Overview
Eight groups, across two categories, and the split is not cosmetic. Channels are grouped by where the record lives, because that decides how it is obtained and how long there is to obtain it. Claims are grouped by what has to be proven, because that decides whether the record is worth getting. Those are two different questions and an attorney arriving here usually has an answer to one of them and not the other.
The pairing is the point. A dispute is always some claim arising in some channel, and the combination determines the work. Invalid traffic in paid search is a different engagement from invalid traffic in affiliate marketing: different records, different holders, different windows, different defenses. Scope disputes about analytics implementation and scope disputes about creative production share a legal shape and share nothing else. So the site does not force a single hierarchy. It gives both axes and lets a reader enter from whichever one they already know.
The four channel groups
Paid media. Start here when money moved through an auction platform and the argument is about what it bought. Search advertising, display and programmatic, paid social, and online advertising generally. The distinguishing feature of this group is that the authoritative record belongs to the platform rather than to either party, which is good news for obtainability and bad news for timing.
Organic and owned. Start here when the channel is one a party ran rather than bought — social accounts, email programs, published content, affiliate relationships. The record is mostly party-held, so the analysis has to address completeness rather than assume it, and the most useful copy of a record is often the one a vendor holds rather than the one the party produced.
Commerce and retail. Start here when the transaction itself is in issue: marketplace selling, retail media and merchandising, or the conversion behavior of a website. The structural surprise in this group is how often a seller does not hold its own transaction and advertising record, and how quickly marketplace reporting windows close.
Measurement. Start here when the dispute is about the numbers rather than about the activity — analytics implementation and tracking, or attribution and return on investment. This is the group where the honest answer is sometimes that no primary record of the thing in question exists, and where the configuration history is usually more probative than any figure.
The four claim groups
Performance and contract. Start here when the argument is that the work was not what was promised: agency and consultant disputes, scope and deliverable disputes, performance-based compensation. The element in dispute is generally the standard of care, and the load-bearing documents are the contract and the statement of work rather than anything about industry practice — for reasons this group’s pages set out at length and without flinching.
Fraud and abuse. Start here when the traffic, clicks or conduct were not what they appeared: click fraud and invalid traffic, affiliate and incentive fraud, bot and impression fraud. The element is usually causation, and the recurring limit is that records establish patterns rather than identities.
Claims and conduct. Start here when the dispute is about what was published: false and deceptive advertising, and trademark use in advertising. The evidentiary problem is preservation, because the creative, the landing page, the listing and the disclosure are all transient and most of what will ever exist has to be captured early.
Loss and quantification. Start here when liability is largely conceded and the fight is over the number: traffic and revenue loss, and marketing damages models. The element is quantum, the leaps between impressions, sessions, orders and dollars are where models come apart, and part of the work properly belongs to an economist rather than to a marketing expert.
Why the two scales are different
Each page opens with one judgment, and the scale changes with the category because the reader’s question changes. On a channel page the question is can this be obtained, so the scale runs platform-held, party-held, modeled. On a claim page the question is what must be shown, so the scale runs standard of care, causation, quantum. Putting both on one scale would have produced a tidier design and a less useful one.
The value of that arrangement is that it makes a mismatch visible. A quantum dispute resting entirely on a channel whose figures are modeled is a matter with a problem in it, and the problem is worth knowing about at retention rather than at deposition. A standard-of-care claim in a channel where the platform holds a complete configuration history is a much stronger position than the same claim in a channel where the only record is a party’s own spreadsheet.
Using the grouping to decide what to ask for
The practical sequence is short. Identify the channel, and read what records that channel produces, who holds each, and how long each survives. Identify the claim, and read which element is contested and what that element requires. Then compare: the records the claim needs against the records the channel actually produces, on the dates in question. What falls out of that comparison is a discovery list, and often it also produces the more valuable finding — that a particular record no longer exists, which changes the theory rather than merely the request.
What this scheme cannot do is tell anyone which group their matter belongs to when the pleadings are ambiguous, and they frequently are. A claim pleaded as breach of contract may turn out to be a causation dispute in substance; a claim pleaded as fraud may resolve into an arithmetic reconciliation. The grouping is a way into the material, not a classification anyone should defend. When the two entries you land on disagree about what the case is, the disagreement itself is usually the most useful thing on the page.
Published