Which differences should we keep when we organize work?

You asked whether customer stories and user stories differ. The older suggestion divided them by whether the speaker paid. That was a proposal, not your definition.

The Singular Bank and Boston Children’s Hospital examples are stored as case studies. The NVIDIA speaker’s account is stored as a testimonial. These neighbors show why the definitions must be considered together.

I recommend Customer Story as the shared browsing name, with User Story reaching it too. Preserve the subject, company and relationship to the product separately, including when the relationship is unknown.

This would cover an account of use, not every endorsement or review. A short endorsement alone can remain a testimonial. The boundaries with Case Study and Product Review still need examples in the definitions review.

Your decision that a vlog can remain a useful kind is carried forward. A vlog may also tell a customer’s story when both descriptions apply to the whole piece.

The alternative is distinct Customer Story and User Story choices. That needs a reliable difference visible in the collected evidence, rather than an assumed purchase history.

No current piece is renamed by this answer. Prompt examples, search wording and neighboring definitions must be aligned before implementation.

Your retail brief named Best Buy endcaps, product pages and in-store displays. The old question treated aisle display as a possible synonym for endcap. That equivalence is not established by the available example.

I recommend retaining the distinction in discovery. People could enter through in-store displays, then narrow by the evidenced kind or placement without duplicating the work.

Whether the final distinction belongs in a display type or a placement choice depends on the broader retail review. This answer preserves the meaning; it does not force a particular tree.

The image titled That wheel merch tho — image three illustrates the current evidence gap. It cannot demonstrate the difference between end-of-aisle and other display placements.

A photograph keeps its identity and links to the display it shows when that display is known. The placement belongs to the displayed work’s occurrence, rather than becoming the photo’s identity.

The alternative is one broad In-store Display choice without a narrower distinction. That is simpler to maintain, but a researcher could not reliably filter for endcaps.

Neither proposed type appears in the checked field. That does not prove the library contains no photographed displays.

The old recommendation folded Corporate Gift into Merchandise and said the fold was free. That would be unsafe if useful gifting information had nowhere to go.

I recommend describing the object separately from its gifting purpose. The hoodie and keyring keep their own descriptions; a confirmed gifting use becomes an additional searchable fact.

An organized gift set may also be useful as a whole. It should link to its independently useful parts, rather than force the whole set and each object into one identity.

Unknown gifting status remains unknown. Neither the recipient’s identity nor the presence of a logo proves that an object was a gift.

The alternative is a separate Corporate Gift type alongside Merchandise. It makes gifting easy to browse, but the definitions must allow a gifted object to retain its physical kind.

This answer does not approve an automatic rename or invent a storage field. The retail definitions and relationship review must show how the gifting fact is preserved before any fold is implemented.

The earlier page recommended merging the questions because some words overlapped. I disagree with that reasoning. Similar words do not make the questions identical.

The live records include 767 analyzer descriptions pairing an interface on screen with the product as the focus. Another 99 pair an interface on screen with active use. Choosing only one of those existing answers would lose one of those facts.

The local prompt already asks these separately. It specifies a short visibility list and another list for product role. The active vocabulary agrees with that visibility list, yet 238 stored records use other visibility words.

That mismatch needs investigation through the actual prompt, writer and older analysis runs. It does not establish that the product-role question should absorb the visibility question.

The Sora release demo, the Sora generated-video post and the Codex sales demo remain useful source examples. Watching the exact intervals is still required to judge their recorded descriptions.

I recommend keeping the meanings separate and reviewing the choices. A physical product resting on a table, multiple products and an off-screen reference need clearer examples than the current list provides.

The alternative is a combined description that preserves both facts together. That can support search, but it must not reduce every moment to one value that erases a valid combination.

No automatic conversion or guarantee of avoiding reanalysis follows from this answer. Unknown values and human corrections must survive any later repair. Whole-piece types, campaign membership and credits keep their distinct meanings.

You suggested ongoing product campaign and questioned whether another container name was needed. I recommend allowing that possibility within Campaign, provided evidence supports a coordinated body of work.

The Ramp story and its captured page image are stored pieces. Humans of HOKA supplies another source example. Their stored memberships do not prove that each campaign record is correctly named or bounded.

Keep duration separate from purpose. A product launch and an ongoing product campaign can both promote a product; changing duration should not erase what they are trying to achieve.

The earlier proposal added always-on as another campaign kind. I would not make that the only way to represent duration, and I would not infer it merely from a blank end date.

A saved search for all customer stories can gather work across campaigns without asserting that the results form a real campaign. Your boards, trackers and reels can organize a useful selection independently.

The alternative is a separate program or series kind. That becomes useful if we find a real body of work whose relationships cannot be represented honestly by the approved grouping kinds.

This answer will guide grouping detection, time filters and campaign pages. It does not certify the existing campaign records or mean their current storage already supports every approved relationship.