Most SaaS teams underbudget file conversion, and the reason is not that the prices are hidden. It is that flat-rate plans are sold in large steps, so your bill does not track your usage - it tracks which step you have landed on. Crossing one conversion over a plan boundary can multiply your cost fivefold.
This guide computes real cost per MAU from enforced quotas, using ChangeThisFile's published ladder as the worked example, and tells you where flat-rate pricing stops being the cheap option. All figures are as of 26 July 2026 - check current rates before you commit a budget.
TL;DR - budget per MAU by product stage
The enforced ladder, which is what your bill is actually determined by:
| Plan | Price | Conversions/month | Per conversion at full use |
|---|---|---|---|
| Free | $0 | 25 | — |
| Hobby | $29 | 1,500 | $0.0193 |
| Startup | $99 | 6,000 | $0.0165 |
| Scale | $499 | 40,000 | $0.0125 |
| Growth | $1,999 | 200,000 | $0.0100 |
| Enterprise | custom | above 200,000 | not published |
Applied to product stages, where "conversions/user" is per active user per month:
| Stage | MAU | Conversions/user | Monthly conversions | Plan needed | Cost/MAU |
|---|---|---|---|---|---|
| Pre-launch / beta | 0-500 | 2 | 0-1,000 | Hobby $29 | $0.058 and down |
| Early traction | 500-2,000 | 3 | 1,500-6,000 | Hobby to Startup | $0.050-$0.058 |
| Growth | 2,000-10,000 | 5 | 10,000-50,000 | Scale $499 to Growth $1,999 | $0.050-$0.250 |
| Scale | 10,000-50,000 | 5 | 50,000-250,000 | Growth $1,999, then Enterprise | $0.040-$0.200 |
Rule of thumb: budget $0.05-$0.25 per MAU per month for a product where conversion is a supporting feature, and $0.25-$0.70 where it is the core workflow. Figures an order of magnitude below that - the "one cent per user" numbers that circulate - come from dividing a plan price by its quota rather than by the conversions you will actually perform.
The most important line in your model is not volume, it is the plan boundary. Between 6,000 and 40,000 conversions there is no intermediate plan, so 6,001 conversions costs $499. Model the boundary you are approaching, not your average.
Estimating your conversion volume per MAU
The number that drives everything: how many conversions does one active user perform per month?
Low intensity (document SaaS, HR tools, reporting): 1-3 per MAU. Conversion supports the product rather than being it - an occasional export or upload.
Medium intensity (content tools, real estate, e-commerce): 5-15 per MAU. Converting listing images, generating PDFs from templates, processing uploaded assets.
High intensity (file tools, media production, developer tooling): 20-100 per MAU. Conversion is why users showed up.
Worked examples, priced against the real ladder:
- HR document tool - 5,000 MAU × 2 = 10,000/month. Startup stops at 6,000, so this is Scale at $499 → $0.100/MAU. Note the trap: at 3,000 MAU the same product needs 6,000 conversions and pays $99. Adding 2,000 users multiplies the bill by five.
- Real estate platform - 2,000 MAU × 8 = 16,000/month → Scale $499 → $0.250/MAU.
- Media production SaaS - 3,000 MAU × 30 = 90,000/month → Growth $1,999 → $0.666/MAU. At this intensity conversion is a genuine COGS line and belongs in your pricing.
Each of those is roughly ten times what a naive plan-price-over-quota calculation suggests, because none of them consumes its full quota. That gap between what you buy and what you use is the single largest source of error in conversion budgets.
Budget line items beyond the API cost
Development time (one-time). Integrating a conversion API takes 2-8 hours depending on complexity. At $75/hour that is $150-$600, or $12-$50/month amortised over the first year.
Failed conversions still consume quota-adjacent effort. Two to five per cent of conversions fail in production on malformed files, unsupported edge cases and timeouts. Budget retry logic and user-facing error states, and add headroom above your expected count.
Hard stops, not overage. This is a budgeting fact, not a technical one. ChangeThisFile bills no overage on any plan: once the monthly quota is spent the API returns HTTP 429 until the period rolls over. That caps your downside - you cannot get a surprise invoice - but it converts a spike into user-visible failure. Metered providers make the opposite trade. Decide which failure mode your product can tolerate, and size for your peak month rather than your average either way.
Storage for outputs. If you keep converted files rather than streaming them through, add object storage: 1 MB average × 100,000 conversions is 100 GB a month, roughly $2.30 on S3. Usually negligible, but it compounds if you never expire anything.
Browser-side conversion is a real lever - but you build it. Common image, font and data-format conversions can run in the user's browser with open-source JavaScript libraries, costing you nothing per conversion and never uploading the file. That is worth auditing for. Be clear about what it is, though: the 193 client-side routes on changethisfile.com belong to its own web tool and are not exposed through /v1/convert, so this is an architecture decision you implement in your own front end, not a discount you can switch on.
Plan selection by volume - and when flat rate stops winning
Under 25/month: the free tier. Enough to validate an integration, not enough to run a beta.
Up to 1,500/month: Hobby, $29. Effective rate $0.0193 at full use.
Up to 6,000/month: Startup, $99. Effective rate $0.0165.
Up to 40,000/month: Scale, $499. This is the widest and most expensive step on the ladder - anything from 6,001 to 40,000 conversions costs the same $499, so a product at 8,000 conversions pays $0.062 each while one at 40,000 pays $0.0125.
Up to 200,000/month: Growth, $1,999.
Above 200,000/month: there is no published price. Budget a sales conversation, and get a quote before you build a financial model that depends on it.
When to price a metered provider instead
Flat-rate pricing is a good deal when you fill the quota and your conversions are slow or heavy. It is a poor deal when you sit just above a boundary, or when your conversions are quick. Metered competitors bill for work done rather than for a step you landed on, and at typical volumes they are cheaper per conversion - CloudConvert's subscription rate curve and Convertio's prepaid API minutes both undercut the ladder above for fast, ordinary conversions.
The honest test is your average processing time per file. If your conversions finish in seconds, a metered provider will almost certainly cost less and you should price one. If they run into minutes - video, large documents, heavy PDFs - the flat unit wins, because you pay the same for a four-minute job as for a four-second one. And if your volume sits in a plan gap, compare against metered pricing before buying the next step up.
COGS or operating expense?
For most products conversion is a COGS line - it scales with users, like hosting or payment processing. Treat it that way:
- Track conversions per paying user, not per MAU, so the cost is measured against the revenue it supports.
- Above roughly 5% of ARPU, act: audit conversion types, move what you can into the browser, cache aggressively, or price a metered provider.
- Below 1% of ARPU it is noise. Do not spend engineering time on it.
Comfortable example. $29/month ARPU, 1,000 paying users, 5 conversions each = 5,000 a month on the $99 Startup plan. That is $99 against $29,000 of revenue - 0.34%. Ignore it.
Uncomfortable example. $9/month ARPU, 5,000 paying users, 30 conversions each = 150,000 a month, which is the $1,999 Growth plan against $45,000 of revenue - 4.4%. That is at the threshold where it belongs in pricing discussions, and it is the realistic shape for a low-ARPU, conversion-heavy product.
Red flag. Above 5% of revenue, something is structurally wrong: you are in a plan gap, converting things that could run client-side, re-converting files you could cache, or on the wrong billing model entirely.
Budget model for three product archetypes
Archetype 1: document-focused SaaS (HR, legal, finance)
Core feature: generate PDFs, export spreadsheets, accept DOCX uploads.
Intensity: low, 2-5 per MAU.
At 1,000 MAU × 3 = 3,000/month → Startup $99 → $0.099/MAU. At 5,000 MAU × 2 = 10,000/month → Scale $499 → $0.100/MAU.
Archetype 2: media-heavy product (real estate, content marketing, e-commerce)
Core feature: image processing, thumbnails, document export.
Intensity: medium, 5-20 per MAU.
At 2,000 MAU × 8 = 16,000/month → Scale $499 → $0.250/MAU. Moving common image conversions into the browser is the highest-leverage saving available to this archetype, and can move it a whole plan step down.
Archetype 3: developer tool or file-first SaaS
Core feature: conversion is the product.
Intensity: high, 30-100 per MAU.
At 3,000 MAU × 30 = 90,000/month → Growth $1,999 → $0.666/MAU. Above 200,000 conversions there is no published plan at all, so model Enterprise as an unknown and get a quote early. At this intensity, price conversion into the product and compare metered providers seriously.
Complete budget sizing table
| MAU | Conversions/MAU | Monthly conversions | Plan required | Monthly cost | Cost/MAU |
|---|---|---|---|---|---|
| 500 | 2 | 1,000 | Hobby (1,500) | $29 | $0.058 |
| 1,000 | 1 | 1,000 | Hobby (1,500) | $29 | $0.029 |
| 1,000 | 5 | 5,000 | Startup (6,000) | $99 | $0.099 |
| 2,000 | 3 | 6,000 | Startup (6,000) | $99 | $0.050 |
| 2,000 | 5 | 10,000 | Scale (40,000) | $499 | $0.250 |
| 5,000 | 8 | 40,000 | Scale (40,000) | $499 | $0.100 |
| 10,000 | 5 | 50,000 | Growth (200,000) | $1,999 | $0.200 |
| 25,000 | 8 | 200,000 | Growth (200,000) | $1,999 | $0.080 |
| 50,000 | 10 | 500,000 | Enterprise | not published | — |
Read the two rows at 2,000 MAU together - they are the whole lesson. Going from 3 to 5 conversions per user takes you from 6,000 to 10,000 conversions, which crosses the Startup ceiling, which takes the bill from $99 to $499 and the cost per MAU from $0.050 to $0.250. Nothing about the product changed by five times. Model the boundaries.
No plan bills overage: every one hard-stops at its quota with HTTP 429. Budget for your peak month, not your mean.
File conversion is a predictable COGS line, but only if you model the right thing. It does not scale smoothly with usage on flat-rate pricing - it scales in steps, and the step between 6,000 and 40,000 conversions is where most budgets break. Work out which boundary your growth curve hits next quarter and price that, not today's average.
Two failure modes to avoid: over-engineering early by self-hosting before you know your real volume, and assuming a cent per user because someone divided a plan price by its quota. Instrument conversions-per-active-user from day one - it is the only input that matters, and it is the one nobody measures until the bill arrives. A free key gives you 25 conversions a month to validate the integration; use real usage data, not estimates, to pick the plan.