Pre-sales audit
An agency has a prospect call and needs a website roast, SEO gap scan, or competitor teardown before spending senior strategy time.
Only buy the first report if it can save or create billable time this week.
This card turns the likely price/value objection into a yes/no agency test. Task Agents should not be sold as interesting AI. It should be tested when a small research packet helps an agency win a call, prep an audit, rescue messaging, or brief creative without burning senior time.
The first order needs a payback path, not curiosity.
Client retention
An account lead needs review mining, FAQ copy, or proof-backed content angles to make the next client conversation less vague.
Creative brief
A freelancer needs shootable hooks or UGC angles today, not another empty content-calendar brainstorm.
Run this before forwarding the page.
+2The buyer can name the prospect, client, or campaign where the report will be used.No named job means no revenue signal.
+2The report could save at least 20 minutes of skilled research, call prep, or creative briefing.At agency rates, an $11–$19 packet only needs a small time save to be rational.
+1The output can be judged from the proof pack or sample in under five minutes.Slow evaluation usually means the task is not narrow enough.
+1The buyer would know exactly where to paste or use the result: sales notes, client email, brief, landing-page fix list, or ad-angle doc.A home for the output increases conversion.
-2The buyer asks for a dashboard, automation platform, or ongoing SaaS workflow.That is a different product. Task Agents sells finished packets.
-2The use case is “might be useful someday.”Do not count politeness as demand.
Threshold: 4+ points is worth a manual 1:1 forward or starter order. Under 4, ask for a named client job first.
Qualified revenue conversion.
Count only a paid starter order, a named referral with a live job, a proof/sample request, or the exact price/value objection. The card is designed to keep the next warm test from producing soft compliments that do not move revenue.
Do not sell “AI reports.” Sell one saved workflow.
If the buyer cannot point to the call, client, pitch, script, page, or brief where the packet will be used, the referral is not ready. Route them back to the scorecard instead of pushing checkout.
Personal 1:1 copy Dimitri can send, not automated.
I’m not trying to sell generic AI work here — I’m testing whether an $11–$19 research packet can save billable agency time. If you had one prospect/client deliverable this week, would this ROI card make a starter website roast rational, or is price/value still unclear? https://tasks.irislabs.dev/partners/agency-roi-card/