A request for production of social media records usually arrives broader than anyone can answer and narrower than the client hopes. “All posts, messages, photos and comments on any social media account from the date of separation to the present” covers years of content across several platforms, most of it irrelevant, some of it sensitive and a little of it important. Responding to it well means settling the scope, collecting from the platform rather than the screen, and producing something the other side can check.

This guide covers the responding side: reading the request, collecting the account archive, reviewing and redacting, producing and serving the written response. Rule references are to the Federal Rules of Civil Procedure; state rules usually track the same concepts with different numbering and deadlines. For the requesting and collection side, see our guide to social media evidence collection.

What does a social media request usually ask for?

Break the request into its parts before you object to any of them. Most requests combine some or all of these:

  • Accounts. Named platforms, “any account,” or accounts under other names the party controls
  • Content types. Posts, stories, comments, reactions, photos and videos, direct messages, check-ins, friend or follower lists
  • Dates. A range tied to an event, or “to the present,” which raises supplementation questions
  • Subjects. References to the other party, the children, a relationship, travel, alcohol, employment or income
  • Form. Native archives, screenshots, PDF, or no form at all
  • Access. Occasionally, passwords or a demand to log in for opposing counsel, which almost always draws an objection

List every account the client has used during the relevant period, including dormant ones, business pages, secondary profiles and accounts on platforms the request did not name but its definitions cover. A missed account found later means a supplemental production and an awkward conversation.

How do you scope and object without stonewalling?

Answer the defensible core and object to the rest with specific grounds. Rule 34(b)(2)(B) requires objections to be stated with specificity, and Rule 34(b)(2)(C) requires a statement of whether responsive material is being withheld on the basis of an objection.

Request languageCommon objectionNarrowing that often resolves it
“All content on all accounts”Overbroad, not tied to a claim or defense; disproportionateContent referencing named people or subjects, within a date range
“To the present”Ongoing burden; unclear cutoffA cutoff date plus an agreed supplementation schedule
Friend or follower listsThird-party privacy; low relevanceIdentify specific people relevant to the claims
Passwords or login accessOverbroad; exposes privileged and unrelated content; securityProduction from the account download instead
Deleted contentNot reasonably accessible, if trueDescribe what the archive includes and what it does not

Proportionality arguments land better when they come with a counterproposal. Our guide to proportionality in ESI discovery covers the factors, and the meet-and-confer is the place to agree on search terms, named people and date ranges before anyone spends a weekend reviewing.

How do you collect social media records from a client’s account?

Use the platform’s own account download, run it from the client’s account with the client present or by the client following written instructions, and log it. Screenshots are a supplement for display, not the collection.

PlatformDownload route (names change; check current settings)Check before you rely on it
FacebookAccounts Center, “Download your information”Full date range, messages included, media quality set to high
MessengerAccount download plus the separate export for end-to-end encrypted chats, where offeredEncrypted chats may be missing from the standard archive
InstagramAccounts Center, “Download your information”Messages and comments included; JSON or HTML chosen deliberately
XSettings, “Download an archive of your data”Direct messages and media present in the archive
TikTokSettings, “Download your data”Messages and comment history categories included
SnapchatAccount settings, “My Data”Only saved chats and limited history survive; say so
YouTube and GoogleGoogle TakeoutComments and uploads selected; full range

For each archive, record the account name and profile address, the date requested and downloaded, the categories, date range, format and media quality chosen, and who ran it. Hash the downloaded file on arrival and keep it untouched; review copies come from it. Download links often expire within days, so tell the client to send the file the same day. Our article on hash values in litigation covers the fingerprinting step.

Two choices matter more than they look. HTML archives are easy for humans to read; JSON archives keep cleaner timestamps and identifiers for processing. Many teams request both. And low media quality can shrink photos and videos to the point where they are hard to review, so set it to high even when the file gets large.

What should the review cover?

Review conversations and posts in context, not as isolated lines, and document every decision to withhold or redact.

  1. Filter to scope. Apply the agreed date ranges, named participants and subjects, and write down every filter. Silent filtering is how cherry-picking accusations start.
  2. Read whole threads. A direct-message conversation is a thread like any text exchange. Produce reasonable context around responsive messages rather than single lines.
  3. Check privilege. Clients sometimes message friends about what their lawyer said, or forward counsel’s advice into a group chat. Those messages raise waiver questions for the supervising attorney, not a quiet redaction. See the privilege log guide.
  4. Protect third parties and children. Unrelated people’s names, photos of children, addresses and health details may call for redaction or a confidentiality designation under a protective order; see attorneys’ eyes only designations.
  5. Flag what is missing. Deleted posts, expired stories and disappearing messages will not be in the archive. Note them for the written response instead of letting the other side discover the gap.

Direct messages in a Facebook or Instagram download are the part of a social media production that most resembles a text message production. Message Extractor reads Facebook and Instagram message exports alongside phone exports, merges them into one thread model with date and participant filters, and renders a paginated, Bates-stamped log with an index and a source manifest.

How should a social media production be formatted?

Produce a rendering people can read and cite, plus the native archive for the same content, unless the parties agreed otherwise.

  • Rendering. A chronological PDF of responsive posts and conversations, showing account name, date, time with a stated time zone, the content, and media inline or in an appendix. Every page carries a Bates number.
  • Native files. The archive files for the responsive content, produced as numbered items with a slip sheet where a page stamp cannot be applied. Our comparison of PDF, native and load file production formats covers the trade-offs.
  • Manifest. A list of every archive collected, with its date, account and hash, so any page can be traced back to its source.
  • Redaction log. Each redaction labeled on the page with a reason and listed in the log.

Sample written response

Response to Request No. [number]: [Party] objects to this Request as overbroad and not proportional to the needs of the case to the extent it seeks all content from all social media accounts without limitation as to subject. Subject to that objection, [Party] will produce posts, comments and direct messages from the [platform] account [account name] between [date] and [date] that refer to [named people or subjects], as a Bates-numbered PDF with times shown in [time zone], together with the native account archive files for that content. The archive was downloaded from the account on [date]. Content outside those subjects is being withheld on the basis of this objection. The archive does not include [stories older than the platform retains / content deleted before (date)]; no other responsive content is being withheld.

Serve the production with a production cover letter listing the Bates range, the accounts and the manifest, and keep a production log. A request that runs “to the present” makes supplementation part of the job; calendar it.

What goes wrong in social media productions?

MistakeConsequenceFix
Screenshots instead of the account downloadMissing dates, identifiers and context; deficiency letterCollect the archive; use screenshots only for display
Low media quality settingUnreadable photos and video; re-collectionDownload at high quality
Only the client’s side of a conversationCompleteness dispute and credibility questionsProduce the whole responsive thread
Encrypted or disappearing chats silently absentAccusations of withholdingSay what the archive does not contain
Client deletes posts after the requestSpoliation exposurePreservation instructions and an early full download

Lost content needs the same handling as lost text messages; our article on spoliation of text messages covers how courts approach it, and a litigation hold letter to the client should name social media accounts expressly.

Objections, privilege and waiver questions, and any advice to a client about privacy settings or deleting content are decisions for the supervising attorney. Family, housing and state trial courts often have their own expectations for social media discovery, so confirm local practice with counsel licensed where the matter is pending.