A consulting firm we talked to last year quoted the same 6-week discovery engagement three different ways in one month: $18,000, $24,000, and $31,000, depending on which partner wrote the proposal. None of the three had a bad reason. They just didn't have a shared process, so every quote was really a guess dressed up in a PDF. That's the normal state for agencies, consultancies, and IT services shops that sell custom-scoped work instead of a fixed-price SKU. If your average deal needs a scoping call before you can even name a price, you don't have a sales process, you have a set of habits living in a few people's heads.
That's fixable without a 40-slide methodology. It takes four things done the same way every time: a scoping call that produces the inputs a quote actually needs, a pricing model that isn't improvised, a quote-to-signature handoff that doesn't let deals go quiet, and a system that remembers what was promised so delivery doesn't get blindsided six weeks in.
In this guide
Why custom quoting breaks down as you grow Running a scoping call that produces a real quote Pricing custom work without guessing What has to happen between the quote and the signature Frequently asked questions
Why Custom Quoting Breaks Down as You Grow
At 3 people, your sales process is whoever founded the company having a conversation and sending a number that feels right. That works, barely, because one person holds all the context. At 15 people it stops working: four different account leads are now quoting the same service categories at different margins, with different assumptions about how many hours a "typical" project takes, and nobody notices until finance runs a margin report and finds projects that lost money the day they were signed.
The root problem isn't sloppiness. Custom-quoted work resists templates by nature, every client's scope is a little different, so firms give up on process entirely and let each seller freelance it. The fix isn't forcing engagements into a fixed price grid, it's standardizing the inputs (the scoping call, the rate card, the approval step) while leaving the actual scope flexible. A CRM built for project-based selling earns its keep here: it gives every rep the same required fields before a quote can go out, which forces consistency without slowing anyone down.
Running a Scoping Call That Actually Produces a Quote
Most scoping calls fail quietly. The rep asks good questions, the client gives vague answers, everyone leaves feeling productive, and then the rep sits down to write a quote and realizes they don't know the three things that determine price: what "done" looks like, who on the client side needs to be involved, and what happens if scope shifts mid-project. Guessing fills the gap, and guessing is exactly how a $24,000 engagement becomes an $18,000 one for no reason other than which partner ran the call.
A scoping call worth the name gets five things on record before it ends:
- A written definition of done, in the client's own words if possible, so a checkout redesign doesn't quietly turn into the checkout, the account settings page, and the mobile app three weeks in
- Who the client-side stakeholders are, and whether the person on the call actually has budget authority
- A rough timeline constraint, since a 6-week ask compressed into 3 weeks changes staffing and price, and that needs to surface before the quote
- Which delivery roles the work needs, since a project staffed with two seniors prices nothing like the same scope staffed with one senior and a mid-level
- An explicit change-order trigger, agreed to before signature, so scope creep has a named process instead of an awkward conversation in week 4
Make these five a required field set in your CRM so no quote can go out without them. That one rule catches more margin-eroding quotes than any amount of sales coaching.
Pricing Custom Work Without Guessing
Here's an opinion that tends to annoy account leads at first and relieve them once they try it: nobody should be pricing custom engagements from memory. Memory is where a senior consultant who's underpriced the same service for two years keeps doing it, because the number "feels right" and nobody's checked it against actual delivered hours.
Price instead from a rate card crossed with your own delivery history. If your last 12 "brand refresh" engagements averaged 140 hours at a blended $165/hr rate, that's your anchor, not a partner's gut feeling. Adjust up for tight timelines or a client with a known history of scope creep, and adjust down only for a strategic reason (a logo you want, a foot in the door for a bigger account), not because the client pushed back and caving was easier. The firms that hold their pricing best can point to their own delivery data and say what a service actually costs to deliver.
The margin leak usually isn't the initial quote, it's the unbilled scope after it. A project quoted at 140 hours that actually takes 175 doesn't register as a pricing mistake, it shows up as a margin miss three weeks later, after the client has already been told the number won't change. Comparing quoted hours against logged hours in real time, inside the same system that generated the quote, is the difference between catching that at hour 150 and catching it at invoicing.
This is also where the gap between sales tools and delivery tools shows up. A quote built in one tool and hours tracked in another means nobody can check the quote against reality until it's too late to fix. Keeping the pipeline, the project, and the timesheet in one record is a practical reason firms move to a connected CRM and project system instead of stitching spreadsheets to a proposal tool.
What Has to Happen Between the Quote and the Signature
A surprising number of custom-quote deals die not because the price was wrong, but because the quote went out and nothing happened for 11 days. The client didn't say no, they got busy, and the rep had no system nudging them to follow up on day 3, 7, and 10. Custom-quoted cycles already run 3 to 8 weeks from first call to signature, so a process that lets deals go quiet loses more to inertia than to competitors. Fix it with a follow-up cadence tied to the quote itself, and a "quoted, awaiting decision" pipeline stage separate from "in discussion," so leadership can see how many dollars are sitting in limbo.
The bigger fix is the handoff once a deal does close. A signed quote that gets manually re-typed as a new project is where the original scope and pricing assumptions quietly go missing. Six weeks later nobody on delivery remembers the quote assumed 30-minute check-ins, the actual check-ins run an hour, and the project bleeds margin for a reason nobody can trace back to the sale. Firms with a good process close that gap by having the CRM and the project system share the same record, so the deal converts directly into staffed work instead of ending at signature. If quoting and delivery still live in separate tools, the plans and pricing page is a fast way to see what a connected setup costs to run instead.
See how a quote turns into a staffed project without re-typing anything
Get a live walkthrough of how Autovella connects your pipeline, quotes, projects, and time tracking into one record.
Frequently asked
A normal sales process moves a known product at a known price through a pipeline. Custom-quoted work requires a scoping step before a price even exists, so the quote itself is a deliverable that has to be produced consistently, not a number pulled from a price list. Most breakdowns in agency and consulting sales happen when firms run custom quoting through a process built for fixed-price selling.
Most engagements under roughly $30,000 can be scoped properly in a single 30 to 45 minute call with a checklist. Larger or more technical engagements often need a second call. The length matters less than whether the five key inputs, definition of done, stakeholders, timeline, staffing, and change-order terms, actually get captured before the quote is written.
Pricing from memory instead of from delivery history. Reps tend to anchor on what a similar project "felt like" it should cost rather than what past similar projects actually took in hours and cost to deliver. Without a rate card grounded in real data, underpricing repeats itself indefinitely.