Flat monthly price · No percentage of spend See pricing
Twine
Menu
Product · measure

Know what worked. Prove it to anyone.

Twine connects results to pipeline instead of form fills, runs experiments with decision rules, learns only from measured outcomes, and freezes the record into reporting leadership can interrogate.

01
Outcomes

Optimize for pipeline, not the easiest form fill.

Tell Twine which outcomes matter. When CRM data is available, it connects campaign decisions to qualified leads and opportunities instead of treating every conversion as equal.

  • Thin data stays inconclusive. Twine does not invent certainty.
  • Your sales cycle matters. Results are judged after they mature.
  • Negative outcomes count. A wrong call is recorded, reversed, and remembered.
Signal tiers
TIER 1 CRM deal stages HubSpot / Salesforce · matched to campaigns
TIER 2 Named conversions weighted by the value you rank
TIER 3 Engagement quality GA4 engaged sessions · directional only
Measured Jul 9 bid change: inconclusive
№ 02 · Experiments

Test a decision, not a hunch.

Every experiment starts with a hypothesis, a primary metric, minimum evidence, and a decision rule, recorded before anything runs. Starting an experiment does not touch the ad platform, and a winner becomes a draft proposal for review. It is not promoted automatically.

Before · EXP-031
Hypothesis
Benefit-led headline outperforms feature-led on Enterprise Search.
control + 1 variant · primary: qualified CPL
minimum evidence: 20 conversions · window: 21 days
guardrail: pause if CPL > $220 · planned spend: $2,400
After · EXP-031
Experiment result
Variant B produced a 17% lower platform CPL. Qualified-opportunity evidence remains directional.
Operator decision
Stage Variant B as a draft proposal for Enterprise Search. Do not apply it to other campaigns.
03
Account memory

Memory built from measured outcomes.

Twine builds account-specific memory from measured decisions, outcomes, experiments, and operator feedback. Directional signals stay visible without becoming beliefs. Only reliable measured outcomes strengthen the patterns Twine is allowed to reuse.

Targeting choices live in a reusable ledger of their own: in market now, saved for reuse, and worth trying.

Account beliefs
Strengthened
Search-term hygiene pays for itself here
supported by 6 measured outcomes over 90 days · confidence 0.9
Weakened
Bid increases on competitor terms
negative outcome Jul 16 · future instances require human review
Visible, not believed
Tuesday sends outperform
directional signal · below evidence floor · not reusable yet
04
Reporting

Report from the operating record, not reconstructed memory.

Freeze a snapshot into a quarterly review that keeps observed facts, interpretation, operator decisions, and unapproved proposals visibly separate. The artifact does not silently update when the dashboard changes. Later corrections append, they never rewrite.

Share it in Slack, email, CSV, or PDF. Every output stays tied to the same frozen snapshot, so there is no deck version that quietly diverges from the record.

Quarterly business review Frozen · SNP-0729
Q3 2026 · frozen Jul 29 · template v4
Observed facts

Paid media produced 9 qualified opportunities against a target of 12. Three more remain inside the sales-cycle lag.

Twine interpretation

Search efficiency improved after term exclusions. The evidence does not yet support a cross-platform reallocation.

Operator decisions

Hold the current allocation and review again August 14. Owner: sarah@

Future proposals · not approved

Move $3,000 from reserve to Enterprise Search if two open opportunities advance.

SlackEmailCSVPDF

Know what worked, with the evidence to prove it.