ICE Prioritisation Coaching with AI: Is Your Team Actually Learning?
ICE — Impact, Confidence, Ease — keeps growth teams honest about which experiments to run first. Or it should. In practice, impact gets stated without a mechanism, confidence is uniformly high, ease ignores the design and legal work, and — the biggest failure — nobody ever compares what an experiment actually did to what the team predicted, so the scoring never gets better.
Ask Loopion Coach audits all of it from the meetings themselves. It checks whether impact came with a number and a mechanism, whether confidence was grounded in prior experiments or data, whether ease was estimated by whoever executes — and whether the team runs the feedback loop: revisiting predictions after each experiment to calibrate the next round.
What the Coach scores — element by element
Ask Loopion Coach evaluates ICE the way a rigorous practitioner would: each element scored Strong / Partial / Missing, each judgement backed by a quote or specific moment from the real conversation — never a single overall number with no breakdown.
Impact
Was the expected effect on the target metric stated with a number and a mechanism — 'this should lift activation ~5% because…'?
Confidence
Is confidence grounded in prior experiments, data or research — and does the team honestly score low confidence for novel bets? Uniformly high confidence means the dimension is decorative.
Ease
Was effort estimated by whoever executes, with hidden costs included — design, legal review, engineering?
The discipline test
Are scores recorded consistently, do decisions follow the ranking (with explicit reasons when overridden), and does the team compare actual impact vs predicted after each experiment? A team that never revisits predictions is not learning from ICE.
The failure patterns it catches
Frameworks fail quietly. These are the patterns the Coach flags specifically — with the evidence quoted, so the coaching conversation starts from facts rather than recollections:
- !
Impact claims with no number and no mechanism — enthusiasm dressed as estimation
- !
Confidence scores that are always high, making the dimension decorative
- !
Ease estimates from people who won't execute, missing design/legal/engineering costs
- !
No feedback loop: predictions never revisited after experiments run, so the team never calibrates
What you can ask
The Coach lives inside Ask Loopion. Ask in plain language — about one call, one person, one deal, or the whole team across a quarter:
Audit this week's growth meeting against ICE
Which experiments were decided against their ranking, and was a reason recorded?
Did we ever score anything low confidence this quarter?
Compare the predicted vs actual impact of the last five experiments we shipped
Whose ease estimates keep missing hidden design and legal costs?
What mechanism did we claim for the onboarding experiment's impact?
Who uses it
Growth leads use ICE auditing to keep experiment velocity honest; marketing ops runs the predicted-vs-actual loop across quarters; founders use it to see whether the growth process is actually learning or just busy.
How the platform backs it up
Every meeting captured
Loopion joins your Teams and Zoom meetings automatically and transcribes them in 99 languages — the Coach evaluates real conversations, not self-reported summaries.
Evidence, not opinion
Every score is backed by a quote or specific moment from the transcript. You can check every judgement.
Coaching that gets executed
Evaluations end with concrete actions — and Loopion's carry-forward brings unfinished coaching actions back into future meetings until they're done.
Defaults per team
Set ICE as the default framework for the right team in Settings → Ask Loopion Coaching, and every evaluation automatically uses it.
Or bring your own
Workspaces can define custom frameworks — your in-house methodology, scorecard or rubric — and the Coach evaluates against your criteria with the same element-by-element rigour.
Coach every growth & marketing conversation against ICE — automatically
Ask Loopion Coach is included in the Business and Enterprise plans. Every call scored, every judgement evidenced, every coaching action tracked to completion.