Perfex CRM Estimate vs Proposal: Which One to Send
Perfex CRM estimate vs proposal, settled: which document each situation needs, exactly what survives every conversion to an invoice, and what does not.
A client approves a $6,400 rebuild on a Thursday afternoon. The quote went out as a proposal, because it needed a scope section, three phases and a page arguing against the cheap option. You open it, hit convert to invoice, and get four line items and none of the six paragraphs the client actually agreed to.
Nothing is broken. The Perfex CRM estimate vs proposal decision is a choice between two genuinely different objects, and most teams make it by reflex at the moment they need a document, then meet the consequences weeks later at conversion time. The two look interchangeable in the Sales menu. They are not.
The difference that settles most cases: who you can address it to
A proposal can be related to a lead or a customer. An estimate can only be related to a customer.
That single constraint decides more real cases than any feature comparison. At the point in a deal where you are still selling, the other party frequently does not exist as a customer record yet. Choosing an estimate forces you to create one, and now your customer list contains people who never bought anything. Every count on the dashboard, every bulk portal invite and every report grouped by customer inherits that pollution, and it is tedious to unwind later, because deleting a customer that already has documents attached is not a one-click operation.
So the first question is not about content or formatting. It is: does this recipient already exist as a customer, and are you confident they will be? If the answer is no, you want a proposal, whatever the document looks like.
What is actually inside each document
Beyond addressability, the two objects differ in shape. Perfex moves these screens around between releases, so treat the labels below as a map rather than a screenshot of your install.
| Proposal | Estimate | |
|---|---|---|
| Related to | Lead or customer | Customer only |
| Body | Full rich-content area around the item table: headings, images, tables, merge fields | Item table, plus a short client note and terms |
| Deadline field | Open till | Expiry date |
| Statuses | Draft, Sent, Open, Revised, Declined, Accepted | Draft, Sent, Expired, Declined, Accepted |
| Recipient can | Accept, decline, and comment in a thread | Accept or decline |
| Converts to | Estimate or invoice | Invoice |
The load-bearing row is the second one. A proposal has somewhere to put an argument. An estimate has somewhere to put a total. Almost everything else follows from that.
The status sets differ in one telling way too. An estimate can go Expired, because it is a priced offer with a shelf life. A proposal can go Revised, because it is a negotiation with versions. If your sales motion involves three rounds of back and forth about what is in scope, the object that models revision is the one you want.
Perfex CRM estimate vs proposal: a five-second decision rule
Ask two questions, in order.
- Is the recipient a customer record? No, and not certainly going to become one > proposal. There is no second question.
- Does the document have to argue, or only to total? Argue > proposal. Total > estimate.
That is the whole rule. Applied honestly it produces some outcomes teams resist:
- A repeat client asking “what would two extra seats cost” gets an estimate, even though your template library is full of handsome proposals. They are already sold. A proposal here costs you the clean estimate-to-invoice hop and buys nothing.
- A tender or RFP response gets a proposal, even if the buyer claims to care only about the number, because the supporting material has to travel with the price and stay revisable.
- A public-sector buyer who needs a formal quotation reference to raise a purchase order gets an estimate, because what they are filing is a number and its reference, not your positioning.
The anti-pattern is using proposals for everything because they look better. Presentation is not free. You are choosing a document whose body content cannot survive conversion.
What survives each conversion, and what silently does not
This is where the choice actually gets paid for.
Estimate to invoice
The clean hop. What crosses: the customer, the currency, every item line with its description, quantity, rate and unit, per-line tax rates, the discount, the project link if there is one, the sale agent, and the terms and client note.
What does not cross, and surprises people:
- The number. The invoice draws from the invoice series, not the estimate series.
EST-000148does not becomeINV-000148, and assuming it will eventually causes a reconciliation argument with somebody in accounts. - The dates. The invoice date defaults to today and the due date is recalculated from your payment terms. The estimate’s expiry date is irrelevant to the invoice and is carried nowhere.
- Attachments. Files uploaded to the estimate stay on the estimate. If the client signed off against a spec PDF attached there, that PDF is not on the invoice.
- The activity trail and staff notes. They remain on the estimate record, which is correct but not where people go looking.
What is preserved is the link between the two. An accepted estimate keeps a reference to the invoice it produced, and that is the only thing that lets you walk backwards from a payment to the offer that was accepted. Do not delete estimates after invoicing to tidy the list.
Custom fields deserve a test rather than an assumption. A value carries across the hop only when a field of the same name exists on the target object. A PO Number captured on the estimate but never defined on invoices stays on the estimate, and the invoice your client’s accounts payable system reads has no purchase order on it. Create the field on both objects, push one throwaway record through the conversion, and look at the result before you trust it with a real deal.
Proposal to estimate or invoice
The line items cross. The content body does not, because neither target object has anywhere to put it.
That is a consequence of the data model rather than a defect, but the effect is real: if the wording of the scope is what the client agreed to, it now lives only on the proposal. Two workable responses:
- Keep the proposal permanently and reference its number in the invoice’s client note, so anyone auditing the invoice can find the agreed scope in one search.
- Put anything binding in a contract instead and treat the proposal as a sales document rather than a legal one. This is the better answer for retainers and for anything with a term.
A proposal addressed to a lead
An invoice needs a customer, so a lead proposal has one extra hop: convert the lead to a customer, then invoice. Do that through the CRM’s own lead conversion rather than hand-creating a customer with the same company name, or the lead’s notes, activity and attachments are stranded on a record nobody opens again, and you have quietly created a duplicate.
The pipeline number that goes wrong when you use both
Each list screen shows totals by status, per document type. Those totals do not know about each other.
A $6,400 deal that lived as an open proposal, then a sent estimate, then an unpaid invoice appears in three separate summaries. Anyone adding open proposals to sent estimates for a pipeline figure double counts every deal that took the two-hop route, and the more disciplined your team is about converting, the more wrong that number gets.
Pick one document as the pipeline of record, and write the choice down:
- If your team sends proposals and converts them straight to invoices, proposals are the pipeline, and estimates are an exception you can leave out of reporting.
- If proposals are reserved for competitive pitches and everything routine is quoted by estimate, estimates are the pipeline.
Then run the check that keeps finding real money: filter estimates by Accepted and look for ones with no invoice against them. An accepted estimate that never became an invoice is agreed revenue nobody billed. On an install with a few hundred documents this takes about five minutes, and it is rarely an empty list.
Settings to fix before the next quote goes out
Under Setup > Settings > Sales, four decisions are worth making deliberately rather than inheriting:
- Separate number formats and prefixes for estimates, proposals and invoices. If two series share a format, somebody eventually matches the wrong pair during a reconciliation.
- Expiry and open-till defaults. A quote with no deadline is a price you have promised indefinitely, and it disables the Expired status that would otherwise tell you the offer is stale. Fourteen or thirty days, chosen once.
- Auto-convert accepted estimates to invoices, on or off, on purpose. On means nothing is forgotten, but the invoice is dated at acceptance whether or not the work has started, which is wrong for deposit-based projects. Off keeps control and hands you the monthly accepted-with-no-invoice check above.
- Default terms and client notes per document type, so nobody retypes payment terms into a line item description where no report will ever find them.
Do this on the next quote
- Check whether the recipient exists as a customer record. If not, and you are not sure they will, it is a proposal - stop there.
- Decide whether the document has to argue or only to total.
- If it is a proposal containing anything binding, decide now where that text lives after conversion: a retained proposal, or a contract.
- Create matching custom fields on both objects for any value that must reach the invoice, and test the conversion once with a throwaway record.
- Set the expiry date. Every time.
Before any of that, open the estimates list, filter by Accepted, and sort by date. Every accepted estimate more than a month old with no invoice against it is work you have already won and never asked to be paid for. Clear that list first. The taxonomy argument can wait a week.