Likely: probabilities in cells

Likely adds three formulas to Excel that answer a question about every row of text and give you back a number you can sort, filter and threshold.

Type =LIKELY.IS(A2:A5000, "Is this customer asking for a refund?") and every cell comes back with a probability between 0 and 1. Rows near 1 are near-certain, rows near 0 are safe to skip, and the ones in the middle are the handful worth reading yourself. Sorting by that column turns five thousand support tickets, survey answers, reviews or leads into a short list.

Two more formulas cover the rest of the work. =LIKELY.CHOICE(range, question, labels) picks the best-fitting label for each row from a list you supply, so a column of messages becomes a column of teams, topics or stages. =LIKELY.SCORE(range, question, levels) scores each row on a scale you define, such as calm to furious, and returns a number rather than a word.

The task pane adds the things a spreadsheet person actually needs. Bulk run judges a whole column at once and writes plain values next to your data, so nothing recalculates while you work. Highlight uncertain shades the rows between 0.35 and 0.65 - the ones where a human should look. Freeze to values turns a finished column into static numbers you can share.

It is fast enough to use on real data: five thousand rows come back in about a second, and repeated cells cost nothing, so re-running a column you have already judged is free.

Typical uses: routing support tickets to the right team, finding refund and cancellation requests in an inbox export, scoring open survey answers, filtering leads by intent, flagging negative reviews, tagging research notes. Anyone who keeps a column of free text in a spreadsheet has a use for it.

Likely is free for 2,000 rows a month. Paid plans start at $12 a month for 250,000 rows. You need a free key from getlikely.app, which the task pane can create for you in one step.