can change any line's tier and the totals recalculate live.
Overhead is added on top, and it differs per tool. These are the defaults, and every one of them is adjustable in the estimate's Adjust panel:
| Tool | Testing | Contingency | Project management |
|---|---|---|---|
| Migration Quoting | 5% | 5% | none |
| HubSpot Onboarding | none | none | none |
| HubSpot Implementation | 5% | 5% | 10% |
| Integration | 5% | 5% | 10% |
| Website Build | none | 5% | 5% |
If a wizard shows a percentage that differs from this table, someone has adjusted it for that quote or an administrator has set a different default in Tools Settings. The number on your screen is the one being used. Website Build's two rates are the exception: no settings screen changes them, so they can be changed only per quote. The Audit Remediation wizard follows the same rule and is described on its own page.
The timeline is a capacity band, not a confidence interval. A quote reading "5 to 7 weeks" does not mean the work might take five weeks or might take seven. It means: at the full working pace, five; at 70% utilization, which is what real weeks look like once meetings and other clients intrude, seven. The hours figure itself is a single number, not a range.
Complexity is a label for the total. Up to 40 hours is Simple, up to 120 Moderate, up to 300 Complex, up to 600 Very Complex, and beyond that Enterprise.
Only the narrative is written by a model. Once the numbers are settled they are handed to the model as fixed facts it is forbidden to recalculate, along with a rule that it must never show its arithmetic. It writes the scope, the assumptions, the risks, and the phasing around numbers it cannot touch.
One tool sits outside this spine. The Proposal Generator asks a model for base hours because at proposal time there is no catalog to look them up in: the brief is a paragraph of prose. It is also the only tool that produces a low-to-high range on hours rather than a single figure.
Where an estimate wizard offers Branded Output on step 1, turning it on adds a Branded PDF button next to Print PDF, Copy Markdown and Download XLSX. The branded PDF carries your agency's name and print styling.
Tier Time
When your instance's Skill Tiers setting (in Tools Settings, under Global Defaults) is on, step 1 of every wizard in the spine above, plus the Audit Remediation wizard and the Proposal Generator, asks you to choose Tier Time: "Show Effort/Base Hours Only" or "Include Tier Time". Effort-only is the default for a new quote. When the global Skill Tiers setting is off, this choice is hidden everywhere and every quote is effort-only.
This is a per-quote choice, not a live setting. Flipping it re-prices the same quote in place: base hours do not change, the tier columns appear or disappear, any inline edits you made survive the switch, and the choice you land on is saved with the quote. Everything downstream, exports, the SOW, the narrative, and the copy pushed to ClickUp or PandaDoc, follows that quote's own frozen tier choice, not whatever the global setting happens to be set to later. ClickUp always receives base hours regardless of a quote's tier choice; see the ClickUp Plan Builder for why.
AI Acceleration
A separate instance setting, also in Tools Settings under Global Defaults and off by default, controls a second per-quote choice: Delivery Model, "Manual Delivery" or "AI-Assisted Delivery". It appears on step 1 of the Implementation, Onboarding, Integration, Website Build, and Audit Remediation wizards only, never on Migration or the Proposal Generator, and only when the instance setting is on. Off, the feature is invisible everywhere.
On an AI-assisted quote, catalog lines that can be accelerated show the manual hours struck through beside the AI-assisted hours, each with its own "AI" checkbox so you can return any single line to manual. A line you edit by hand stays manual and its checkbox disables, labeled "Edited lines stay manual". Judgment work, QA and testing, training delivery, and project management are never accelerated. Acceleration is also evidence-gated: only work classes where delivery data shows AI speeding the work up carry a discount at all (integration and data-sync work, pipeline configuration). Everything else keeps its manual hours even on an AI-assisted quote. The estimate summary, the exports, and the SOW show the net effect as its own "AI Acceleration Adjustment" line before the Grand Total; ClickUp receives the AI-assisted hours actually being delivered. A manual quote carries no AI language anywhere in its documents.
The Document Steps
Every wizard that produces a document offers Regenerate and Humanize, and most offer Client Version. They mean the same thing everywhere.
Regenerate runs the writing pass again from the same numbers. Use it when the prose is wrong, not when the estimate is.
Humanize rewrites the prose to read less like a machine wrote it. It is explicitly forbidden from changing any number, table, or fact, and it only ever touches style.
Client Version strips the internal layer: the tier columns, the multipliers, the variance commentary, and any internal risk notes. It keeps every scope description, subtotal, and grand total, so the client's numbers match yours exactly. What you send a client should almost always be this version. It appears only when the quote was priced with Tier Time on. With Tier Time off, the document is already written client-style and there is no separate client version to switch to.
Who Uses It
Anyone on your team who can sign in to the tools uses the wizards, usually whoever scopes, prices and plans client work between sales and delivery. Administrators change the defaults, catalogs and prompts in Tools Settings.
Related
- Where You Work: how the tools fit beside the app, the Operator Console and Slack.
- Tools Settings: the defaults, catalogs and prompts behind every wizard.
- Glossary: plain-English definitions of the terms used here.