A roofer we were talking to last month had a quote form on his homepage with eleven fields. Name, email, phone, address, city, state, ZIP, roof type, roof age, preferred contact, and a 500-character problem description. The form worked — people filled it out. Twelve times a month, on a site that was pulling in about 3,400 visitors.
Twelve quote requests, from 3,400 visitors, in a business where the owner was spending $4,200 a month on Google Ads. That is a 0.35% conversion rate on a page where the intent is "I need a roofer, right now."
We cut the form to one field.
One field
The field was ZIP code. You tap a five-digit ZIP, you get told whether Henderson Roofing can be at your house same-day or next-day, and you get a phone number with a "Call now" CTA under it. If you want to book without calling, there is a second screen with three fields: name, phone, and "briefly, what is going on." Most people call.
The month after we launched, the site did 41 quote requests. Same traffic, same season, same ad spend. The form itself was doing a slightly smaller fraction of those — because the bigger lift was on the calls — but the form conversions tripled.
This is not a case study about roofing. This is a case study about how much money gets left at the door by form fields that nobody asked for.
The seven-percent rule
There is a rough rule in conversion work that every additional field on a form costs you somewhere between seven and ten percent of the submissions you would have otherwise gotten. It is not a law of physics — the number depends on the field, the audience, the page, the price of the thing being bought. A two-field email signup does not lose seven percent of signups by adding a first-name field. A thirty-field mortgage application does not gain seven percent by removing one.
But for the kind of form that sits on a small-business homepage — "request a quote," "book a consult," "get in touch" — the rule is close enough to be useful. A six-field form is losing roughly half the submissions an equivalent one-field form would pull. A ten-field form is losing two-thirds.
We look at a lot of these forms. The median we see on a rebuild engagement is eight fields. Eight fields is usually the business's fault, not the form builder's. Someone, at some point, added a field because they got a lead they could not use, and never removed it.
What fields actually cost
Every field is a little question the visitor has to answer, in a context where the visitor is already partway to abandoning the page. Some questions are cheap:
- Name. Low friction. People give a first name without thinking.
- Email. Low friction. Most people have one handy.
- Phone. Medium friction. Some people do not want to be called.
- ZIP code. Low friction when it is useful — a service-area filter, a pricing tier. High friction when it feels like data collection.
Some questions are expensive:
- Address. High friction. Requires autocomplete or six more fields underneath. Almost never needed at the lead stage.
- Budget. Very high friction. People do not know, do not want to say, or will lie. Move this to the sales call.
- "Briefly describe your project." High friction. A blank text area is a 30-second wall. Most visitors will bounce rather than write a paragraph.
- Preferred contact method. Medium friction, and usually pointless — you are going to call them anyway.
The test, when deciding whether to keep a field, is simple: what would happen if this were missing and you followed up to ask? If the answer is "we would just ask on the sales call," cut it.
Why forms get bloated
Nobody designs an eleven-field form on purpose. The bloat is incremental.
The owner gets a lead who turned out to be twenty miles out of service area. "We need to ask for ZIP."
The owner gets a lead who wanted a color the company does not offer. "We need a color field."
The owner gets a lead who cannot actually afford the service. "We need to ask about budget."
Each of these is a real problem. But the problem is almost never solved by adding a field. The problem is solved by pre-qualifying on the page itself — "we serve Bucks and Montgomery counties," "we install architectural shingles only," "jobs start at $8,500" — or by having a thirty-second phone screen at the top of the funnel.
A form is a machine for losing leads at the exact moment they are most willing to become customers. Every field you add subtracts from the pile on the far side of it. That is the tax.
What to ship
When we rebuild a quote form, the default is two or three fields. One is something the site can respond to instantly — a ZIP, an address, a service. That field earns its place because the visitor gets something back: "we serve your area," "we can be there tomorrow," "that service is $3,500 to $5,000 depending on the scope."
The other fields are whatever the business genuinely cannot follow up without. For most service businesses, that is name, phone, and one short description — and even the description is optional, because the phone call will cover it faster than typing ever would.
Everything else moves to the call.
This is not a design preference. It is an arithmetic preference. A form with three fields converts somewhere between two and four times the same form with eight. A small-business website that gets twice as many leads is, approximately, a small-business website that is worth twice as much. There is no equivalent optimization anywhere else on the site.
The easiest SEO gain on most of these rebuilds is not a meta tag or a sitemap entry. It is the form. Fix the form and the site starts earning the traffic it was already getting.