OriginalUniqueArt.com has 4 setup steps left.

The website pages analysis writes it: AkOOre reads your pages and summarises the company, its market, its segments, its products and its buyers. Re-running the analysis from the website page rewrites the two descriptions here, but leaves the three tables alone — see the next question.

An ordinary re-analysis fills the three tables only while they are empty. Once a list holds a row, no analysis ever adds, renames or deletes one — that rule is what protects the typical deal values, margins and repeat rates you type, and it protects the market demand measured against your segments. The price of it is that a taxonomy that has gone stale stays stale through every re-run.

Refresh the whole profile is the way out: it empties the three tables first, so the analysis that follows bootstraps all of them from scratch. It is destructive on purpose — the market demand measured against the current segments is deleted with them, and your typed figures do not come back.

Company market segments is that taxonomy written down: one row per segment, each one being one kind of buyer after one kind of thing — not a product line, and not a page of your site. It is the only list a report reads as a structure rather than as prose. The market demand analysis clusters every keyword it collects into these rows and then measures each one: searches a month, who ranks today, what share of it is yours, what the gap is worth. A segment you never declare is never measured, and deleting a segment deletes the demand measured against it. Keep them few and distinct — up to 20 rows, a general name of 2 to 5 words each. Off-market is reserved: the analysis uses that name for the searches it sets aside as not your market.

Typical deal value is what one won buyer in that segment spends on a first purchase, in US dollars — a plain number, no currency symbol and no thousands separator (2500, not $2,500). It prices the sale, while every money figure in the market demand report prices traffic, at what Google would charge you for it: the two are not the same money and must never be added together. The first analysis estimates the figure from your pages while the cell is empty; once you type one, nothing overwrites it.

Typical deal value, on the segments and on the products/services, is money: what ONE won buyer spends on a first purchase, in US dollars. Type it as a plain number — no symbol, no thousands separator, no other currency. If you sell in euros or pounds, convert before typing; nothing on this page does the conversion for you.

Margin and Repeat / year, on the personas, are not money at all: they are ratios against one full-price sale. Margin is the share of the price you keep (0.5 to 1), repeat is how many times that buyer comes back in a year (1 to 5), and their product is what one buyer of that persona is worth against a plain single sale.

One warning worth repeating: the deal value prices a sale, while the dollar figures in the market demand report price traffic, at what Google would charge you in ads for it. They are different money and adding them together means nothing.

Rows are matched by their name, ignoring case, spacing and order. So re-ordering the table, fixing capitals or collapsing a double space keeps the row — and with it the deal value or the margin you typed, and, on a segment, the demand already measured against it.

Changing the words is different: to AkOOre "Boutique hotels" and "Independent hotels" are two different rows, so a reword is a delete plus a create. The new row starts empty, and the old row's figures and measurements go with the old row. If you only want to change how a row reads, put the rewording in its description; if you really do mean a different segment, expect to re-run the market demand analysis and to type the figures again.

The first analysis proposes the rows from your pages; after that the table is yours — no later run adds, renames or deletes one. Same limits as the other two lists: up to 20 rows, a general name of 2 to 5 words a row.

Typical deal value is what one won buyer spends on that product on a first purchase, in US dollars — a plain number, no currency symbol and no thousands separator (2500, not $2,500). Enter the average first order, not the lifetime value: how often a buyer comes back is the personas' job, below. The analysis estimates the figure from the prices it read on your pages, but only while the cell is empty — a figure you type is never overwritten, and a cell you empty is open to being estimated again on the next run. Renaming a row replaces that product, and its deal value starts again from empty.

The two figures beside each persona say what a won buyer of that kind is worth relative to one full-price sale. They are ratios, not money — no currency applies to either, and neither is a deal value: the dollars live on the segments and the products above.

Margin is the share of the sale price left as margin on this buyer's purchases, usually 0.5 to 1 — use 1 for someone who pays list price, around 0.8 for a discount-driven shopper, around 0.6 for a reseller buying on volume pricing. Repeat / year is how many times a won buyer of that kind buys in a year, usually 1 to 5 — 1 is a one-shot buyer who never comes back, 3 buys about three times a year. Multiplied together they give that buyer's value multiplier: 0.8 × 3 = 2.4, a persona worth nearly two and a half full-price sales. The first analysis estimates both from your pages while the cells are empty; a figure you type is never overwritten.

All of them, but not in the same way. The market demand report is the only one that reads a list as a structure: it clusters your keywords into the declared segments and measures each one, which is why removing a segment removes its numbers.

Everything else reads these rows as context. The articles, the content briefs and the ad copy write about the products/services and to the personas; the competitor comparison scores your offer, your segments and your personas against each competitor's; the positioning gap and the keyword and ads introductions weigh what you found against what you say you sell and who you say you sell to; and AkBot answers from the three tables directly.

Which is the practical point of this page: a row that is wrong here is wrong in every report built on it, and it costs a minute to fix here against a re-run everywhere else.