Quick Answer
A web design proposal closes when it removes doubt, not when it lowers price. The buyer needs to see that you understand the problem, that the scope is countable, that the timeline depends on them as much as on you, and that the number will not move after signature. Most proposals fail on the second point. This is how to write one that holds.
By Cody Wise, founder of Wise Media. This article covers commercial and operational practice, not legal or tax advice. Confirm contract terms with a lawyer and tax treatment with an accountant before relying on them.
Summary
- A proposal is a sales document. A contract is a legal document. Confusing the two is why proposals read as defensive and still fail to protect you.
- Scope holds when it is countable: number of templates, number of unique page layouts, number of revision rounds, number of integrations. Adjectives do not hold. Numbers do.
- Three options beat one price. A single number turns the conversation into yes or no. Three turns it into which one.
- The clause that actually stops scope creep is not a warning. It is a named, priced, non-punitive change order path that makes saying yes to extra work easy and billable.
- In Canada, which sales tax you charge is usually set by the recipient address you obtain in the ordinary course of business, not by where you sit. Quote the rate that matches the client province.
- Put the price on page one or two. Buyers scroll to it first regardless, and burying it reads as a lack of conviction.

Table of contents
- What should a web design proposal include?
- Why do most proposals lose the deal before anyone reads the price?
- How do you write scope that actually holds?
- How should you price a web design proposal?
- What actually stops scope creep after signature?
- Proposal or contract: what belongs in which document?
- Common mistakes that cost the deal
- The pre-send checklist
- Frequently asked questions
What should a web design proposal include?
A web design proposal should include nine sections: a restatement of the client problem, the outcome you are selling, a countable scope, what is explicitly excluded, the process and timeline, what the client has to supply, three priced options, payment terms, and a single next step. Anything else is decoration.
Notice what is not on that list. No company history. No team bios. No twelve slides on your design philosophy. A buyer reading three proposals does not choose the agency with the best origin story, they choose the one whose document made them feel like the project was already under control. Every section below earns its place by removing a specific doubt.
| Section | The doubt it removes | Length |
|---|---|---|
| Problem restatement | “Do they even understand my business?” | 1 short page |
| Outcome | “What am I actually buying?” | 3 to 5 bullets |
| Scope, counted | “What exactly do I get?” | 1 page, a list |
| Exclusions | “What will they bill me for later?” | 5 to 8 lines |
| Process and timeline | “When is this done?” | 1 page |
| Client responsibilities | “Why do projects stall?” | 5 to 8 lines |
| Three priced options | “Is this the right size?” | 1 page |
| Payment terms | “What is the cash commitment?” | 5 lines |
| Next step | “What do I do now?” | 2 lines |
Total: roughly six to eight pages. If yours runs twenty, you are selling to your own insecurity rather than to the buyer. This is the same discipline that makes a conversion system work on a website: fewer elements, each doing one job, each removing one objection in order.
Why do most proposals lose the deal before anyone reads the price?
Most proposals lose because they describe the agency instead of the project. The buyer opens the document looking for evidence that you listened, and finds a capabilities deck. By the time they reach the price there is no context for it, so the number has to defend itself alone. It never wins that fight.
The opening page test
Read your first page and ask one question: could this page have been sent to any other client without editing? If yes, rewrite it. The opening page should contain at least three specifics that only came out of the discovery call. The product line they are discontinuing. The seasonal dip in July. The fact that their best salesperson leaves in March and the site has to carry the lead flow alone after that.
Specifics are expensive to fake, which is exactly why they signal competence. A competitor who did a shallower discovery call physically cannot match that page.
Selling the artifact instead of the outcome
“A twelve page WordPress website” is an artifact. “A site that turns the three hundred monthly searches for your service in your city into booked consultations, and that your office manager can update without calling us” is an outcome. The artifact is what you build. The outcome is what they are buying. Lead with the second and the first becomes a specification rather than a price tag.
This matters more than it used to. Buyers now compare a custom build against an AI site builder, a template, and a freelancer in the same afternoon. If you sell the artifact you are competing on price with all three. If you sell the outcome, and you can show what mechanism produces it, you are the only quote on the table that addresses the actual business problem. Our breakdown of why a site produces poor quality leads is the mechanism argument in long form, and it works inside a proposal just as well.
Hiding the price
Buyers scroll to price first. Every time. Putting it on page fourteen does not delay the comparison, it just means they make the comparison without the context you spent thirteen pages building. Put pricing on page one or two, with the three options visible together, then let the rest of the document justify the one they are already drawn to.
How do you write scope that actually holds?
Scope holds when every line is countable and every boundary is named. Replace adjectives with numbers: not “a modern multi page site” but “nine unique page layouts, twenty two total pages, two rounds of revision per layout, three third party integrations.” A client cannot argue with a count. They can always argue with “modern.”
Count the things
Every ambiguous noun in a web project has a number attached to it. Find the number before you send the document.
| Vague | Countable |
|---|---|
| Custom design | 9 unique page layouts, designed once, applied to 22 pages |
| Mobile friendly | Tested at 390px, 768px and 1440px on current Safari and Chrome |
| Revisions included | 2 consolidated rounds per layout, feedback returned in one document |
| Content migration | Up to 40 existing pages migrated, client supplies any new copy |
| SEO included | Title tags, meta descriptions, heading structure and schema on all 22 pages. Ongoing SEO not included |
| Integrations | 3 named systems. A fourth is a change order |
| Training | One 60 minute recorded session plus a written handover |
| Support after launch | 30 days of defect fixes. New requests are billable |
Writing the right hand column takes about forty minutes. It is the highest return forty minutes in the entire sales process, because every row is an argument you will not have in month three.
Name the exclusions out loud
Agencies avoid exclusion lists because they feel negative. They are the opposite. An exclusion list is the single clearest signal that you have run this project before and know where it goes wrong. Keep it to five or eight lines and keep the tone neutral.
- Copywriting beyond editing supplied text
- Photography, videography and paid stock licences
- Ongoing SEO, advertising or content production after launch
- Third party software subscriptions, which are billed to the client directly
- Translation or a second language version of the site
- Data migration from a system not named in the scope
- Accessibility remediation of content the client publishes after handover
Note the third party software line. Licences are the quiet killer of project margin, and they are also where a client feels ambushed. Our web design contract checklist for Canadian projects goes deeper on licence ownership, intellectual property and what happens to accounts at the end of a build. The proposal flags the cost. The contract settles who owns what.
The revision clause people actually honour
“Two rounds of revisions” fails because nobody agrees on what a round is. Define it mechanically:
A revision round is one consolidated set of written feedback, returned within five business days, covering one design stage. Feedback arriving after the round closes, or arriving piecemeal from multiple people across several days, starts a new round.
That definition does two things at once. It caps your exposure, and it trains the client to collect internal opinions before sending them, which genuinely makes the project better. Clients rarely object, because the rule is obviously fair in both directions.
Client dependencies belong in the proposal, not the post mortem
Most web projects run late because of the client, and most clients genuinely do not know that in advance. Put the dependency list in the proposal with dates attached and the timeline stops being a promise you alone carry.
- Logo files in vector format, plus brand colours and fonts, by the kickoff date
- Final copy for any page you are not writing, by the end of design week two
- Photography or a decision to licence stock, by the same date
- Registrar and hosting access, or written authority to transfer, before build starts
- One named decision maker with authority to approve, for the whole project
- Feedback returned within five business days at each stage
Add one sentence after the list: timeline assumes these dates are met, and shifts day for day if they are not. That sentence has saved more project margin than any rate increase.

How should you price a web design proposal?
Present three options rather than one number. A single price forces a yes or no decision. Three options change the question to which one, which is a decision the buyer is far more willing to make. Price the middle option as the one you expect to sell, make the top option genuinely better rather than padded, and keep the bottom option real enough that choosing it is not embarrassing.
Structuring the three tiers
| Essential | Recommended | Growth | |
|---|---|---|---|
| What it is | The site, built properly | The site plus the conversion layer | The site plus the engine that feeds it |
| Typical inclusions | Design, build, launch, training | Essential plus analytics, tracking, lead routing, CRM connection | Recommended plus content, SEO and ongoing optimisation |
| Commitment | Project fee | Project fee | Project fee plus monthly |
| Who picks it | Budget constrained, wants the asset | Most buyers | Buyers with an existing demand problem |
The Growth tier does the heaviest lifting in the document even when nobody buys it, because it reframes the middle option as the sensible choice rather than the expensive one. It also plants the retainer conversation at the moment of highest trust rather than six months later. If recurring revenue is the goal, the proposal is where that starts. Our piece on building a recurring revenue stack covers what belongs in the ongoing tier, and how Canadian agencies price retainers covers the number itself.
Deposits and the payment schedule
Tie payments to stages rather than to dates. A date based schedule bills a client for a month in which nothing visibly moved, usually because they were the bottleneck, and that is where invoice disputes start.
- Deposit on signature: commonly 30 to 50 percent of the project fee. Non refundable once discovery begins, and say so plainly.
- Design approval: a middle instalment released when layouts are signed off.
- Launch: the balance, due on launch or on a stated number of days after the site is ready and the client has delayed launching.
That last clause matters. Sites sit finished and unlaunched for months while a client sources one photo. Without a readiness trigger the final invoice sits with it.
GST, HST and which rate goes on the quote
For Canadian agencies, the sales tax line on a proposal is set by where the client is, not by where you are. The Canada Revenue Agency general place of supply rule for services is that a service is supplied in a province if the supplier obtains, in the ordinary course of business, a home or business address of the recipient in that province. The address has to be a physical one, so an email address or an IP address does not count.
Where you hold more than one address for a client, the rule points to the address most closely connected with the supply, and the CRA guidance treats the contracting address, meaning the address the client hires you from, as generally the most closely connected one. Practically, for an Alberta agency building a site for an Ontario company contracting from its Toronto office, the quote carries Ontario HST rather than Alberta GST.
Three habits follow from that. Collect a physical business address during discovery rather than at invoicing. Show tax as a separate line with the province named, so nobody reads the subtotal as the total. Confirm the treatment with your accountant, because the specific rules for services have special cases and this is tax, not design.
Source: CRA GST/HST Memorandum 3-3-6, Place of Supply in a Province, General Rules for Services, read 18 November 2026.
What actually stops scope creep after signature?
Scope creep stops when the change order path is easier to use than the argument. Most agencies write a defensive clause saying extra work will be billed, then never invoke it because invoking it feels confrontational. The fix is procedural: a named form, a published rate, a short turnaround, and a tone that treats additions as normal rather than as a breach.
The mechanism, in four parts
- A trigger definition. Anything not named in the scope list is a change, including things that feel small. Say this once, in neutral language, in the proposal.
- A standing rate. An hourly or half day rate published in the proposal so the number is never a negotiation later.
- A one page form. Description, price, schedule impact, approval line. Sent the same day the request arrives.
- A schedule consequence. Every change order states its effect on the launch date. This is what keeps a client from stacking “small” requests.
The schedule line does more work than the price line. Clients will often pay for an addition without hesitation and then be genuinely surprised that launch moved. Showing both every time makes the trade visible at the moment of the request, which is the only moment it can be reconsidered.
The sentence to use when the request arrives
Scope conversations go badly when they start with no. Start with yes and a price instead:
“Good call, that would work well. It sits outside the nine layouts we scoped, so I will send the change order this afternoon. It is roughly four hours and it moves design approval out by about three days. Want me to send it, or park it for phase two?”
Three things happen in that reply. The idea is validated, the boundary is restated without accusation, and a parking lot is offered. The parking lot is the underrated part. Half of all creep requests are enthusiasm rather than need, and a phase two list captures them without either a fight or free work. It also becomes the agenda for the retainer conversation after launch.
Pros and cons of a strict change order process
| Pros | Cons |
|---|---|
| Protects margin on fixed fee work | Adds administrative steps to small requests |
| Makes the schedule honest | Can feel bureaucratic to a small owner operator client |
| Converts creep into revenue | Requires you to actually send the form every time |
| Creates a natural phase two pipeline | A rigid reading can damage goodwill on genuinely trivial asks |
Mitigate the cons with a stated allowance: a small pool of minor change hours included in the project, drawn down visibly. Clients get latitude for the trivial, you get a clean line for everything else, and the drawdown itself communicates that the hours are real.
Proposal or contract: what belongs in which document?
The proposal sells and scopes. The contract allocates risk. Mixing them produces a document that is too legal to be persuasive and too loose to be enforceable. Keep them separate and have the contract incorporate the proposal scope by reference.
| Proposal | Contract |
|---|---|
| Problem, outcome, approach | Intellectual property and licence ownership |
| Countable scope and exclusions | Liability limits and indemnity |
| Timeline and client dependencies | Termination and kill fee |
| Three priced options | Warranty window and defect definition |
| Payment schedule in plain language | Late payment interest, governing law |
| Change order rate and process | Confidentiality and data handling |
One practical note for Canadian agencies: ownership of design and code does not transfer automatically on payment. It transfers when a written assignment says it does. That is a contract question, not a proposal question, and it is covered in detail in the contract checklist. Keep the proposal honest by not promising ownership language the contract does not deliver.
Common mistakes that cost the deal
- Sending the proposal without a call. A document emailed cold has to do all the persuading alone. Walk the buyer through it live, then send it as the record.
- Quoting before discovery is finished. A price given from an incomplete picture is either padded, which loses, or wrong, which costs you later.
- One price, no options. Turns the decision binary at the worst possible moment.
- Timeline with no client dependencies. You have just accepted sole responsibility for a schedule you do not control.
- Listing hours instead of outcomes. Hours invite line item negotiation. Nobody negotiates an outcome down, they negotiate it away.
- No expiry date. Prices held open indefinitely get used as leverage months later. Thirty days is normal and reasonable.
- Inventing results. Never attach a performance number you cannot evidence. In Canada, representations made to promote a product or service are subject to the misleading representations provisions of the Competition Act, and an unsupported performance claim is a commercial risk as well as a trust one.
- Burying the next step. End with one action, one link, one name.
The pre-send checklist
- Page one names at least three specifics from the discovery call
- Every scope line carries a number, not an adjective
- Exclusions listed, five to eight lines, neutral tone
- Revision round defined mechanically, not by count alone
- Client dependencies listed with dates and a day for day clause
- Three options, visible together, middle one expected to sell
- Payment tied to stages, with a launch readiness trigger
- Sales tax shown as a separate line at the client province rate
- Change order rate and process stated in the proposal itself
- Expiry date on the pricing
- One next step, one link, one named person
- No claim in the document you could not evidence if asked
Frequently asked questions
How long should a web design proposal be?
Six to eight pages. Long enough to carry a countable scope, a timeline and three options, short enough that a busy owner reads all of it. Length signals effort to the writer and signals indecision to the reader.
Should I put pricing in the proposal or wait until the client asks?
Put it in, on page one or two. Buyers scroll to price first whatever you do. Withholding it does not create suspense, it creates the impression that the number is negotiable or that you are not confident in it.
What deposit should I ask for on a website project?
Between 30 and 50 percent on signature is standard practice for project based web work, with the balance tied to design approval and launch. State plainly that the deposit becomes non refundable once discovery work starts, because that is the point at which your cost is real.
How do I handle a client who keeps adding small requests?
Send a change order for every one of them, including the ones you intend to absorb, priced at zero where you are absorbing it. The client sees the running value of what they are getting, the boundary stays visible, and nothing becomes a precedent by accident.
Which sales tax do I charge a client in another province?
Generally the rate of the province of the client physical address that you obtain in the ordinary course of business, with the contracting address treated as the most closely connected one where you hold several. That usually means an Alberta agency charges Ontario HST on work for a Toronto client. Confirm with your accountant, since the services rules carry special cases.
Do I need a separate contract if my proposal is signed?
Yes. A signed proposal can form an agreement, but it will not contain intellectual property assignment, liability limits, termination terms or a warranty definition. Have a contract that incorporates the proposal scope by reference, and have a lawyer review it once rather than improvising per project.
What is a reasonable proposal to close rate?
There is no published benchmark worth quoting, because close rate depends almost entirely on lead quality. The useful measure is your own trend after a change. Track proposals sent, proposals closed and average value monthly, change one variable at a time, and judge the document against your own baseline rather than someone else’s number.
The short version
A proposal that closes is a proposal that has already answered the questions the buyer was going to ask their accountant, their partner and themselves. Restate the problem in their words. Count every deliverable. Name what you are not doing. Put the dependencies on them in writing. Give three options and expect the middle one. Publish the change order rate before you need it.
The scope creep part is the same discipline running one step later. You are not trying to prevent change, you are trying to make change visible, priced and scheduled. Agencies that do this well end up with more revenue per project and fewer difficult conversations, which is the same outcome from two directions.
Want the proposal and scoping system built for you?
Wise Media builds the systems that agencies and founders run on, from the website itself through to the conversion system that turns traffic into scoped, qualified opportunities. If your proposals are closing below where they should, or every project is leaking hours after signature, that is a systems problem and it is fixable.
Start with the Wise Media intake form and tell us where the deals are stalling. We will come back with a scoped recommendation, not a brochure.
Related reading: booking calendar or contact form for how the first touch is captured, digital agency pricing in Canada for the number behind the options, and what it costs to take payments on a Canadian website for the deposit collection side.



