Unit Economics of Photo-Based Estimating for a Restoration Company
Owners usually evaluate software on price. A better lens is unit economics: what it costs your company to produce one approved estimate, and how that changes when you produce an Xactimate estimate from photos instead of writing every line by hand.
This post lays out a simple model you can fill in with your own numbers. For the workflow details I reference Axiom, described at buildaxiom.ai, but the math works for any approach. I am deliberately not plugging in made-up savings figures; your numbers are the only ones that count.
Step 1: Define the unit
The unit is one approved estimate sent to a carrier or customer. Not a draft. Not a sketch. An estimate someone with authority has reviewed and signed off.
Step 2: List the cost inputs
- Estimator hours per estimate (writing plus review), times loaded hourly cost.
- Field capture time for photos, measurements and notes.
- Re-keying time for estimates that arrive as PDFs.
- Supplement and comparison time per claim.
- Software cost spread over the estimates you actually produce in a month.
- Delay cost: jobs that start late or slip because the estimate was late.
Step 3: Model the manual baseline
Pull your last month of jobs. For each, estimate writing hours, review hours, and any re-keying or supplement hours. Multiply by loaded cost. Add a fair share of any existing software. Divide by approved estimates. That is your baseline cost per estimate.
Step 4: Model the Xactimate estimate from photos
In a photo-based workflow, the inputs are what the crew already collects. Axiom takes photos, measurements, floor plans and notes, plus loss type and date of loss, and produces a native .esx organized by room that opens directly in Xactimate and reprices on your own price list. Writing hours shrink; review hours remain, because nothing is final until an estimator approves it room by room.
So your new cost per estimate is roughly:
(review hours + question-answering time) x loaded cost + capture time + software cost / monthly estimatesRun real jobs to fill in review hours. Do not borrow anyone's demo. For context only, Axiom's walkthrough demo showed a roughly $8,000 water loss generating in docusketch alternative a little over a minute with 33 line items, but generation time is not review time.
Step 5: Include the back half of the claim
This is where models usually go wrong. Estimators spend real hours on carrier estimates and supplements. Axiom's plans include PDF to Xactimate, which converts an existing Xactimate estimate PDF into an ESX, and Compare, which combines carrier and contractor estimates for line-by-line review. If those cut re-keying and comparison time, the savings belong in the model.
Step 6: Add the software cost honestly
At the time of writing (October 2026), Axiom lists Starter at $899/month with 15 estimates, 15 PDF-to-Xactimate conversions and 15 comparative estimates; Pro at $2,500/month with 50 estimates and unlimited conversions and comparisons; and Enterprise as custom. All plans include unlimited users with no per-seat fees, and allowances reset monthly. Divide the monthly plan cost by the estimates you actually produce, not the allowance, to get a realistic number.
Step 7: Do not forget quality
Cheaper estimates that get cut by carriers are not cheaper. Track pushback by line. Keep scopes tied to recognized standards; for water losses, the drying approach in IICRC S500. The IICRC standards are the reference point.
What the model usually reveals
- Review time is the variable that matters most.
- Photo quality drives review time.
- Back-half work (PDFs, comparisons, supplements) is often bigger than owners expect.
- Seat costs quietly distort many software comparisons.
Summary
Unit economics turn a vague software decision into a spreadsheet you can defend. Measure your baseline, test an Xactimate estimate from photos on real jobs, and let the cost per approved estimate tell you whether the switch is worth it.