By now you know what a good session needs: a shared currency, independent estimates from the people who do the work, and clustering that turns a scatter of ideas into a portfolio. The practical question is what you actually run it with. There are three honest families of option, and each is genuinely good at something — so let's compare them without pretending any of them is magic.
Option one: whiteboard workshop templates
The Miro and FigJam route. You grab an impact-effort template, everyone joins the board, and you fill it with sticky notes. These tools are excellent at what they're for — collaborative online whiteboards have become a default for distributed teams, and the big platforms serve tens of millions of users precisely because open-canvas brainstorming is a real need done well.
Where it fits: open-ended ideation, diagramming, any session where you want maximum flexibility and don't mind building the structure yourself.
What it costs, honestly: the flexibility is the cost. A blank template enforces nothing. It won't make people use shared units, won't stop one person placing every note, and won't cluster the result for you — you're back to the three failure modes from the last post, now on a nicer canvas. And the sizing is your problem: the sticky note that says "big win" carries no math behind it. The template gives you a place to have the conversation, not a mechanism that makes the conversation come out right.
Option two: scoring frameworks (RICE, ICE)
The spreadsheet route. Score each idea with a formula: RICE — Reach, Impact, Confidence, Effort — was popularized by Intercom's product team; its lighter cousin ICE — Impact, Confidence, Ease — comes from the growth-experimentation world. You assign numbers, the formula ranks them, and you have a defensible ordered list.
Where it fits: a product manager triaging a known backlog, where you want a repeatable, auditable score and you already have the estimates.
What it costs, honestly: two things. First, scoring frameworks are built for one estimator working through a list — they were designed for the PM at their desk, not for twelve people surfacing tacit knowledge together. They quantify a decision; they don't collect the raw material from across a team. Second, the numbers feel more objective than they are. A framework's own advocates warn that the scores are only as good as the guesses that feed them, and confidence is the input people fudge most. RICE and ICE are great at ranking what you already know; they're not a way to find out what your team knows.
A whiteboard gives you a place to talk. A scoring formula gives you a way to rank. Neither one gets twelve people to honestly surface, size, and cluster the work — which is the actual hard part.
Option three: a discovery-workshop consultant
The expert-facilitator route. You hire someone to run structured discovery sessions — interviews, workshops, an opportunity map, a prioritized roadmap at the end. Done well, this is the highest-quality option on the list: a skilled facilitator enforces shared units, draws out the quiet people, and clusters the result, because that's the craft you're paying for.
Where it fits: high-stakes, ambiguous, cross-functional strategy where the cost of getting it wrong dwarfs the fee, and you want an outside party to hold the room.
What it costs, honestly: money and calendar. Professional discovery and strategy engagements run into serious day rates, and the good ones take weeks — scheduling, interviews, synthesis, readout. For a foundational replatforming decision, that's proportionate. For "which of our thirty AI ideas do we start with next month," it's a cannon aimed at a paperwork jam. You also, at the end, own a document produced by someone who's now leaving — the tacit knowledge went into their synthesis, not into a living artifact your team can rerun.
Where this leaves you
Line them up and the gap is specific. Whiteboards are flexible but enforce nothing. Scoring frameworks rank but don't gather. Consultants do it all but at a cost and cadence that only makes sense for the biggest decisions. For the very common middle case — a team that wants to turn its collective AI ideas into a starting portfolio, quickly, without a five-figure engagement — none of the three is shaped right.
That middle case is exactly the shape worth naming: many people, tacit knowledge spread across them, ideas that need shared units and clustering, and a decision that's important but not existential. It wants the structure of a consultant, the self-serve speed of a whiteboard, and the honest math of a scoring framework — all at once, and in one short session. In the final post we'll get concrete about what that looks like: the three-bet portfolio logic, and when a focused thirty-minute collective session genuinely beats a quarter of analysis.