ChatGPT
Claude
PerplexitySalesforce Implementation Proposal Checklist
A Salesforce implementation proposal needs eight sections: client context, objectives, scope of services, approach and timeline, team and credentials, assumptions, commercial model, and acceptance terms. Buyers rarely score the first two and almost always score the middle four, which is the opposite of where most effort goes.
Before working through the list, hold one distinction. A proposal persuades and may stay approximate. A statement of work binds and cannot. Salesforce’s own customer agreements make the point contractually, defining a deliverable narrowly as any output identified as a deliverable under a statement of work or order form, which means nothing in a proposal is a deliverable until it survives into the contract.
Eight sections carry a Salesforce implementation proposal
Section | Job it does | Length that works |
|---|---|---|
1. Client context | Proves you listened, in their language rather than yours | Half a page |
2. Objectives | States the measurable outcome the project serves | Half a page |
3. Scope of services | Names deliverables that can be inspected and handed over | Two to three pages |
4. Approach and timeline | Shows the sequence, dependencies and what you need from them | Two pages |
5. Team and credentials | Names the people, their certifications and their availability | One page |
6. Assumptions | States what must be true, and what happens if it is not | One page |
7. Commercial model | Price, structure, payment schedule, what triggers more | One to two pages |
8. Acceptance and terms | How the proposal converts into an engagement | Half a page |
Eight to fifteen pages covers a mid-sized implementation. Length beyond that reduces the chance anyone reads section 6, which is the section that decides whether the engagement makes money.
Buyers score proposals on risk reduction, not presentation
Where proposal effort usually goes and where it earns are different places.
Cover design, company history, awards and empathetic restatements of the client’s pain carry almost no weight in a scored evaluation. The buyer already knows their pain. Restating it consumes a page and proves only that you were listening, which is table stakes rather than a differentiator.
What moves a score is anything that reduces the buyer’s perceived risk of choosing you. Named people with real availability. Assumptions written as testable statements. A timeline that accounts for their decision speed rather than assuming perfect responsiveness. References in their industry at their size. A commercial model whose structure matches how certain the scope actually is.
The practical rewrite is to take a page from sections 1 and 2, and give it to sections 5 and 6.
Scope belongs in deliverables, not activities
Section 3 fails when it lists activities. “Discovery workshops” is an activity. “Solution design document covering objects, sharing model and integration contracts” is a deliverable, because it can be handed over and inspected.
Write each deliverable with a one-line definition, then write the out-of-scope list underneath it. That second list is the part experienced buyers read first, because it tells them what they will be asked to pay extra for later. Naming the adjacent things a reasonable reader might assume are included, such as additional clouds, historical data beyond a stated cut-off, or training beyond the stated sessions, reads as confidence rather than as hedging.
Four sections most Salesforce proposals leave out in 2026
These follow from platform changes rather than from proposal convention, and their absence is now noticeable to any buyer who has been through a recent implementation.
Addition | What to state | Why it matters |
|---|---|---|
Agent scope | Which agents, how many subagents and actions each | Agentforce work is rarely a line item in older proposal templates |
Consumption ownership | That Agentforce and Data 360 usage is metered by Salesforce and billed to the client | It is a recurring cost that continues after you leave |
Agent testing approach | Scoring against named quality dimensions, and who sets the threshold | Agent tests are scored rather than passed, so sign-off needs a number |
AI usage in delivery | Where AI is used and where client data goes | Buyers now ask, and a vague answer costs more than an honest one |
The consumption point is the one that damages relationships when it is missing. A client who discovers a metered bill after go-live remembers that the proposal did not mention it, regardless of what the contract said.
The assumptions register decides whether the proposal survives delivery
Section 6 is short, unglamorous and the highest-leverage page in the document.
Write each assumption as a statement someone other than its author can check, paired with what happens commercially if it proves false. “Client data is of reasonable quality” is not checkable. “Source data comprises three systems and fewer than five percent of records require manual remediation, with remediation above that threshold billable at the stated rate” is.
Three assumptions belong in almost every Salesforce proposal: the data quality threshold, the client decision response time, and integration endpoint documentation. Each is a common cause of overrun, and each becomes a priced event rather than an argument once it is written down.
Proposal content maps to SOW clauses at signature
The proposal is the draft of the contract, whether or not anyone treats it that way. Knowing which sections carry over saves a negotiation round.
Proposal section | Becomes in the SOW |
|---|---|
Scope of services | Scope and deliverables schedule, plus the out-of-scope clause |
Approach and timeline | Deliverables schedule with dependencies |
Team and credentials | Named resources with allocations, if the client asks for it |
Assumptions | Assumptions register, contractually binding |
Commercial model | Pricing and payment clause, plus change control threshold |
Acceptance and terms | Acceptance mechanism with a review window |
Write section 6 in contract-ready language at proposal stage and the transition is a copy exercise. Write it as marketing language and you will renegotiate it under time pressure. The full clause set is in from proposal to SOW.
Where Salesforce implementation proposals lose
Mistake | What the buyer concludes | Correction |
|---|---|---|
Activities listed as deliverables | Nothing can be accepted or compared | Every deliverable is inspectable |
No out-of-scope list | Change orders are coming | Name the adjacent exclusions |
Generic team page | The pitch team is not the delivery team | Name individuals with allocations |
Assumptions with no consequences | The number is soft | Attach a consequence to each |
Fixed price on undefined scope | Either optimistic or hiding contingency | Match commercial model to scope certainty |
Timeline assuming instant approvals | They have not worked with an organisation like ours | State the response time you assumed |
Consumption unmentioned | Surprise cost after go-live | State who owns it, plainly |
Two of these are worth pairing. A generic team page next to an aggressive timeline reads as a proposal written by people who will not deliver it, which is the single fastest way to lose a scored evaluation to a smaller competitor.
The checklist, condensed
Item | Reference |
|---|---|
Sections | Eight, of which scope, approach, team and assumptions carry the score |
Length | Eight to fifteen pages for a mid-sized implementation |
Deliverable test | Can it be handed over and inspected |
Assumption test | Can someone other than the author check it |
Sections to shorten | Cover, company history, pain-point restatement |
Sections to lengthen | Team, assumptions, out of scope |
2026 additions | Agent scope, consumption ownership, testing approach, AI usage |
Proposal versus SOW | Persuades and may approximate, rather than binds and cannot |
Producing proposals without the weekend
Proposal work is unbillable, competitive and time-boxed, which is the worst combination in consulting. GetGenerative.ai’s Discovery Agent generates the roadmap, business case, scope and resourcing plan that sections 2, 3 and 4 rest on, with up to 70 percent effort saved on business case creation, and the same structure carries into the SOW rather than being rewritten. If you are drafting from notes rather than from a blank page, drafting the proposal with AI covers the workflow.
For teams producing these regularly, the Pro plan for consultants includes a trial period with full agent access, which is long enough to run one live opportunity through it.
Questions consultants ask
What should a Salesforce implementation proposal include?
Eight sections: client context, objectives, scope of services, approach and timeline, team and credentials, assumptions, commercial model, and acceptance terms. In 2026, add agent scope, consumption ownership, agent testing approach and a statement on AI usage in delivery.
How long should a Salesforce proposal be?
Eight to fifteen pages for a mid-sized implementation. Longer documents reduce the chance anyone reads the assumptions section, which is where the commercial protection lives. Depth belongs in scope, team and assumptions rather than in company background.
What is the difference between a proposal and a statement of work?
A proposal persuades and may stay approximate. A statement of work binds and cannot. Nothing in a proposal is contractually a deliverable until it is identified as one in the signed document, which is why proposal language should already be close to contract language.
Should a proposal include pricing?
Yes, with structure rather than a single total. Show cost by phase or workstream, name what is excluded, and state what triggers additional fees. A total with nothing behind it invites the buyer to negotiate the total rather than the scope.
How do I make a proposal stand out?
By reducing the buyer’s risk rather than by design. Named people with real availability, testable assumptions, an honest out-of-scope list, and a timeline that accounts for their decision speed will beat a better-looking document from a firm that promised everything.
Do proposals need to mention Agentforce consumption?
If agents are in scope, yes. Consumption is metered by Salesforce, billed to the client, and continues after the engagement ends. Raising it in the proposal costs one paragraph. Discovering it after go-live costs the relationship.
