Skip to content
BH TechnologyROI FOR AI Book a consultation
← Back to the guide

Capacity / One person

A routine that prepares. A person who decides.

One person can start with their own tools, files and a repeatable routine. Assistance prepares the work; the owner checks it before permitted steps are completed.

Approved files pass through an AI-prepared draft, a person’s review and permitted execution. Corrections loop back to the next run.
Illustrative reconstruction of the work. The picture does not establish a result.
  1. Approved inputs

    Use the right customer context, pricing, scope and examples.

  2. A human decision

    Review, corrections and approval stay inside the work.

  3. A routine that improves

    Schedule the preparation and review what the next run should do better.

Same accepted work. Review and upkeep included.
Before30 hours
With the routine20 hours

10 hours released on the same batch30 ÷ 20 = 1.5× output per hour (+50%). Time reduction and productivity increase are different measures.

Hypothetical batch, not a whole workweek, a payroll cut or a forecast. Accepted quality must remain comparable.

Open the detail

Keep the setup, assumptions and checks in view.

What could improve when one person gets practical AI support?

What could improve when one person gets practical AI support?

This example focuses on repeated work, while judgment stays with the person responsible. Here is one standard sales proposal, before and after.

See what changes
Moves
Searching, organizing and first drafts go to an assistant.
Stays
You decide prices, scope, customer fit and what gets sent.
Repeats
A schedule finds the work. A review keeps improving it.
One person, their own desk and their usual work.
Hypothetical workflow · the same standard sales proposal

Before: you do every step by hand

  1. You     Find information

    Search emails, notes, pricing, scope documents and customer history.

  2. You     Prepare the proposal

    Build the scope, confirm prices, write the explanation and format the document.

  3. You     Review and send

    Check accuracy, adjust judgment calls, approve the final version and send it.

  4. You     Next proposal

    Starts the same way, from the beginning.

After: the routine prepares, you decide, and approved steps get finished

  1. Recurring schedule     Find the work

    A scheduled routine lists new requests, missing information, follow-ups and proposals needing attention.

  2. Assistance     Prepare a ready-to-check pack

    It organizes what is relevant and drafts the scope, explanation, checklist and questions.

  3. Human check     Decide what is correct

    You confirm prices, scope, customer fit, risks and wording, then approve.

  4. Execution     Complete permitted steps

    After approval, a connected workflow may file it, update a record or start the next task. This needs that setup and permission.

  5. Recurring review     Improve the routine

    Review errors, missed information and unnecessary steps. Update the routine so the next proposal starts better.

Back to the first step: the next scheduled run uses the improved routine.
10 hours saved and +50% productivity: how the same work gives bothWorked example, with review and upkeep included
Hours for the same batch of workScale: 0 to 30 hours
  • Today30 h
  • With the routine20 h

10 hours released. The striped section is the time back.

Output per hour on that workScale: 0 to 1.5x today’s rate
  • Today1.0x
  • With the routine1.5x

30 ÷ 20 = 1.5x, or +50%.

The same change reads two ways. The work takes 33% less time, and each hour produces 50% more. Those are different measures of one change, so 33% and 50% are not interchangeable.

This assumes comparable accepted work. The 30 hours are one batch of proposal work, not a 40-hour week. It is an illustrative model, not a promise or a measured result.

Assistance prepares. Execution completes. You decide in between.

  1. Assistance
    Gives you a draft

    It organizes, summarizes, drafts and lists questions. What you get is something to check and use.

  2. You
    Approve or change it

    Prices, scope, customer fit, risks and final wording remain your call.

  3. Execution
    Does an allowed step

    It files the proposal, updates a record or opens the next task. This needs a real connection, permission for that step and a log you can check. Not every product does all of this.

What a schedule adds: it sets when the routine runs, such as every weekday morning. It repeats the check and the preparation. It does not decide what is correct. Automation links steps so you don’t have to restart each one, and your approval stays in the chain where it matters.
  • Straightforward tools
  • Relevant context
  • A routine you repeat

Together, these can produce strong gains. Buying tools covers only the first part.

Try it on one routine
The routine
Standard proposals, from new request to approved draft.
Input you can check
Approved price list, scope template and one recent accepted proposal.
Output you can check
A pack with the draft, the sources it used and its open questions.
What to measure
Over the next ten: minutes per accepted proposal including review, corrections per draft and missed information. Time a few of the same kind first.
What happens next
Fix the most common miss, put the routine on a schedule and run it again.

Three different results

  1. Hours releasedTime back on the same work.
  2. More accepted outputDepends on useful redeployment, demand and accepted quality.
  3. Financial returnAlso depends on delivered work and total costs. Collected cash is a separate check.

What this person needs

How much could one narrow task change?An assumed 3x to 10x example for one task, not a forecast or a whole week
Assumed minutes for one standard follow-up note from approved inputs, including reviewScale: 0 to 60 minutes
  • By hand60 min
  • Assisted20 min (3x)
  • Automated routine6 min (10x)
Gains can be large when the work is repeated, structured, information-heavy and easy to check.
Gains are smaller when every case is new or needs deep judgment.

These minutes are assumptions, not a rate or a forecast, and they need testing on your own work. A week holds many other tasks, so a task multiple does not carry over to the week. The main illustration stays at 10 hours and +50%.

Worked proposal setup, tools and costsWho does what, one deal start to finish, monthly cost and sources

Overview

One person can often move first. They usually start with the tools, files and routines they already use, which makes setup faster, cheaper and easier to control.

In this illustrative example, that person saves 10 hours a week and shows a 50% productivity increase. Those are two separate measures. Neither is a promise, and ten hours don’t turn into cash on their own. They are usable capacity, and the next question is what the person does with it.

Example values for one person who prepares standard sales proposals on their own accounts. A person approves every proposal before it goes out.

Hypothetical example · same weekly workload

  • 30 hToday
  • 20 hWith the setup
  • 10 hBack each week

One person completes their usual batch of work in fewer hours, with review and upkeep included.

How one proposal moves

  1. Deal in HubSpotProposal stage reached
  2. Approved sourceTemplate, pricing and examples in Drive
  3. Agent draftsGrok Bot drafts and organizes recurring steps
  4. Person checksClaude compares the draft with approved examples
  5. Person approvesOwner signs off and sends
  6. RecordedNote and file link saved in HubSpot

Who does what each day

  • Grok BotReads the approved queue or an exported deal and drafts proposals and recurring steps
  • ClaudeWrites longer pieces and reviews drafts against approved examples the owner provides
  • CodexBuilds and tests small scripts and reports in VS Code or a repo
  • HubSpotHolds deals, tasks and customer status
  • DriveHolds approved docs and working files

One deal, start to finish

  1. A deal in HubSpot reaches the proposal stage. The owner exports the deal details or adds them to the approved queue.
  2. Grok Bot reads that entry plus the approved template and pricing from Drive, then drafts the proposal and a checklist of recurring steps.
  3. The owner hands Claude the draft and approved examples as files. Claude flags gaps and differences. The owner checks facts, scope, pricing and any exceptions.
  4. The owner approves and sends the proposal. The note, file link and next task go back into HubSpot.
  5. When the routine needs a fix, the owner uses Codex in VS Code to update the script and tests it before the next run.

The owner stays in charge. They hold every account, review each draft and approve what goes out. Tools hand work over by file, link or repo commit.

Entry $30 Grok Premium + $20 ChatGPT Plus + $20 Claude Pro = $70 a month. Heavier tiers $300. New Workspace adds $16.80. CRM and extra usage are separate.

USD before tax. Sources xAI, ChatGPT, Claude.

Go deeper

Their own tools

Their own tools

What has to work together for useful work to improve? Five jobs around one decision-maker. AI supports the workflow. It is not the workflow.

Example setup · open a role to see its limits and where you decide
  • Approved sourceHolds what the work may use

    In this example: Google Workspace holds approved docs, email and working files.

    May
    Give the routine current, approved material to read.
    May not
    Stand in for customer status. That stays in the record system.
    You decide
    Which files count as approved. Each one has an owner, a version and a review date.
  • Draft and assistPrepares a first version

    In this example: Grok Bot runs as a recurring agent, with weekday follow-up drafts and a morning brief.

    May
    Organize information and prepare a draft to check.
    May not
    Auto-send, change pricing or edit deal stages.
    You decide
    Whether the draft is right and whether it goes out.
  • ReviewChecks against good examples

    In this example: Claude writes, analyzes and reviews against approved examples. One sample task is a client brief from dated, authorized files.

    May
    Point out gaps and compare a draft with the quality bar.
    May not
    Replace your approval. A review pass is still input to you.
    You decide
    Which points to accept and what changes.
  • You approveEvery role hands work to one person, who checks it and sends it

    Why you sit in the middle: in this example each tool runs on its own account. You pass work between them by file, link or repo commit.

    A decision that stays yours: a proposal draft can be prepared automatically, but you still check scope and prices before sending.

    This keeps one person’s setup simple. Connections between tools need permissions and upkeep, which is the department’s harder job.

  • Record systemKeeps status, tasks and next-action dates

    In this example: HubSpot holds deals, tasks and next-action dates, and every send gets a HubSpot note.

    May
    Show what happened on a deal and what comes next.
    May not
    Be changed by the agent at the deal-stage level in this setup.
    You decide
    When a deal moves and what the next action is.
  • BuildMakes the small tools the routine needs

    In this example: Codex builds and tests a small script, report or internal tool in VS Code or a repo. One week, it tests a script that fixes dates and removes duplicate contacts, then Git commits it with history.

    May
    Build and test a small, named piece of work.
    May not
    Decide what goes to a client.
    You decide
    You check the tested result before relying on it.

What to take from this: choose tools by the job they do in your routine. The routine matters more than the logo.

Named tools, one example week and monthly licensesWho does what, how work passes, license tiers and pricing sources

Their own tools

Who does what

  1. Grok BotPrepares
  2. ClaudeReviews
  3. OwnerApproves
  4. HubSpotLogs
Grok Bot
Recurring agent. Weekday follow-up drafts and a morning brief.
Codex
Builds and tests a small script, report or internal tool in VS Code or a repo.
Claude
Writes, analyzes and reviews against approved examples.
HubSpot
Deals, tasks and next-action dates.
Google Workspace
Approved docs, email and working files.

How work passes

In this example each tool runs on its own account. The owner passes work by file, link or repo commit.

One example week

  1. Codex in VS CodeTests a script that fixes dates and removes duplicate contacts
  2. GitCommits the tested script with history
  3. ClaudeDrafts and reviews a client brief from dated, authorized files
  4. OwnerChecks both, then sends

Monthly licenses

Illustrative planning prices in USD a month before tax, checked Nov 4, 2026. Plans, terms, usage and regional pricing change, so check current plans before relying on them.

Entry tiers
Grok Premium $30 + ChatGPT Plus $20 + Claude Pro $20 = $70
Heavier tiers
Grok Super $100 + ChatGPT Pro $100 tier + Claude Max $100 = $300
Optional
New Google Workspace Business Standard adds $16.80, so entry becomes $86.80. VS Code editor $0. CRM and extra usage are separate.
Their own files

Their own files

How do your files become usable work instead of a pile of documents? The routine reads approved files, writes only to a draft area, and you check what leaves.

Sample folder setup · open a file or zone for its rule

  1. Approved inputsAgent: read only

    Drive / Sales-Approved

    • followup-templates-v3.docxTone and structure

      Reviewed Oct 2026. It gives each draft its structure and voice, so you correct facts and judgment instead of rewriting from a blank page.

    • pricing-reference-2026-10.pdfCurrent prices

      Recheck monthly. Stale pricing is replaced, never edited in place, so an old version cannot quietly change.

    • good-examples-v2.mdQuality bar for review

      Accepted examples give the review step something concrete to compare against. Like every approved file, this one has an owner, a version and a review date.

  2. Working draftsAgent: can write

    • Drafts /Unfinished until you check it

      The one writable document folder in this example. A proposal might use the approved pricing and template, with the draft created here. Nothing in this folder has been accepted yet.

    • repo / scripts / weekly-report.pyCode, Git versioned

      Code lives in a separate repo with history, apart from documents. A bad change to a script can be traced and undone without touching approved files.

  3. Checked outputYou: approve

    • Owner checks, then sendsYour version is the one that counts

      The draft becomes accepted work only after you check it. If it later becomes a model for other work, it joins the approved folder as a new version with an owner and review date.

    • HubSpot noteThe record of what happened

      Every send gets a HubSpot note. HubSpot stays the record for customer status, so that status is not scattered across folders.

Outside the routine: Clients-Restricted / and personal foldersNo agent access

The owner’s personal folders and Clients-Restricted stay unshared with agent accounts, and restricted client data stays out of agent folders.

Having access to a file is not permission to use every piece of information inside it.

Check your own setup: can you name who owns each approved file, which version is current and when it is next reviewed? If not, start there.

Complete sample folder and file rulesThe full tree, its rules and example permissions

Their own files

Sample folder

  • Drive / Sales-Approved
  • followup-templates-v3.docxTone and structureReviewed Oct 2026
  • pricing-reference-2026-10.pdfCurrent pricesRecheck monthly
  • good-examples-v2.mdQuality bar for review
  • Drafts /Agents can write
  • repo / scripts / weekly-report.pyGit versioned
  • Clients-Restricted /No agent access

Rules for files

  • Every approved file has an owner, a version and a review date.
  • Stale pricing is replaced, never edited in place.
  • Code lives in a separate repo with history, apart from documents.
  • Restricted client data stays out of agent folders.

Example permissions. Sales-Approved is shared read-only and Drafts is writable. The owner's personal folders and Clients-Restricted stay unshared with agent accounts. HubSpot stays the record for customer status.

Quick to start

Quick to start

What is the smallest useful routine you can test? One real task, draft-only, with you approving every send. The goal is a controlled first win, not instant autonomy.

Sample plan · one routine, in this order

First month

  1. 1Time one real taskStart to approved finish: your baseline

    Pick one repeated job, such as weekday follow-ups. Time it now, including review and corrections. Without this number, you cannot tell later whether anything changed.

  2. 2Set up the draft and approved contextGrok Bot draft, plus filed approved material

    Connect the drafting routine to approved inputs only: current templates, prices and good examples. The sample plan assumes 12–24 h of initial setup.

    See which files are approved in Their own files

  3. 3Run draft-onlyYou approve every send

    This is the checked first use. The routine’s safeguards are:

    • No auto-send, pricing changes or deal-stage edits by the agent.
    • Every send gets a HubSpot note.
    • Watch plan usage limits. Extra credits cost more.
  4. 4Compare with the baselineCompleted, accepted results only

    Time completed, accepted results against step 1. A quick draft that you end up rewriting does not count as a gain, because the review and fixes are in the timing.

After that

  1. 5Repeat on a schedule and improve itEach run still ends with your approval

    Fix the most common miss, put the routine on a schedule and run it again. When the same issue repeats, update the routine or its source file instead of correcting it by hand each time.

    The sample plan allows 2–4 h a month for care and upkeep.

Planning assumptions

12–24 hInitial setup

2–4 h a monthCare and upkeep

1 routineTested first

What the sample hours do and do not sayA hypothetical comparison, not a quote or a promise

In the sample, the same workload takes 30 h today and 20 h with the setup, which is 10 h back for this one routine. That is a hypothetical batch with review and upkeep included. It is not a 40-hour week or a proven result, so test it against your own baseline.

Full first-month plan, safeguards and license starting pointPlanning figures, the sample calculation and the usage source

Quick to start

Planning assumptions

  • 12–24 hInitial setup
  • 2–4 h a monthCare and upkeep
  • 1 routineTested first

First month

  1. Time one real task from start to approved finish.
  2. Set up the Grok Bot draft and file approved context.
  3. Run draft-only. You approve every send.
  4. Time completed, accepted results against the baseline.

Safeguards

  • No auto-send, pricing changes or deal-stage edits by the agent.
  • Every send gets a HubSpot note.
  • Watch plan usage limits. Extra credits cost more.

Same weekly workload · 30 h today → 20 h with the setup = 10 h a week back for this one routine.

Licenses start at $70 a month, or $86.80 with new Workspace. Sample hours only. Not a quote.