Credits without surprises: read the badge before you run

8 min readBeginner

What this costs

Free to read. The point of the lesson is that nothing surprises you afterwards.

Before you start

Nothing.

This is the most important lesson in the Akademy and the one the old version never had. Kuration charges in four different ways, every tool tells you which one it uses, and almost every "why did that cost so much" support ticket is someone who never looked.

Read the badge on the tool card

Everything in this lesson comes down to one habit. Open the Toolbox, find the tool you are about to use, and read the badge on its card. Every tool states its own price and its own charging behaviour there, and that badge is always current in a way no lesson or price list can be.

The Kuration Toolbox with four different cost badges visible: Per result, If result, FREE, and a flat credit figure — All four badges, on one screen. The number is only half the answer. The badge beside it decides what a miss costs you.

The four badges, and what each one means for a miss

Badge on the card	You are charged	A miss costs you	Run it
FREE	Never	Nothing	As often as you like. Scoring lives here
A credit figure on its own	For every row the tool processes	The same as a hit	After filtering, never before
A figure plus If result	Only on rows where it returned a value	Nothing	Wide, on a filtered list
A figure plus Per result	Once per item returned, so one row can charge several times	Nothing	Wide, but cap what it returns per row

The tools whose behaviour you should know by heart

Tool	What to expect on the card	Why it matters
Scoring Agent	FREE	It reasons over data already in the row rather than researching. Score early, score often, and use the score to gate the paid columns
Official Website Finder	A flat per row figure	The dependency almost every other tool has. Run it on the rows that are missing a website, not on all of them
Find People	A figure plus Per result	One company can return several people. Budget on results
Waterfall Logic	FREE	Picks the first non empty value from several inputs, so you can chain cheap attempts before an expensive one
Verification and QA tools	A flat per row figure	A fail costs the same as a pass, because deciding it was a fail is the work
Custom Research Tool	A per row figure	Preview on five rows before you create the column across the table

Predicting a run before you press it

Work it out per column, not per project. For each column you are about to create: how many rows will it actually run on after filters, what does the card say one unit costs, and does a miss cost anything.

Here is the same 1,000 sourced companies built two ways. The unit prices below are made up for the arithmetic. Substitute whatever your own tool cards say, because the shape of the result is the point rather than the totals.

Worked example · The same 1,000 companies, flat versus curated first

Step	Without curating	Curating first
Rows after sourcing	1,000	1,000
Filter on what the source gave you, free	skipped	380 remain
Website Finder, say 2 credits a row	2,000	760
Score against your ICP, free	0	0
Cut to rows scoring 7 or above, free	skipped	145 remain
One decision maker each, say 3 credits a result	about 3,000	about 435
Total	about 5,000	about 1,195

Same source, same tools, same outcome for the rows you were ever going to email. The difference is entirely the order, and two filters that cost nothing.

Where to check what you actually spent

The Kuration project analytics panel showing total credits consumed and average enrichment for a project — Every project reports its own lifetime credit usage, alongside how complete the enrichment actually is. Those two numbers together are the only honest measure of whether a list was worth it.

Your remaining credits sit in the top navigation of the dashboard. For a breakdown by action, open Settings and then Usage. If a run surprised you, that breakdown tells you which column did it, which is usually enough to spot the missing condition.

Ran an expensive tool before filtering

What you see: The balance drops by thousands on a list you then cut to a tenth
The fix: Filter, then select the visible rows, then run. When filters are active only visible rows are processed.

Deleted a column expecting a refund

What you see: Column gone, credits still spent
The fix: Deleting a column does not refund credits, the work was already done. Preview on a handful of rows before creating a column on all of them.

Read the unit price but not the behaviour badge

What you see: A per result tool costs far more than rows times unit price
The fix: One row can return several results. Read both badges on the card, not just the number.

Re ran a whole column to fix a few empty cells

What you see: Paying twice for rows that already worked
The fix: Successfully enriched rows are not re processed on a normal re run. Course 2.3 covers the advanced run options that control this.

Check yourself

0 of 3 answered

1.A tool card shows a credit figure and the words "If result". What does a miss cost you?

2.A tool card shows a credit figure with no behaviour badge beside it. What does that mean?

3.You delete a column you regret creating. What happens to the credits?

Try it yourself

Take a project with at least 50 rows. Before running anything, write down what it would cost to add websites to every row, and what it would cost to add websites only to rows missing one. Then check the difference in the Usage breakdown after you run the second version.

You have done it when: Your predicted cost for the filtered run is within 10 percent of what was actually deducted.
Credit budget: Small, if you filter first. That is the whole exercise.

Related and next

Recipes that use this skill

Make a messy list outbound ready