For solo and boutique consultants, 1 to 10 people
An automated proposals consulting workflow drives the four nodes you already have. It does not replace them with a new platform.
The pages that currently answer this question all sell you a templating platform: Proposify, PandaDoc, Better Proposals, Bonsai, Lindy. Their argument is that you migrate your proposals into a new environment, build a content library, set up merge fields, and send from inside their system. The honest mechanism is the opposite. Your Google Doc proposal template, your DocuSign account, your CRM, and your transcript tool already exist. Clone drives those four things directly. This page walks through how, section by section.
Direct answer · verified 2026-05-06
How do you automate consulting proposals end to end?
You drive the four nodes that already exist (the discovery call transcript, your past proposal Google Doc, the e-signature send, the CRM stage update) instead of replacing them with a templating SaaS. Clone reads the transcript on screen, opens your existing Google Doc proposal template, drafts each of the eight canonical sections using language retrieved from your last twelve sent proposals, queues a DocuSign envelope for your review, and moves the deal to Proposal Sent in your CRM. The first draft lands in 5-10 minutes after the call ends; the source workflow doc lists the same task as 1-3 days by hand.
Authoritative source for the mechanism and pricing: cl0ne.ai. Re-verified 2026-05-06.
The thesis in one paragraph
"Proposal automation" has come to mean adopting a templating platform. It does not have to.
Open the first ten guides that come up for this topic and they all describe the same flow. You move your proposals into Proposify or PandaDoc, you build a content library, you set up token placeholders, you wire role-based approvals, you connect a CRM integration on the Business tier, and you send from inside the platform. Every guide that ranks treats this as the only mechanism available, then differentiates by feature checklist (token granularity, signature options, analytics depth).
That mechanism solves one real problem (the rep typing the same paragraphs over and over) and creates three. The proposal lives inside a hosted platform that your client clicks a link into, which reads as a corporate sales document rather than a Google Doc from a consultant they trust. The merge fields are recognizable to any client who has received one before. The content library locks the proposal body the day it was written, so every prospect at a similar deal stage gets language built once for nobody in particular.
A different mechanism exists. The four nodes that produce a real consulting proposal (the call transcript, your past proposal Doc, your e-signature account, your CRM) all already live on your machine and in your accounts. Driving those four directly, with the discovery transcript as the spec and your past twelve sent proposals as the voice corpus, produces a draft in your Google Doc that no client can tell apart from one you wrote on a good morning. The rest of this page walks through how that compose chain works, section by section.
The compose chain
Four inputs you already have. Three outputs that already exist. One Memory layer that retrieves your voice.
Nothing in this diagram is a new tool you have to install. The transcript came out of the call you just finished. The folder of past proposals is the one you already keep in Drive. The terms doc is the boilerplate you already paste from. The CRM is wherever you track deals today. Clone is the layer in the middle that retrieves voice from those sources, structures the eight sections, and writes them into the destinations.
What feeds the proposal, and what the proposal turns into
The eight sections, in order
Every consulting proposal is the same eight sections. Clone populates each from a different source.
The canonical structure is in the source workflow doc on this product's repo (consulting-business-workflow.md, Phase 3 Step 2, lines 106-119). It is not Clone's invention; it is the structure every working consultant converges on after enough engagements. Below is what each section needs and where Clone retrieves it from.
1. Executive summary
Restate the prospect's problem in their own words. This is the single section every existing platform gets wrong because it cannot retrieve the prospect's actual phrasing from the call.
Clone pulls the two or three sentences where the prospect described the pain in their own words from the discovery transcript. It quotes those lines verbatim into the executive summary. No paraphrase, no consultant-speak overlay.
2. Current state
What you observed about their business during the call: the tools, the bottleneck, the cost. Not your interpretation; their description, organized.
The Planner layer classifies transcript sentences into observations, commitments, decisions, and risks. The observations are quoted into section two with timestamps preserved in a comment so you can verify before you send.
3. Recommended approach
Phased plan, broken into steps. The reasoning here is yours; the structural pattern (how you phrase phases, how you break them up) is retrieved from your last twelve sent proposals.
Clone does not invent the approach. It writes the approach you sketched in the call (where you said things like "the first thirty days are about getting visibility") in the structural pattern your past proposals follow. You edit the substance. The form is already yours.
4. Deliverables
Specific, tangible outputs the engagement will produce. Reports, systems configured, processes documented, training delivered.
Pulled from the commitments you made on the call ("I'll document the new intake flow", "I'll train your team on the dashboard"). Bulleted in the format your past proposals used. If your past proposals listed deliverables as "By end of week 4: ..." Clone follows that.
5. Timeline
Week-by-week or phase-by-phase breakdown. The dates come from the call; the layout (calendar grid, table, prose) follows the pattern from your last twelve sent proposals.
Clone reads the dates the prospect named ("we want to be live by end of June") and back-solves the milestones. The Memory layer remembers whether you bullet timelines, table them, or write them in prose. It does not flip formats between proposals to the same client.
6. Investment (three options)
Option A: smaller scope (assessment only). Option B: full engagement. Option C: full engagement plus retainer. Three options anchors the conversation and lifts close rate.
The 3-option pattern is in the source workflow doc at lines 114-117. Clone preserves the structure but populates each option with numbers derived from your past proposals at similar scope, with the actual line items the prospect asked for. Pricing is never invented; if Clone cannot anchor a number to a precedent, it leaves a flag in the doc for you to fill.
7. About you
Two or three sentences of relevant experience and one testimonial. The shortest section in every proposal, and the one most consultants over-optimize.
Pulled from your existing About-you boilerplate Doc, with a single-sentence customization based on the closest analogue engagement in your past work. If you have a testimonial that specifically references the bottleneck the prospect described, Clone surfaces it.
8. Terms
Payment schedule, what's included and excluded, assumptions. The boring section that protects you when scope drifts.
Pulled verbatim from your master terms boilerplate. The only per-proposal changes are the payment schedule (which references the Investment option you anchored) and any client-specific assumption you mentioned on the call.
What the prospect actually reads
The single section that gives away which mechanism produced the proposal: the executive summary.
Every other section can be wrong in similar ways. The executive summary is the section a prospect reads first, the section where they decide whether the document is talking to them or talking past them. Templated proposals get this section structurally right and substantively generic. Retrieved proposals quote the prospect's own phrasing back at them.
Same prospect. Same call. Two different mechanisms.
Dear {{first_name}},
Thank you for the conversation last
{{call_day}}. Following up on our
discussion about {{pain_point}}, please
find our proposal below.
[CONTENT BLOCK: Standard Approach]
[CONTENT BLOCK: Standard Deliverables]
[CONTENT BLOCK: Standard Timeline]
[PRICING: Pull from {{deal.amount}}]
We look forward to partnering with
{{company_name}} on this engagement.
Best,
{{rep_name}}Templating SaaS vs. driving your existing stack
A direct line-by-line comparison of what changes when you adopt a proposal platform versus when Clone drives the apps you already have.
| Feature | Proposify / PandaDoc | Clone |
|---|---|---|
| Where the proposal lives | Inside the platform; you and your client both need an account | Inside your Google Doc, the same template you've used for years |
| Voice in the executive summary | Merge fields and a content-library snippet | The prospect's actual sentences from the call, quoted verbatim |
| Voice in the recommended approach | Pre-written content blocks, picked by deal stage | Your structural pattern retrieved from your last 12 sent proposals |
| Pricing source of truth | Whatever the rep types in, no precedent check | Anchored to your past engagements at similar scope; flagged when no precedent exists |
| Send mechanism | Native e-sign; you switch tools | Drives your existing DocuSign or HelloSign account |
| CRM stage update | Native if you bought their integration tier | Drives your existing HubSpot or Pipedrive (or a Sheet) directly |
| Time from call end to first draft | 30-90 minutes if all merge fields hit | 5-10 minutes; you spend the rest on the edit pass |
| Migration cost | Onboard team, rebuild templates, retrain on a new UI | Zero. Clone drives the apps already on your machine |
| Cost | $40-100+ per seat per month | $49 per month flat on Solo, $129 per seat on Boutique |
Pricing for Proposify and PandaDoc reflects published Team and Business tier rates as of 2026-05-06. Verify on the vendor's pricing page; both platforms run frequent promotions.
The honest counterargument
When a templating platform actually is the right answer.
If your proposals are sold by a sales team rather than a consultant, if the same person rarely sees the engagement they are pricing, and if compliance requires every draft route through named approvers before it leaves, the role-based approval workflow inside Proposify or PandaDoc is doing real work. That is not the consulting use case this page is about, but it is a real use case and the platforms exist for a reason.
If your proposal volume is in the hundreds per month and the content is genuinely commodity (procurement responses, RFPs with fixed answer banks), the content library actually pays for itself. A solo consultant or a five-person boutique firm that sells custom advisory engagements is the opposite of that customer. The whole point of the engagement is that the proposal is specific to the prospect. Templating works against the differentiator.
The decision is not "templating SaaS bad, screen-driving good". The decision is which mechanism matches the size and shape of your practice. For 1-10 person consulting and advisory firms billing on retainer or project, the math points one way. For inside sales teams of fifty selling repeatable SKUs, it points the other.
Setup, in order
What it takes to get the first proposal drafted.
- Point Clone at your proposal Drive folder. Clone reads your last twelve sent proposals at similar scope and surfaces the structural patterns it observes (sign-off, deliverable bullet style, timeline layout, three-option pricing shape, About-you boilerplate). You confirm or correct each pattern once. Total time: about fifteen minutes.
- Connect your transcript tool. Open it once with Clone watching. tl;dv free, Fireflies free, Otter, Granola, and native Zoom cloud all work without an API key because the Computer Agent reads the transcript page on screen. Total time: two minutes.
- Name your CRM and the deal stage. "When a draft is ready, move the deal to Proposal Drafted in HubSpot." Plain English. No field mappings, no API setup. If you run a Google Sheet as your CRM, the same instruction works against the Sheet.
- Write rituals/proposal.md. Three or four lines of plain English describing what to do after a discovery call. Where the transcript is, where the proposal template lives, how you want the three-option pricing anchored, what the default DocuSign envelope settings are. The file lives on your disk, not in a Clone account.
- Run it after your next discovery call. One instruction at the end of the call: "Read the transcript, draft the proposal in the standard template, queue DocuSign, move the deal to Proposal Drafted." First draft in five to ten minutes. You spend the rest of the post-call hour on the edit pass and the framing of options A, B, and C.
What you still do
The edit pass is where the proposal earns its rate.
Clone does the retrieval and the structural population. The 5-10 minutes between the end of the call and the first draft is the speed-up. The 30-45 minutes after that, where you sit with the draft and shape the recommended approach, frame the three options, and decide which deliverables to push and which to quietly drop, is still you. That work is the reason the engagement is worth what it is worth. Clone removes the typing, not the judgment.
In practice, the edit pass on a Clone-drafted proposal looks different from one on a hand-drafted proposal. With a hand draft, you are still figuring out what to say in section three; with a Clone draft, the structural skeleton is already correct and you are tightening language and pressure-testing the option B scope. The cognitive shift is from blank-page to red-pen, which is the same shift that happens when a senior associate hands you their draft. Clone is the senior associate's draft. You are still the partner.
The 30-minute proposal presentation call (the one in step 3 of the source workflow doc, line 121, where you "never just email the proposal") is unchanged. You walk the prospect through the doc page by page, stop talking when you hit pricing, and handle objections. Clone does not replace that conversation. It just ensures that the document you are walking through was built from the prospect's actual call, not from a template library.
Want to see the proposal compose chain run on a real call?
A 20-minute walkthrough of what Clone drafts after a discovery call ends. Bring a transcript from a recent call you have permission to share, or use a sample. We'll watch the four-node chain run end to end and answer the questions this page didn't.
Frequently asked questions about automating consulting proposals
What does an automated consulting proposal workflow actually do?
It compresses the post-discovery, pre-contract bridge in your sales cycle. The canonical structure is in the source workflow doc on this product's repo (consulting-business-workflow.md, Phase 3 Step 2): a 2-5 page proposal with eight named sections (executive summary, current state, recommended approach, deliverables, timeline, investment with three options, about you, terms). The doc lists the time tax as 1-3 days. An automated workflow turns that into a 5-10 minute first draft plus your edit pass, by reading the call transcript and your last twelve sent proposals to populate every section in your voice. The reading and writing happen in the apps you already use, not inside a separate platform.
How is this different from Proposify, PandaDoc, Better Proposals, or Bonsai?
Those four are templating platforms. You move your proposals into their environment, build a content library, set up token placeholders, and send from inside their system. The pages that rank on this topic almost all describe that flow. The trade-off is that every proposal that leaves your inbox now lives in a third-party platform, your client clicks a link to a hosted page rather than opening a Google Doc, and the merge fields are recognizable to any client who has received one before. Clone takes the opposite approach. Your Google Doc proposal template stays where it is. Your e-signature account stays where it is. Your CRM stays where it is. Clone drives those four things from a plain English instruction. There is no platform to migrate to and no template language for your clients to recognize.
How does Clone retrieve voice from twelve past proposals?
The mechanism is shown verbatim on the live product page in src/components/how-it-works.tsx step 03 (line 45 names the count). Clone reads your existing proposal folder in Google Drive, observes the patterns in your last twelve sent proposals at similar scope, and surfaces them as a candidate template ('You start with a personal line, you always attach the SOW PDF, you cc your assistant when the contract is above $10K. Should I apply this template going forward?'). You confirm once. The pattern goes into the Memory layer on your disk and is reused. Twelve is the working number where structural patterns are stable but per-client variation is still preserved. Below that the pattern is noisy. Above that, returns flatten.
What about the discovery call transcript? How is that read?
Clone's Computer Agent layer (described in src/components/architecture.tsx around line 18) opens the transcript page in your browser, the same way you would. It reads the page on screen. This means it works against tl;dv free, Fireflies free, Otter, Granola, or native Zoom cloud transcripts without an API key. Inside the transcript, the Planner classifies sentences into commitments, decisions, observations, and small talk. Commitments and observations populate the Deliverables and Current state sections. Quoted phrasing populates the Executive summary. Timestamps for each quoted line are preserved as comments in the draft so you can verify before sending.
What are the eight sections of a consulting proposal Clone actually drafts?
Pulled directly from the canonical structure in the source workflow doc (consulting-business-workflow.md, lines 106-119): (1) Executive summary, (2) Current state, (3) Recommended approach, (4) Deliverables, (5) Timeline, (6) Investment with three options (A: smaller scope, B: full, C: full plus retainer), (7) About you, (8) Terms. The three-option pattern in section six is load-bearing because it anchors the conversation and lifts close rate (lines 114-117). Clone preserves the structure exactly. Every retrieval pass populates these eight slots and only these eight slots.
Does Clone auto-send the proposal?
No, not on the first pass and not on any review_first stage. Every draft lands in your existing Google Doc as a draft (with comments showing where each quoted line came from in the transcript). The DocuSign send is queued, not fired. You walk through the proposal in a 30-minute call with the prospect (the source doc, line 122, is explicit: 'Never just email the proposal'). After the prospect agrees, you review the queued DocuSign and send it yourself. Once you have observed Clone for two or three engagements, you can flip specific stages from review_first to auto, but the default is human-in-the-loop.
What if my proposal Google Doc has custom formatting or a complicated cover page?
Clone drives Google Docs the way you do: by opening the doc, navigating to the right heading, and typing into the section. If your template has a cover page with a logo and a signed cover letter, Clone leaves that block alone and writes underneath it. If you bullet your deliverables in a specific style, Clone matches the style because the Memory layer observed it across your last twelve proposals. Custom formatting is the case where this approach beats a templating SaaS most clearly: the SaaS wants to own the layout, but your existing Doc already owns it correctly.
Does my client data leave my computer?
No. The transcript stays in your transcript tool. The Memory patterns Clone observed across your past proposals stay in plain text files on your disk. The Doc stays in your Drive (under your account, not a Clone-owned account). The DocuSign envelope is created in your DocuSign account. The architectural principle is on the live product page (src/components/architecture.tsx, lines 46-49): Clone operates your desktop apps from your desktop, and client files, emails, contracts, and transcripts never leave your computer. The Planner may call a model to interpret your plain English instruction, but client artifacts are not sent along with the call.
What does this cost compared to Proposify or PandaDoc?
Clone Solo is $49 per month, flat. Boutique is $129 per seat per month. PandaDoc Essentials is currently $19 per seat per month but the proposal-automation features (CRM integrations, conditional content, role-based approval, advanced analytics) require Business at $49 per seat per month or higher. Proposify Team starts at $49 per seat per month and the Business tier (with Salesforce integration and content library limits removed) runs higher. The non-monetary cost difference is bigger: with Clone you do not migrate templates, you do not retrain on a new UI, and your clients keep receiving proposals in the format they expect.
How long does it take to set up?
On the order of an hour. You point Clone at your proposal Drive folder so it can read your last twelve sent proposals. You confirm the structural patterns it surfaces (sign-off, deliverable bullet style, timeline format, three-option pricing layout, your boilerplate About-you line). You connect your transcript tool by opening it once. You name your CRM and the deal stage that should trigger 'proposal sent'. You write a short rituals file (proposal.md) with three or four lines of plain English describing what to do after a discovery call. After that, the workflow runs from a single instruction at the end of each call.
What does the first instruction look like in practice?
Plain English, no syntax. Something like: 'Read the transcript from this morning's Acme call. Pull the bottleneck and the dates they named. Open the Acme folder in Drive, copy my standard proposal template, draft sections one through five using the call. Use my standard three-option pricing pattern for section six, anchor the numbers to the Stellar engagement we ran last quarter, and flag option B if it crosses $25K. Leave About-you and Terms verbatim. Queue a DocuSign envelope but do not send. Move the deal to Proposal Drafted in HubSpot.' Clone reads that, executes against your existing apps, and surfaces the draft for your review.
How does this fit with the rest of the consulting back-office automation Clone does?
The proposal step is one node in a longer chain. Discovery call, proposal, contract, kickoff, weekly status, milestone invoice, follow-up, monthly report, closeout. Clone runs all of these against your existing tools using the same retrieval mechanism applied to different artifacts (transcripts, sent emails, past invoices, past status reports). The proposal step is the highest-leverage one because the source doc lists it as a 1-3 day task that recurs once per engagement, but the same mechanism removes the equivalent admin tax from every other step in Phase 3, 4, and 5 of the canonical workflow.
The same retrieval mechanism applied to other parts of the consulting workflow.
Related guides
Automate Consulting Workflow Without New Software
The general principle behind this proposal-specific page. Why the right answer is screen-driving the apps you already have, not adopting a workflow platform.
Active Recall Client Follow-Up Generator
The same retrieval mechanism applied to follow-up emails. Twelve sent emails per stage, voice retrieved instead of recognized.
Consulting Workflow Automation
How the post-call admin chain (transcript → CRM → invoice → follow-up) automates end to end against your existing stack.