Every business that quotes custom work has a quoting spreadsheet, and every quoting spreadsheet has a person. The person knows which tab to update when steel goes up, why the labour rate on row 41 is different, and that you have to copy the file rather than open the original because the macros do something odd otherwise. Quotes go out on time while that person is available. When the business grows, or the person is away, or a second estimator starts and produces different numbers for the same job, the spreadsheet stops being a tool and becomes a risk. What replaces it depends on who needs to produce the quote and how much of your competitive edge is in the pricing logic itself.
Quick answer: A spreadsheet is fine until more than one person quotes, quotes need version history, or pricing changes break formulas. Off-the-shelf CPQ software suits businesses with a product catalogue and rules that fit its model. A customer-facing configurator suits standardised custom products where the customer can specify the job. A custom quoting app is right when the pricing logic is the business’s advantage and it doesn’t fit anyone else’s template. Most job shops we see land on the last option, built small.
When has the spreadsheet stopped working?
The signs are consistent. Two estimators quote the same job and get numbers that differ by more than the margin. Nobody can say which version of the spreadsheet a quote from last March was made with, which matters when the customer accepts it in September. A material price change means editing formulas across a dozen tabs, and one of them gets missed. The file has grown to the point where it takes a minute to recalculate and someone has turned automatic calculation off, so numbers on screen are sometimes stale. Quotes exist as PDFs in email and nowhere else, so win rates and margins by job type can’t be reported without a manual trawl.
Most of these come down to a spreadsheet having no concept of a record: there’s no “quote 1042, version 3, sent Tuesday, accepted Friday”, there’s a file. Our post on when to replace a spreadsheet workflow covers the general version of this; quoting is the case where the cost of the spreadsheet is most directly visible, because a wrong quote is either lost work or lost margin.
What does off-the-shelf CPQ offer?
Configure-price-quote software, from Salesforce’s own product down to small tools aimed at manufacturers, gives you a catalogue, rules for which options go together, price books, discount approval workflows, and quote documents, usually tied to a CRM. It’s the right answer when your products are configurable from a defined set of options and your pricing follows rules the tool can express: base price plus options, volume breaks, regional price books.
Where it fits badly is the true job shop, where every quote involves estimating labour hours from a drawing, material from a cut list, and setup from experience. The tools can be made to do this, with effort, and the result is often a configurator that estimators work around. Pricing is also a factor: enterprise CPQ is priced for enterprise sales teams, and the small-business tools tend to assume a catalogue rather than a workshop.
When does a customer-facing configurator make sense?
When the customer can specify the job without a conversation. Cabinets in standard sizes with a choice of finish, signage, printed goods, fabricated parts from a set of templates. The configurator lives on the website, the customer chooses options, sees a price, and either orders or requests a formal quote. It’s a sales tool as much as a quoting one; it removes the wait for a human, and it filters out the enquiries that were never going to buy.
The trap is putting a configurator in front of work that needs judgement. If the estimator regularly looks at a request and says “that won’t work, but this would”, a configurator will quote things you can’t make at prices you can’t hold. Configurators are for the standard part of the range, and the estimator still quotes the rest.
When should you build a custom quoting app?
When the pricing logic is yours. If the way you estimate a job, the rules of thumb for setup and waste, the adjustments for a difficult customer or a rush, the knowledge of which supplier’s price to use this month, is the thing that makes your quotes both competitive and profitable, then that logic is an asset and it belongs in something you own and can change.
A custom quoting app doesn’t need to be large. The ones we build for job shops usually have: a quote record with versions and a status, line items built from a materials list with current prices, labour estimated from a small set of operations with rates, margin and rounding rules, a PDF that looks like the company’s quote, and a conversion step that turns an accepted quote into an order or a job. A rules editor so the owner can change rates without a developer. Reporting on win rate and realised margin by job type, which is the report that pays for the whole thing, because it shows which work to chase.
The cost lands in the range covered in our post on what a custom web application costs, toward the lower end for a quoting tool with a few integrations. Integration with the accounting system for invoicing and with whatever tracks production is where the scope grows, so decide up front whether phase one includes them.
How do you choose?
Ask three questions. Can a customer specify the job without talking to us? If yes for a meaningful share of enquiries, a configurator for that share. Does our pricing follow rules a catalogue tool could hold? If yes, look at CPQ tools and expect to adapt to them. Is the estimating judgement the thing we’re better at than competitors? If yes, build, and build small. Many businesses need two of the three: a configurator for the standard range and a custom estimating tool for the rest, sharing the same price data.
Whichever you choose, get the pricing data out of the spreadsheet and into one place first. Material prices, labour rates, and margin rules in a table that every tool reads from. That’s the step that removes the “which version” problem regardless of what sits on top.
The estimating tool for a fabricator or a print shop is one of the projects we do most often under custom software development for small business, and it usually starts with a morning spent watching the estimator quote three real jobs.
Frequently asked questions
Can we keep the spreadsheet and add a front end?
For a while. Tools that put a form in front of a spreadsheet exist, and they solve the multiple-estimators problem without solving versioning or reporting. They’re a reasonable stopgap and a poor destination, because the formulas are still the fragile part.
How long does a custom quoting app take to build?
A first usable version for a job shop is typically four to eight weeks, with the estimating rules being the slow part, since they have to be extracted from the estimator’s head and the spreadsheet’s formulas and then checked against past quotes.
Should the quoting tool connect to our accounting system?
Eventually. Pushing an accepted quote to an invoice or a sales order saves re-keying and removes a class of errors. It’s also where scope grows, so do quoting first, prove the numbers, then integrate.
How do we check the new tool’s prices are right?
Re-quote twenty past jobs with known outcomes in the new tool and compare. Differences point at rules the spreadsheet had that nobody mentioned. Run both in parallel for a month before retiring the spreadsheet.
Can a configurator handle quantity breaks and rush pricing?
Yes, both are standard. What configurators handle badly is anything requiring a judgement about feasibility, which is why they suit the standard part of a range rather than the whole of it.
Who should own the pricing rules?
The business, in a form the owner or head estimator can edit without a developer. If changing a labour rate requires a code change, the tool has the same problem the spreadsheet had, with a slower person to call.