How I Estimate Junk Removal Jobs From Photos
(And Why I Stopped Driving to Every Quote)

Published April 2026 · Estimating & pricing

The driving problem

Early on I drove to every quote. Every single one. Neenah, Appleton, Fond du Lac, Berlin — my service area spans maybe 40 miles in any direction from Oshkosh. A quote in Waupaca was an hour round trip minimum. Ripon was 45 minutes. Even jobs across town were 20 minutes each way.

That's time I wasn't earning. It's fuel. It's wear on the truck. And most of the time, the customer had already decided whether to hire me or not based on the ballpark number I gave them on the phone. The site visit was just confirmation.

The shift happened gradually. I started asking customers to text me photos when they called. Not instead of a visit — just to save a trip if the job was obviously too small, too big, or not something I wanted. But pretty quickly I realized I could quote most jobs accurately from photos alone. The customer got a faster answer. I kept my evening. And my closing rate actually went up because I was responding in minutes instead of scheduling for two days out.

After 10 years of this, here's exactly what I look for when a customer sends me photos of their junk.

What I ask customers to send

I don't need professional photos. An iPhone in auto mode is fine. But I do need specific shots, and I'll text back asking for them if they send one photo of a corner.

My standard request: "Send me 3-5 photos from different angles. Stand in each corner of the room and take one toward the middle. If it's a garage, take one from inside the garage door looking back, one from each side wall, and one from the back wall looking forward."

That gives me overlapping views. Anything hidden from one angle shows up in another. If they only send one photo from the doorway, I can only see what's visible from the doorway. The back half of the garage, the stuff behind the couch, the boxes stacked against the far wall — all invisible. I'll tell them: "I can only quote what I can see. Can you send a few more from inside the room?"

Most people send good enough photos once you ask. The ones who send one dark, blurry photo from across the street are the same ones who ghost you after the quote anyway.

Finding reference objects

This is the core skill. When you look at a photo of someone's junk, your brain does something automatic: it sizes objects by comparing them to things it knows the size of. A door is 80 inches tall and about 32 inches wide. A standard interior door frame is the most reliable ruler in any residential photo — it's always there, always the same size, and it gives you both vertical and horizontal scale.

Here's what I use, in order of reliability:

Standard interior door. 80×32 inches. If a pile of debris comes up to the doorknob, that's roughly 36 inches tall. If it spans the width of the door, that's about 32 inches wide. Instant measurement.

Electrical outlets and light switches. Count them on a wall. Outlets are typically 12 inches from the floor, spaced about 6 feet apart horizontally. Light switches are about 48 inches high. If you can see two outlets on a wall behind a pile, you know that wall section is roughly 6 feet wide.

Step height. Standard interior steps are 7 inches tall. If a pile of construction debris comes up to the third step, that's about 21 inches.

Fence height. Residential chain-link fence is usually 4 feet. Privacy fence is 6 feet. If there's a fence in the background of an outdoor pile photo, that's one of your best scale references. Outdoor piles are notoriously hard to estimate because there's nothing indoor to anchor against. A fence solves that.

Cinder blocks. Standard cinder block (CMU) is 16×8×8 inches. If someone has cinder blocks in their pile or as a retaining wall in the background, you can measure against them.

Standard appliances. A typical refrigerator is about 36 inches wide and 68-70 inches tall. A washer is about 27 inches wide. These aren't as consistent as doors because appliance sizes vary, but they're close enough for volume estimation.

The trick is finding two reference objects in the same photo. A standard couch next to a pile gives you a rough width measurement just by comparing the two. The more reference points, the tighter your estimate.

The pile depth problem

This is the one that gets new operators, and it still gets me sometimes. When you look at a photo of a pile, you see the front. You don't see what's behind it or underneath it. That's not a small detail — it's often half the job or more.

I did a garage cleanout last month. The customer sent two photos: one from the garage door looking in, one from the side. From the front, it looked like a moderate pile — a desk, some boxes, a bed frame, a few bags. I gave them a quote.

When I got there, the pile went back another four feet. Behind the desk was a wooden shelving unit full of tools, two more bags, a nightstand, a pedestal fan, and a pile of clothing. The job was easily twice what I'd quoted. I still did it for the original price because I'd given my word, but I left money on the table that day.

That's the pile depth problem. You can't see what you can't see. But you can learn to expect it.

When I see a pile against a wall, I assume it goes back further than what's visible. If items are stacked higher than about four feet, there's almost always stuff behind and underneath that doesn't show in the photo. When a customer says "it's the whole garage," I mentally double whatever I'm seeing, because "whole garage" always means more than the front half.

Construction debris is the worst for this. Roofing shingles, for example. A photo of a shingle pile against a chain-link fence might show what looks like a manageable mound. But shingles are dense — they pack tight with almost no air space. A pile that looks like it'll fill a quarter of the truck bed might actually take up three-quarters. And shingles are heavy. Rain-soaked shingles weigh even more.

Jobs I got wrong (and what they taught me)

Here are the ones I remember, because they cost me money.

The garage with the back wall. Customer sent one photo from the driveway looking into the garage. I could see a couch, a few boxes, a lawn mower, and some scattered tools. Looked straightforward. Got there and the back wall of the garage was lined with shelves full of paint cans, chemicals, and car parts. That was a full truck load, not the quick run I'd quoted. Lesson: always ask for photos from inside the room, not just from the doorway.

The "small" basement. Customer said it was just a few boxes in the basement. Sent one photo that showed three boxes and a broken chair. I quoted my minimum. Walked downstairs and found an entire finished basement that had been gutted — drywall pieces, carpet remnants, a dismantled bar, couches, and construction debris everywhere. That was at least a full day's work I'd quoted as an hour job. Lesson: "a few boxes" means nothing. Ask for photos of every room they mentioned.

The apartment that was "almost empty." The landlord said the tenant had already moved most things out. Photo showed a kitchen with a few bags and a broken chair. The tenant had moved out — into the bedrooms. Both bedrooms were full. Living room was full. Closet was full. Balcony had stuff too. Lesson: "almost empty" from a landlord's perspective means "I already took the good stuff." Always ask for photos of every room.

The shingle pile from one angle. Photo showed a mound of shingles next to a shed. Looked manageable. Got there and the pile was against the shed wall, hiding a pile almost as big behind it. The weight was the real problem — shingles are heavy and I hadn't accounted for how much was actually there. Lesson: for outdoor piles, ask for photos from at least two sides. And always assume outdoor piles are bigger than they look.

Every one of these mistakes had the same root cause: I had incomplete information and I guessed instead of asking for more photos. Now I ask. Every time. Even when I'm busy. Even when the customer seems impatient. I'd rather lose a quote by asking too many questions than lose a few hundred bucks on a job I underbid.

When I still drive out

Photo estimates don't work for everything. I still visit the job site when:

Hoarding situations. Hoarding jobs are hard to photograph completely. There's no "good angle" because every angle shows more stuff. I need to walk through and count rooms, assess biohazard conditions, and figure out how many loads it'll take. No photo captures that.

Jobs with access questions. Fourth-floor walk-up? Steep driveway that the truck can't reach? Items that might require disassembly to fit through a doorway? These aren't volume questions — they're logistics questions that need eyes on site.

Commercial or contractor jobs. Construction debris from a job site, office cleanouts, property management foreclosures. These are bigger and more variable than residential cleanouts, and the quote is usually part of a bidding process where the customer expects you to show up.

When the photos are terrible. Dark, blurry, one photo of a far wall with nothing visible. I'll ask for better photos once. If they can't or won't send better ones, I'll drive out or pass on the job. Quoting from bad photos is worse than not quoting at all.

When something feels off. Sometimes the customer's description doesn't match the photos, or the photos seem staged, or the price they're expecting doesn't line up with what I'm seeing. That gut feeling has saved me from several jobs that would have been disputes. Trust it.

That said, I estimate maybe 80% of my jobs from photos now. The 20% I visit are the ones where the photo estimate would be a guess, and I've learned that guesses cost more than drive time.

Building a tool that does what my brain does

After doing this for years, the process became automatic. Look at the photo. Find the door. Size the couch. Spot the reference objects. Estimate the pile dimensions. Count the visible items. Add some cushion for what you can't see. Come up with a number.

It takes me maybe 30 seconds to do that. But I had 10 years of practice. New operators don't have that. They look at a photo of a cluttered garage and have no idea if that's a couple hundred bucks or most of a truck. They quote low because they don't want to lose the job, and they lose money on every one.

I built WhatShouldICharge to do what my brain does automatically. It takes the same reference points I look for — doors, outlets, fences, known appliance sizes — and uses them to measure items in the photo. It flags heavy items that need premium pricing and generates a price range based on the operator's own rates.

It's not perfect. Neither am I. But it gives a new operator a reasonable starting point in about 30 seconds instead of a drive across town. And for experienced operators, it's a second opinion — a documented estimate you can show the customer instead of a number you pulled out of thin air.

I still look at every estimate it generates and adjust when I know something the photos don't show. But I don't drive to every quote anymore, and neither should you.