ICE scoring framework for founders with too many ideas: how we pick what ships this week

Sunny Ray · Mar 13, 2026 · 5 min read

A founder team with ten good ideas and no scoring system will ship none of them. Three numbers per idea ends the debate in ten minutes.

On our closers call on March 12, we were drowning. Not in problems. In ideas. Good ones. Every one of them had a smart person arguing for it, and every one of them would have eaten a month. We had 19 active clients that day, up from 18, and the business was moving. That is exactly when idea overload is most dangerous, because momentum makes everything look possible.

Charles brought the fix: ICE.

Ideal land: the founder disease

The phrase we used on the call for what was happening was "ideal land." It is where founder teams go when every idea is evaluated on how good it would be if it worked perfectly. In ideal land, every idea wins, because every idea is imagined at its best.

The cost is invisible. Nobody says "let's not ship anything this month." You just end up with five half-started projects, a team pulled in five directions, and the original engine running a little slower because nobody is tending it.

ICE in one line

Score every idea from 1 to 10 on three things. Impact: if this works, how much does it move the business? Confidence: how sure are we that it will work? Ease: how easy is it to execute with the people and time we have now?

Most teams multiply the three. ProductLift's glossary describes it that way: score each factor from 1 to 10, multiply the three values, then rank from highest to lowest. Multiplying punishes any idea that is weak on one dimension, which is the point. A huge idea you have no confidence in should not beat a solid idea you can ship on Monday.

Then you rank. The list decides, not the loudest voice.

Our actual list that day

Here is what was on the table.

Turning daily podcast guests into monthly subscribers. We were taping guests every day, and the idea was a paid tier for guests who wanted ongoing value. High potential impact. Confidence was a guess.

A closed-door roundtable for nine warm VC leads, with no pitching allowed. Just investors in a room talking with each other and with us. Strong confidence, because these were warm relationships, and very easy to run.

A hybrid with a fast-growing network-matching competitor. Interesting, possibly big, but complex and dependent on another company's priorities.

Before ICE, all three got an equal hearing every time they came up. After ICE, the order was obvious within minutes, and the complex partnership dropped to the bottom without anyone having to argue it down.

Confidence is the honest number

Of the three scores, Confidence is the one founders inflate. Impact is easy to dream about. Ease is usually knowable. Confidence asks something uncomfortable: what evidence do we have that this will work?

A good test: if your confidence score is above 6, you should be able to name the evidence. A customer asked for it. A test already showed signal. A similar move worked before. If the only evidence is that everyone on the call likes the idea, score it 3 or 4 and move on.

The roundtable scored well on confidence because we already had nine warm leads who had told us they cared. The subscription idea scored lower because nobody had asked for it yet. That is the difference between a plan and a hope.

ICE also exposed what the high-scoring ideas had in common. The ones that ranked well reduced manual work or built on relationships we already had. That matched a theme from the same call: the business needed systems that ran without constant personal involvement, from educational videos clients watch before sales calls to deeper research on every lead before we meet them. Good scoring does not just rank ideas. It shows you what kind of idea your business needs right now.

Write it down where the team lives

The last item on our next steps list that day was to implement the ICE framework in a shared Notion. That step matters more than the scoring itself.

Ideas come back. The one you killed on Tuesday reappears on Thursday with a new argument. If the scores live in someone's head, you re-run the whole debate. If they live in a shared table, anyone can look and see why it ranked where it did, and the only way to move it up is to bring new evidence that changes a score.

Re-score when something changes, not every week. A new customer request, a failed test, a partner going quiet. Those are moments to update a number.

One more benefit showed up the same day. Once the list was scored, the team could commit to the top items with confidence. Emiliano was launching a climate tech VC podcast that weekend. The roundtable for warm VC leads moved forward. The ideas that scored lower did not die; they waited in the table for evidence. Nobody had to win an argument for that to happen.

Why this matters in a services business

We run a business where the core engine, campaigns and shows and calls, needs daily attention. Every new initiative competes with that engine for the same people. ICE protects the engine by making the cost of each idea explicit. Ease is really a question about whose time it will take and what stops while they do it.

Ten minutes of scoring saved us a month of half-built projects. That trade is available to any founder team willing to put three numbers next to each idea and believe them.

If you want your team's time on the engine that brings in buyers, not on building one from scratch, apply here.

Frequently asked questions