Ask ten agency owners what the hardest part of the job is and most will say the same thing: pricing. Quote too high and the deal dies. Quote too low and you pay for the project out of your own margin. And even when you land somewhere reasonable, one scope change from a client who thinks âsmall tweakâ means âfive minutesâ can erase the profit anyway.
This guide pulls the entire âPricing is Artâ series into one place. Need to estimate a project from almost nothing? Decide between range and fixed pricing? Figure out why proposals keep stalling at the price page? Start here.
Part 1: How to estimate a project with barely any info
It starts the same way every time. A prospect emails with a vague idea, and you have to put a number on it before they lose interest and hire someone else.
The four-step framework
1. Educate. Assume the client has never bought custom software before. Keep a short glossary handy. Schedule milestone checks during the negotiation so they see the work in stages. Bring them into a team meeting once, so they understand why a âsmall changeâ rarely takes five minutes.
2. Explain. Put everything in writing. Annotations, links, comments, side notes. When youâre not sure something is clear, run it past a non-technical person first.
3. Gather, guess and estimate. Some gaps stay open no matter how much you ask. For those, guess conservatively: assume the worst case for the details you donât have, build in reserve time and budget for surprises, and research similar projects through blogs, colleagues, whatever industry data you can find.
4. Be fast. Send the proposal as a draft they can comment on, not a finished PDF. Start with wide brackets and mark them clearly as ballpark figures. Donât wait for perfect information. Collect it in parallel while the conversation is still warm.
When to walk away
If three rounds of questions still leave you guessing at scope, the project isnât ready to be priced. Offer a paid discovery phase instead.
Part 2: When to use range pricing (and when not)
Brackets are a strong tool, and easy to misuse. There are six common approaches; most of them are wrong for your situation.
The six approaches to range pricing
| Approach | What it is | When to use |
|---|---|---|
| Looking for a fool | Lowest bracket = real price; higher brackets are bluffs | Never. It destroys trust |
| The fake brackets | You have a target price but widen the range artificially | Only when testing a new market |
| Reverse psychology | Comically wide brackets ($15-$300/hr) to make the client propose | When you have no market data at all |
| The negotiator | High prices with built-in discounts; the low bracket is your walk-away point | When you expect heavy negotiation |
| Market positioning | Use the industry min and max as your brackets | When entering a new segment |
| Socrates | Honest brackets based on actual uncertainty | This is the one you want |
When brackets actually work
Use them when you depend on subcontractors or external experts, when you have data but no precise number, or when the clientâs documentation is solid and the scope is unlikely to explode.
Skip them when your data is shaky (a wide range doesnât rescue a bad estimate), when the client demands an exact figure, or when the gap between min and max runs past 25%. In that last case, scope is the real problem.
The 25% rule
If your top estimate sits more than 25% above your bottom one, the requirements are still too vague to price. Go back and refine the scope before you send any number.
Part 3: Pricing models vs. pricing strategies
A model is what you charge. A strategy is how you justify it. Mixing the two up is where most agencies lose money.
Common pricing models
| Model | Works best for | Risk |
|---|---|---|
| Fixed price | Well-defined scope, low uncertainty | You eat the overruns |
| Time & materials | Evolving requirements, agile projects | Client carries the uncertainty |
| Retainer | Ongoing support, maintenance | Predictable revenue, capped upside |
| Value-based | High-impact projects, proven track record | Hard to quantify the value |
| Hybrid (fixed + T&M) | Most real-world projects | Needs clear scope boundaries |
Pricing strategies that actually work
Anchor high, negotiate down. Your first number sets the ceiling. A client who talks you down from $100k to $85k walks away feeling like they won. Show $85k first and negotiate to $70k, and the same deal feels like a loss.
Tier your offerings. Three options: good, better, best. Most clients pick the middle, so put the margin where you want the deal to land. The âbestâ option exists mostly to make the middle look reasonable.
Price the outcome, not the hours. If a feature saves the client $50k a month, 40 hours of development time is the wrong yardstick. Quantify the value and charge against it.
Part 4: Measuring how proposals affect conversion
You canât fix a leak you never see. Proposal performance is trackable, and itâs worth the effort.
What to track
- Open rate. Did the client even open it? Web-based proposals make this visible.
- Time spent. How long on each section?
- Drop-off point. Where did they stop reading? Usually the price page.
- Follow-up actions. Questions, change requests, or silence.
- Close rate by proposal type. Do interactive proposals beat PDFs? Spoiler: yes.
How to track it
| Method | What you get | Effort |
|---|---|---|
| Email tracking + follow-up | Basic open/click data | Low |
| Document hosting (Google Docs, Office 365) | Edit history, time spent | Medium |
| Web-based proposal platform | Full analytics, session recording | Best |
Using the data
Session recordings are the most useful thing on that list. Watch where clients hesitate, re-read, or leave. Churn at the price page means your pricing presentation needs work. Ten minutes stuck on the scope section means the scope is written too densely.
The feedback loop
Every proposal produces data, and every closed deal teaches something. Capture it. After a lost deal, ask plainly: price, scope, or timing? After a won one: what actually convinced them? Then wait. Collect ten deals or more before you change the process, because one lost deal proves nothing.
Part 5: Three steps to overhaul your pricing
Step 1: Audit what you charge now
Pull the numbers from your last 20 projects. Estimated versus actual hours. Margin by project type. Change-request volume by client. Win rate by price range. Most agencies find a few project types that print money and a few that quietly bleed. Push toward the first kind.
Step 2: Build pricing on that data
Turn the audit into tiers. Small projects under $20k: fixed price, tight scope. Medium projects between $20k and $100k: hybrid, with a risk buffer. Large projects above $100k: T&M or value-based, paid in milestones.
Step 3: Test and iterate
Pricing is never done. Run experiments. Raise rates 10% for your next five prospects and watch the win rate. Swap PDFs for interactive proposals on the next ten deals and compare close rates. Add a transparent risk buffer to every quote and see who pushes back.
The bottom line
Pricing isnât a math problem. Itâs a strategy problem. The agencies that win arenât the cheapest; theyâre the ones who know what their work is worth, track their own numbers, and understand how their clients think.
Start with the four-step estimation framework. Add range pricing where it fits. Track everything. And keep iterating, because the market wonât wait for you.
Price with confidence
Turn pricing from a guess into a process.
Apropo connects estimates, ranges and proposal tracking so every quote builds on real numbers, not gut feel.
See how Apropo protects marginFor a closer look at how automated estimation tools speed up pricing, see our guide on how to estimate an IT project.