What ExportDecode can read, what it calculates and what it cannot do — one table per provider.
Formats are documented as of mid-2025. Services occasionally change their export structure. If your file fails, report it with the column headings and the error message.
Retail.OrderHistory.1.csv (or any CSV whose name contains "order" and not "refund", "return" or "cancel")
Required columns
Order ID, Order Date, Product Name (or ASIN), and either Unit Price or Total Owed
Analysis available
Total spending, order count, average order value, biggest single item, spending by year or month, spending by weekday, top items by total spend, most-reordered items, price swings on repeat purchases, biggest single orders
CSV exports
All line items; monthly totals
Known limitations
Refunds and returns are in separate Amazon files and are not subtracted — totals show gross spending. Digital purchases (Kindle, Prime Video, apps) are in separate "digital order" exports not currently supported. Multi-currency orders: only the dominant currency is included.
ViewingActivity.csv (full export) or the CSV from the "Download all" button on the viewing activity page
Required columns
Full export: Start Time, Duration, Title. Legacy "Download all": Date, Title
Analysis available
Total watch time, session count, shows and movies, active days, most-watched titles, binge days, devices, countries streamed from, weekday × hour heatmap. Legacy format: title counts and trends only (no watch time).
CSV exports
All sessions; show ranking
Known limitations
Netflix records time in UTC; times are converted to your device's time zone. "Duration" is how long playback was running — falling asleep with autoplay counts. Trailers and previews are automatically excluded. Legacy "Download all" file has no durations.
Total listening time, play count, artist and track counts, top artists by listening time, most-played tracks, skip rate (extended only), platform breakdown (extended only), podcast shows and episodes (extended only), weekday × hour heatmap
CSV exports
Artist ranking; track ranking
Known limitations
Basic account data covers roughly the last 12 months and has no platform or skip information. Extended history is accurate but can take up to 30 days to arrive from Spotify. Totals will differ from Spotify Wrapped, which applies its own date range and listening rules.
watch-history.json (inside the Takeout/YouTube and YouTube Music/history/ folder)
Required fields
time, title, titleUrl (must be a youtube.com link)
Analysis available
Videos watched, distinct videos, estimated watch time, channel count, Shorts share, busiest day, top channels, rewatched videos, YouTube vs YouTube Music split, weekday × hour heatmap
CSV exports
Channel ranking; rewatched videos
Known limitations
Google's export includes timestamps but no video lengths. Watch time is an estimate based on an average duration you choose; Shorts are counted as 30 seconds. Videos watched while history was paused or auto-deleted are absent. The HTML format exported by default in Takeout is not supported — set format to JSON before exporting.
Yes — drop the export ZIP and ExportDecode finds both files inside
Expected file names
followers_1.json AND following.json — both are required
Required fields
JSON objects with string_list_data arrays containing href (instagram.com profile URL) and optionally timestamp and value
Analysis available
Accounts you follow that do not follow back; followers you do not follow back; mutual follows; counts and percentages
CSV exports
Non-followers list; fans list
Known limitations
The export is a snapshot from the day you requested it — accounts followed or unfollowed since then are not reflected. Renamed, deactivated or private accounts may appear incorrectly. ExportDecode cannot log in to Instagram or take any action on your account.
Any CSV with a date column, a description/merchant column and either an amount column or separate debit and credit columns. Delimiters: comma, semicolon or tab. Decimal separators: period or comma.
Date formats supported
ISO (YYYY-MM-DD), US (MM/DD/YYYY), European (DD/MM/YYYY), and common bank variants. Day-first vs. month-first is auto-detected.
Analysis available
Active recurring charges with rhythm (weekly to yearly), monthly and annual cost estimate, next expected charge date, stopped charges, price creep on fixed-amount subscriptions
CSV exports
All detected recurring charges
Known limitations
Detection is rhythm-based. Subscriptions billed under changing merchant names, or with fewer than 3 charges in your file, may be missed. Regular habits (weekly coffee, daily lunch) can look like subscriptions. Export 12–24 months of history for best results. Figures are from your CSV and may differ from official statements.
What is not currently supported
The following are commonly requested but not yet supported. They are listed honestly — no analyzer exists for them:
Amazon digital orders (Kindle, Prime Video, app purchases) — separate files from Amazon, not yet supported
Facebook / Meta data exports
Twitter / X data exports
Apple health or Screen Time exports
ChatGPT conversation exports
LinkedIn exports
Google Pay or Google Shopping history
Uber / Lyft trip data
If you need support for a format, tell us — include the file name and column headings (no personal data).
