What the baseline engine is
The baseline engine grades a style at one account, Winner, Sleeper or Bleeder, from what that account's own doors sold before. It is one of Tightly's four named engines, and it comes with Tightly Essentials+ and Tightly Pro, because it grades an account and Lite and Essentials hold no accounts. Its faces call it Grades. The grade's words, and what measured, borrowed and From the brand mean under it, are in The grade: measured, borrowed, brand.
The baseline engine is not the forecast a never-sold product borrows from comparable past styles. That borrowed forecast belongs to the forecasting engine and is described in Forecasts for products that have never sold; the baseline engine reads its result as its last rung.
What the baseline engine reads, rung by rung
For each account and style it looks for evidence in a fixed order and stops at the first rung with a style the account has reported for at least eight weeks. A style under eight weeks does not vote.
- The account's own sell-out of the style's analog in place: the past product the style launches on, set on Analogs. One predecessor whose shelf was observed every week gives a measured grade; several blended, or weeks with a corrected shelf, read borrowed.
- Where the style is a kit, the account's sell-out of the product inside it. Borrowed.
- The account's sell-out of the style's proposed look-alikes, the comparable past styles the analog proposals rank. Borrowed.
- The account's sell-out of the other styles in the style's family, set on Families. Also borrowed. A style with no family has no family rung; the engine never guesses a family.
- The brand's own level for the style, the borrowed forecast made from your own sales. The row reads From the brand, with no grade word.
Where none of the five answers, the row reads Sell-in only, and the account's demand stands on what it orders.
When the baseline engine runs
It runs inside every sync, immediately after the forecast curve is written, so the brand's level it reads is this sync's and not yesterday's. Syncs run every hour in production. It grades the styles inside active Commitments whose window has not closed. Where no active Commitment covers any style at all, as on Essentials+, which holds no Commitments, it grades the styles in the order book's seasons that are still open instead. It runs again for one account alone when that account's sell-out file lands and writes rows. Nothing is pressed to refresh a grade. A tenant whose plan does not include the engine is skipped, and nothing is written for it. A grading failure never fails the sync: the grades keep their last date and the engine's state reads Failed.
What a planner changes in a grade, and who may
A planner changes a grade by changing its evidence, never the grade itself.
- On Analogs, Use as the analog puts a past product in place, which is rung 1; Break this link takes it away; Not this one takes a proposal out of the look-alikes for good.
- On Families, placing a style in a family gives it the family rung.
- On the account, land its weekly sell-out through Sell-out intake, and its rows are graded from its own doors.
A viewer reads every one of these pages. A member or an admin makes the changes, because every write on Analogs and Families asks for a seat that can edit.
The baseline engine's state sentences
Grades are Late four hours after a sync they should have followed and Stale after twenty-six. The Inbox lists Grades, with its sentence, while the engine is late, stale or failed. The engine's one sentence, state by state, character for character:
Before the first run, with no state word:
Grades have not been made for this account yet.
Late or Stale:
Grades are from {as_of}; the refresh since then did not run.
Partial:
Some grades did not refresh today; each row carries its own date.
Failed:
Today's grading did not finish; showing {as_of}'s grades.
Nothing yet:
No account reports sell-out yet, so there is nothing to grade.
Not on your plan:
Grading is off for this tenant.
Paused:
Paused with the connection; grading waits for the sync.
Not on the run:
This tenant is not on the grading run.
Only Paused and the state before the first run carry a door, and it is Integrations: the engine follows the sync, so the sync is what a person repairs.
What each plan sees of the baseline engine
| Plan | What it sees |
|---|---|
| Tightly Pro | Grades on accounts and on the showroom's Grade column |
| Tightly Essentials+ | Grades on accounts: the Grade column on the account's Book and Sell-out tab, and the source word under demand on the Book |
| Tightly Essentials, Tightly Lite | No grades; these plans hold no accounts |
A read the plan does not include is refused with one sentence:
This organisation's plan does not include the baseline engine. It is sold with Essentials+.
In the retailer chair the same engine grades a vendor's style from your own store history: a second key, not a second engine.
Glossary line
The baseline engine grades a style at one account, Winner, Sleeper or Bleeder, from that account's own sell-out where it exists, and from the brand's own level where it does not.