6 stages. Every order. No exceptions.
Our eBook Conversion Quality Standards — 6 Stages From Manuscript to Delivery
Most conversion services run your manuscript through software and deliver what comes out. We do not work that way. Every file that leaves Ebookifi has been through a six-stage quality process that combines human review, professional validation tools, and a final check against your original source material.
We are not a conversion house using automated pipelines. We are specialists who build eBook files by hand and verify them at every stage. The process below applies to every conversion order — from a short novella to a 500-page multilingual manuscript. No exceptions.
The 6 Stages
1Text Extraction & Source Review
2Hand-Coded Conversion
3PDF Comparison Check
4Code Validation
5Device & Platform Check
6Final Human Review
Every stage applies to every order. Rush delivery reduces time between stages — not the stages themselves.
The Six-Stage Quality Process
Stage 1 — Text Extraction and Source Review
The first stage is understanding what your manuscript contains before any conversion begins.
A specialist opens your source file and reads through it — identifying your paragraph styles, heading hierarchy, footnotes, tables, image placement, special characters, foreign language sections, and any formatting elements that require specific handling in the target format.
PDF Sources
We do not run your file through an automated OCR engine and hope for the best. Every page is reviewed by a human — catching editing issues, formatting inconsistencies, and content problems before they become problems in the converted output.
Word Sources
We review the paragraph style structure before beginning. An improperly styled Word document produces a poorly structured eBook regardless of how carefully the conversion is done. Structural issues are flagged before starting — not after.
InDesign & Other
InDesign, Pages, and HTML source files are reviewed for style consistency and structural integrity before conversion begins. We identify export settings and any elements requiring special handling.
Stage 2 — Hand-Coded Conversion
No automation software runs on your file at any point during conversion.
A specialist builds your eBook file from scratch in HTML and CSS. No Calibre export. No Sigil auto-conversion. No template-based processing pipeline. Every structural element — paragraph classes, heading levels, image handling, footnote coding, table structure, list formatting — is coded by hand to the current published specification of your target format.
Hand-coded files are smaller, cleaner, and more reliable than files produced by automated tools. A typical hand-coded Kindle file from a 300-page manuscript runs 800KB to 1.2MB. The same manuscript processed through an automated converter often returns two to three times that size — full of redundant code the device must process but never uses.
Stage 3 — PDF Comparison Check
For manuscripts submitted as PDF, this stage is unique to Ebookifi. No other conversion service we are aware of includes this as a standard step.
After conversion is complete, every page of the converted output is compared against the original PDF source file using professional PDF comparison software. The comparison flags any difference between the source and the output — cut text, missing paragraphs, spacing changes, altered punctuation, or any content that did not transfer cleanly during extraction.
Every flagged difference is reviewed manually. Most are intentional formatting changes that are correct for the target format. Any that are genuine content errors are corrected before delivery.
Why this step exists:
PDF text extraction — even when done carefully — can occasionally drop a line at the bottom of a page, merge two paragraphs, or misinterpret a special character. The PDF comparison catches these errors systematically, not by chance. This is why authors who have lost content in conversions elsewhere find no such problems in their Ebookifi output.
Stage 4 — Code Validation
Every ePub file is validated with the current release of EPUBCheck — the W3C’s industry-standard validation tool used by Apple Books, Kobo, Google Play Books, Barnes and Noble, IngramSpark, and every other major ePub retailer — before it leaves us.
We do not deliver files with warnings. We do not deliver files with non-critical errors that “probably won’t matter.” We deliver files with zero errors and zero warnings. If a file does not pass clean, we fix it before delivery.
Stage 5 — Device and Platform Compatibility Check
Passing validation confirms the file is technically correct. It does not confirm the file reads correctly on the devices your readers actually use.
After validation, your converted file is checked for correct display on the most common reading environments:
- Kindle Paperwhite, Kindle Scribe, Kindle Basic
- Kindle app — iOS and Android, current versions
- Apple Books — iPad and iPhone
- Kobo Clara, Kobo Libra
- Android e-reader apps — current versions
- Metadata — all fields confirmed at correct spec
We check the visual output — fonts, spacing, image placement, TOC navigation, footnote pop-ups, chapter headings — not just the file structure. A file can pass EPUBCheck and still display incorrectly on older Kindle hardware. We check both. Cover images are embedded to exact pixel dimensions: Amazon KDP 1,600 × 2,560px, Apple Books 1,400 × 1,873px, Kobo 1,600 × 2,400px (width × height).
Stage 6 — Final Human Review
Before any file is delivered, a second specialist reviews the completed eBook against your original manuscript. This is the stage that catches the errors automated validation cannot catch.
Machines catch structural errors. People catch the ones that require reading.
A chapter heading at the wrong level. A footnote that links to the wrong anchor. An image that is technically present but slightly misaligned. These are the errors that automated validation passes over and human review catches. Stage 6 is where they are found — and corrected — before delivery.
Original Styles Retained Throughout
Your formatting decisions are preserved through every stage of this process, within the constraints of each format’s published specification.
Chapter headers, drop caps, first-line indentation, blockquotes, callout boxes, special paragraph styles, custom section breaks — all are preserved exactly as intended in your source file. We do not impose a default template on your work. We convert your manuscript into a format your readers can read, while keeping it recognisably yours.
Formatting elements preserved:
- Chapter headers and sub-headings
- Drop caps and ornamental elements
- First-line indentation and spacing
- Blockquotes and callout boxes
- Special paragraph and section styles
- Custom fonts (with valid licence)
- Tables, lists, and nested elements
Where a specific formatting choice cannot be replicated in the target format, we flag it before delivery and discuss the closest available option with you.
Lifetime Warranty — The Quality Guarantee That Does Not Expire
At the end of this process, you receive a converted file that has been through six stages of human review and professional validation. If an error is found after delivery — today, next year, or five years from now — we fix it at no charge.
No
Expiry date
No
Claim process
No
Exceptions
This warranty exists because we are confident in the process above. If we were not, we could not offer it.
Quality Standards — Frequently Asked Questions
The current release of EPUBCheck, the W3C’s official ePub validator. We update our validation process whenever a new version is released that changes retailer acceptance criteria.
Yes. Every manuscript submitted as a PDF goes through the PDF comparison stage. This applies to standard PDFs, print-ready PDFs, and scanned books after OCR processing.
No. Rush delivery reduces the time between stages, not the stages themselves. Every order — 72-hour standard or 24-hour rush — goes through the same six-stage process.
Send us the rejection notice. We identify the specific error in the code and fix it. Platform rejections caused by our conversion are corrected at no charge — included in every order.
Every page of a PDF manuscript is reviewed by a human during the OCR stage — not just processed by automated software. This allows us to catch errors, inconsistencies, and source problems that automated OCR misses, and to flag them before conversion begins.
See Our Quality Process in Action
Every conversion goes through these six stages. If you would like to see a sample before ordering, contact us — we provide sample conversions for projects over $65.
