# ICE Scoring: Complete Guide 2026 | Enrich Labs > Master ICE scoring in 2026: Impact, Confidence, Ease formulas, Sean Ellis origins, worked examples, RICE vs ICE, and a playbook to prioritize growth and marketing experiments without endless debate. _Source: https://www.enrichlabs.ai/blog/ice-scoring-complete-guide-2026_ --- # ICE Scoring: Complete Guide 2026 ## TLDR ICE scoring is a fast prioritization method that ranks ideas on **Impact**, **Confidence**, and **Ease**. [Sean Ellis](https://www.seanellis.me/) created it for growth teams so experiment backlogs stop dying in debate. [Growth Method](https://growthmethod.com/ice-framework/) (updated June 2026) still calls ICE the original scoring system for growth marketing, with the average formula **ICE = (Impact + Confidence + Ease) / 3** on a 1-10 scale. [ProductPlan](https://www.productplan.com/glossary/ice-scoring-model) and [Dovetail](https://dovetail.com/product-development/ice-scoring-model/) describe the common multiply variant **ICE = Impact × Confidence × Ease**. [Itamar Gilad](https://itamargilad.com/ice-scores/) treats either math as fine because scores only rank ideas relative to each other. Intercom later published [RICE](https://www.intercom.com/blog/rice-simple-prioritization-for-product-managers) (Reach × Impact × Confidence ÷ Effort) to fix ICE's missing audience-size factor. In 2026, ICE still wins when you need a shared language for weekly tests, marketing experiments, and early feature triage. Pair it with a north-star metric, evidence rules for Confidence, and a separate strategic track for big bets. Use [Helena](https://www.enrichlabs.ai) when the backlog is marketing work so scoring ideas and shipping campaigns live in one AI system. **Key takeaways:** - **Define I, C, and E before you score.** Write what a 1 vs a 10 means so Impact and Ease do not drift by scorer mood ([Growth Method](https://growthmethod.com/ice-framework/); [Neal Cabage](https://nealcabage.com/ice-scoring/)). - **Pick one formula and stick to it.** Average keeps scores on a 1-10 band; multiply stretches high/low ideas further ([Growth Method](https://growthmethod.com/ice-framework/); [ProductPlan](https://www.productplan.com/glossary/ice-scoring-model)). - **Score privately first.** Independent scores cut anchoring bias in group sessions ([Growth Method](https://growthmethod.com/ice-framework/)). - **Raise Confidence with evidence.** Gilad's method uses Confidence as the brake on pure opinion ([Itamar Gilad](https://itamargilad.com/ice-scores/)). - **Switch to RICE when reach differs a lot.** Intercom's model multiplies Reach and divides by Effort ([Intercom](https://www.intercom.com/blog/rice-simple-prioritization-for-product-managers)). - **Skip ICE as a full multi-year roadmap law.** ProductPlan flags it as relative triage, not multi-goal portfolio math ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model)). * * * ## What ICE Scoring Is ICE scoring assigns three numbers to each idea, then combines them into one rank score. The letters stand for: - **Impact:** How much the idea moves the metric that matters (activation, conversion, revenue, retention, pipeline). - **Confidence:** How strong the evidence is that Impact (and often Ease) estimates are right. - **Ease:** How easy the idea is to ship relative to time, skills, and dependencies. Some teams call the inverse **Effort**. [Neal Cabage](https://nealcabage.com/ice-scoring/) notes Ellis popularized the model for growth work; product managers later reused it for features and rollouts. Wikipedia’s entry on [Sean Ellis (entrepreneur)](https://en.wikipedia.org/wiki/Sean_Ellis_%28entrepreneur%29) summarizes the growth-hacking origin Cabage cites. [Itamar Gilad](https://itamargilad.com/ice-scores/) frames ICE as an upgrade on classic impact/effort grids because Confidence forces you to admit when a "huge" idea is only a vibe. Savio’s product roadmap guide states the multiply form **ICE Score = (Impact \* Confidence \* Ease)** on a 1-10 scale per factor ([Savio](https://www.savio.io/product-roadmap/ice-scoring-model/)). Two aggregation styles appear in practice: 1. **Average:** `(I + C + E) / 3` → score stays between 1 and 10 ([Growth Method](https://growthmethod.com/ice-framework/)). 2. **Product:** `I × C × E` → score ranges up to 1,000 if each factor is 10 ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model); [Dovetail](https://dovetail.com/product-development/ice-scoring-model/); [Neal Cabage](https://nealcabage.com/ice-scoring/) also shows multiply). Gilad allows either multiply or average because the absolute number is not a forecast. Sort the backlog. Ship or test the top ideas. Re-score when evidence changes. ICE sits next to other decision tools on Enrich Labs: [lead scoring](https://www.enrichlabs.ai/blog/lead-scoring-complete-guide-2026) ranks people, ICE ranks work. [SMART goals](https://www.enrichlabs.ai/blog/smart-goals-complete-guide-2026) set the target metric Impact should move. [SaaS metrics](https://www.enrichlabs.ai/blog/saas-metrics-complete-guide-2026) and [A/B testing](https://www.enrichlabs.ai/blog/what-is-ab-testing) supply the data that raise Confidence. * * * ## Why ICE Scoring Matters in 2026 ### Backlogs grow faster than capacity Every growth, product, and marketing team faces the same constraint Gilad names: more ideas than you can build or even test, a thin slice that will work, and debates driven by opinion and politics ([Itamar Gilad](https://itamargilad.com/ice-scores/)). ICE shortens that debate to three questions. You can still pick wrong ideas. You spend less time arguing about them. ### Experiment velocity still rewards simple filters Growth Method's June 2026 guide positions ICE for early-stage growth teams, high-velocity experiment programs, and cross-functional groups that need a shared language without heavy process ([Growth Method](https://growthmethod.com/ice-framework/)). That matches how modern marketing runs: weekly creative tests, landing page variants, email subject lines, and channel bets. Pair ICE with [conversion rate optimization](https://www.enrichlabs.ai/blog/conversion-rate-optimization-complete-guide-2026), [landing page optimization](https://www.enrichlabs.ai/blog/landing-page-optimization-complete-guide-2026), and [performance marketing](https://www.enrichlabs.ai/blog/performance-marketing-complete-guide-2026) so tests hit the queue in a clear order. ### Speed comes with subjectivity ProductPlan is blunt: ICE is almost completely subjective, two scorers can diverge, and a low Ease score can bury a high-Impact idea even when that idea is the right long-term bet ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model)). Dovetail repeats the same risk: excitement inflates Impact and understates work ([Dovetail](https://dovetail.com/product-development/ice-scoring-model/)). Treat ICE as a **relative** sort for the next sprint of tests, not a five-year roadmap oracle. Put multi-year platform work under [product-market fit](https://www.enrichlabs.ai/blog/product-market-fit-complete-guide-2026) and [brand strategy](https://www.enrichlabs.ai/blog/brand-strategy-complete-guide-2026) reviews instead. ### Marketing ops needs a filter that leads to shipped work When the backlog is ads, content, lifecycle, and SEO tasks, slow prioritization kills pipeline. ICE keeps the filter light. [Helena](https://www.enrichlabs.ai) then executes scored campaigns across channels so the ranking does not sit in a spreadsheet. See [agentic marketing](https://www.enrichlabs.ai/blog/agentic-marketing-complete-guide-2026), [AI marketing for B2B SaaS](https://www.enrichlabs.ai/blog/ai-marketing-for-b2b-saas-scale-pipeline-2026), and [marketing automation for startups](https://www.enrichlabs.ai/blog/marketing-automation-for-startups-complete-guide-2026). * * * ## How to Calculate an ICE Score ### Step 1: Lock the goal metric Write one primary metric for the scoring session: trial starts, demo bookings, paid conversion, activation rate, churn, or pipeline dollars. Impact is meaningless if each scorer invents a different north star. Tie the metric to [demand generation](https://www.enrichlabs.ai/blog/demand-generation-complete-guide-2026) or [B2B lead generation](https://www.enrichlabs.ai/blog/b2b-lead-generation-complete-guide-2026) if growth is the theme. ### Step 2: Agree the 1-10 rubrics Growth Method's calibration table is a useful default ([Growth Method](https://growthmethod.com/ice-framework/)): Factor Low (1-3) Mid (4-6) High (7-10) **Impact** Negligible lift Moderate, measurable lift Significant, data-backed lift **Confidence** Pure gut Some data or precedent Strong data, prior wins **Ease** Weeks of eng, complex deps Medium effort Days, minimal deps Cabage's Ease bands map time: long (3-6 months) near 1-2, short (~2 weeks) near 8-10 ([Neal Cabage](https://nealcabage.com/ice-scoring/)). Write your own version if "Ease" for you means cost, specialist hours, or compliance risk. ### Step 3: Score Impact, Confidence, Ease - **Impact:** Use past lift ranges when you have them. A homepage headline change that already worked on a similar page scores higher than a brand-new referral system with no internal precedent. - **Confidence:** Gilad treats this as evidence strength. User research, quantitative funnels, and prior tests raise Confidence; buzzwords and founder preference keep it low ([Itamar Gilad](https://itamargilad.com/ice-scores/)). - **Ease:** Ask engineering or the channel owner. ProductPlan warns that business people guessing technical Ease is a failure mode ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model)). ### Step 4: Combine with one formula **Average example** (Growth Method worked set, homepage headline): Impact 8, Confidence 7, Ease 9 → `(8 + 7 + 9) / 3 = 8.0` ([Growth Method](https://growthmethod.com/ice-framework/)). **Multiply example** (ProductPlan style): Impact 7, Confidence 6, Ease 5 → `7 × 6 × 5 = 210` ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model)). Cabage's Feature 1 example: Impact 9 × Confidence 8 × Ease 5 = **360** ([Neal Cabage](https://nealcabage.com/ice-scoring/)). Do not mix average and product ranks in the same sheet. Sort descending. Break ties by higher Confidence, then higher Ease for velocity, or by strategic fit outside the score. ### Step 5: Independent scores, then discuss outliers Growth Method's tips: score alone first, discuss big spreads (one person scores Impact 3, another 8), re-score as capacity and market shift, and treat scores as ranking aids not precision forecasts ([Growth Method](https://growthmethod.com/ice-framework/)). Dovetail suggests optional multi-person scoring to surface roadblocks ([Dovetail](https://dovetail.com/product-development/ice-scoring-model/)). ### Worked marketing backlog (average formula) Adapted from Growth Method's experiment table ([Growth Method](https://growthmethod.com/ice-framework/)): Experiment I C E ICE Rewrite homepage headline from customer interviews 8 7 9 8.0 Exit-intent popup with lead magnet 6 6 8 6.7 Referral program with double-sided incentive 9 5 4 6.0 Rebuild onboarding with personalized steps 9 6 3 6.0 The headline wins on **evidence + speed**, not raw Impact. Link the winner to [call to action](https://www.enrichlabs.ai/blog/call-to-action-complete-guide-2026) and [lead magnet](https://www.enrichlabs.ai/blog/lead-magnet-complete-guide-2026) craft, then run a clean test. * * * ## ICE vs RICE, PIE, and Other Frameworks ### RICE (Intercom) [Sean McBride at Intercom](https://www.intercom.com/blog/rice-simple-prioritization-for-product-managers) defined RICE as Reach, Impact, Confidence, and Effort: **RICE = (Reach × Impact × Confidence) / Effort** - **Reach:** customers or events per time window (example: customers per quarter). - **Impact:** multiple-choice scale (3 massive, 2 high, 1 medium, 0.5 low, 0.25 minimal). - **Confidence:** percent (100% high, 80% medium, 50% low). - **Effort:** person-months across product, design, and engineering. RICE measures "total impact per time worked." Use it when ideas hit very different audience sizes. Growth Method notes RICE was built to fix ICE's blind spot on reach ([Growth Method](https://growthmethod.com/ice-framework/)). ProductPlan contrasts the formulas directly ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model); [RICE glossary](https://www.productplan.com/glossary/rice-scoring-model)). GrowthMentor’s 2026 comparison of ICE, RICE, and Kano also credits Ellis for popularizing ICE on growth teams running many experiments ([GrowthMentor](https://www.growthmentor.com/blog/prioritization-frameworks)). Productboard’s framework glossary lists ICE beside RICE, Kano, and weighted scoring for product teams ([Productboard](https://www.productboard.com/glossary/product-prioritization-frameworks/)). ### PIE (CRO) Growth Method summarizes **PIE** (Potential, Importance, Ease) from Chris Goward / WiderFunnel for page-level A/B tests. Importance reflects traffic and value of the page under test ([Growth Method](https://growthmethod.com/ice-framework/)). Prefer PIE when the backlog is mostly on-page CRO, not net-new growth systems. Connect to [conversion rate optimization for ecommerce](https://www.enrichlabs.ai/blog/conversion-rate-optimization-ecommerce-complete-guide-2026) and [what is A/B testing](https://www.enrichlabs.ai/blog/what-is-ab-testing). ### PXL and heavier models Growth Method describes **PXL** (Peep Laja / CXL) as ten binary evidence questions that cut 1-10 subjectivity at the cost of time, and **DRICE** as a deep RICE variant for high-stakes ideas ([Growth Method](https://growthmethod.com/ice-framework/)). Atlassian's prioritization overview stresses that scoring cannot forecast market response and can feel rigid on abstract work ([Atlassian](https://www.atlassian.com/agile/product-management/prioritization-framework)). ### Quick chooser Need Prefer Weekly experiment velocity, similar audience **ICE** Different audience sizes / person-month rigor **RICE** Page CRO backlog **PIE** Heavy anti-bias evidence checklist **PXL** Multi-goal portfolio / roadmap years Weighted score, OKRs, not ICE alone For agency and SaaS GTM stacks, also map ICE winners into [content marketing strategy](https://www.enrichlabs.ai/blog/content-marketing-strategy-complete-guide-2026), [content marketing for SaaS](https://www.enrichlabs.ai/blog/content-marketing-for-saas-complete-guide-2026), [account-based marketing](https://www.enrichlabs.ai/blog/account-based-marketing-complete-guide-2026), and [sales enablement](https://www.enrichlabs.ai/blog/sales-enablement-complete-guide-2026). * * * ## How to Run ICE Scoring Without Gaming the Sheet ### 1\. Separate hygiene work from scored bets Cabage notes required work often scores low and should leave the scoring process ([Neal Cabage](https://nealcabage.com/ice-scoring/)). Keep compliance, security, and broken-funnel fixes on a must-do list. ICE ranks optional experiments and features. ### 2\. Protect Confidence from optimism Gilad's community vs upload-flow example: a trendy community idea can score high Impact on buzz and low Confidence; a user-requested upload fix with early metric lift wins on evidence ([Itamar Gilad](https://itamargilad.com/ice-scores/)). ICE does not kill the community idea. It says **test more** before you staff it. ### 3\. Stop Ease from trapping you in local optima ProductPlan warns sole reliance on ICE can push endless low-hanging fruit while starving larger investments ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model)). Gilad answers with a **strategic track** for big, high-risk ideas and different resource rules ([Itamar Gilad](https://itamargilad.com/ice-scores/)). Run a quarterly strategic ICE-free review for platform bets. ### 4\. Re-score on learning Gilad scores at triage, filtering, new evidence, and periodic hygiene ([Itamar Gilad](https://itamargilad.com/ice-scores/)). Growth Method: re-score when market, capacity, or goals move ([Growth Method](https://growthmethod.com/ice-framework/)). Outdated scores mislead the next sprint. ### 5\. Tie marketing ICE to channel reality Score Meta, Google, email, and SEO ideas with the same metric definition, then execute winners with [Meta ads targeting](https://www.enrichlabs.ai/blog/meta-ads-targeting-and-bidding-strategies-2026), [Google Ads audit](https://www.enrichlabs.ai/blog/google-ads-audit-checklist-2026), [best AI for Google Ads](https://www.enrichlabs.ai/blog/best-ai-for-google-ads-2026), [best AI for email marketing](https://www.enrichlabs.ai/blog/best-ai-for-email-marketing-2026), and [Shopify marketing automation](https://www.enrichlabs.ai/blog/shopify-marketing-automation-complete-guide). For local and vertical teams, score channel plays the same way you would score [HVAC SEO](https://www.enrichlabs.ai/blog/hvac-seo-complete-guide-2026) or [electrician SEO](https://www.enrichlabs.ai/blog/electrician-seo-complete-guide-2026) tests: Impact on booked jobs, Confidence from past GSC data, Ease of content and GBP work. ### 6\. Use tools that store scores and ship work **1\. Helena (Enrich Labs)** Helena is Enrich Labs' AI marketing agent. Score campaign and creative ideas with ICE, then run plan, copy, and channel work in the same system. Best for SaaS, DTC, agencies, and local teams that want one operator instead of a pile of point tools. Start at [enrichlabs.ai](https://www.enrichlabs.ai). **2\. Spreadsheet + product board** Intercom published a RICE sheet pattern you can mirror for ICE columns ([Intercom](https://www.intercom.com/blog/rice-simple-prioritization-for-product-managers)). Product boards (ProductPlan glossary context, Productboard framework docs) store scores next to roadmap items ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model); [Productboard](https://www.productboard.com/glossary/product-prioritization-frameworks/)). **3\. Research workspace for Confidence** Dovetail positions research systems next to ICE so Confidence is not pure guesswork ([Dovetail](https://dovetail.com/product-development/ice-scoring-model/)). Feed interview themes into Impact and Confidence before you staff builds. * * * ## ICE Scoring for Marketing, Product, and Growth Teams ### Growth and product - Build an **idea bank** Gilad-style: triage new ideas, filter for the next cycle, move winners to test or build, archive losers with notes ([Itamar Gilad](https://itamargilad.com/ice-scores/)). - Cap the scored set. Growth Method's starter: top 10 ideas, score in one session, run the top experiment that week ([Growth Method](https://growthmethod.com/ice-framework/)). - Link outcomes to [net revenue retention](https://www.enrichlabs.ai/blog/net-revenue-retention-complete-guide-2026), [churn rate](https://www.enrichlabs.ai/blog/churn-rate-complete-guide-2026), and [lifecycle marketing](https://www.enrichlabs.ai/blog/lifecycle-marketing-complete-guide-2026) when retention is the goal. ### Marketing and agencies - Score client or brand experiments against one KPI per sprint. - Prefer high Ease when media budgets are live and learning windows are short. - Use ICE in [RFP response](https://www.enrichlabs.ai/blog/rfp-response-complete-guide-2026) and [scope of work](https://www.enrichlabs.ai/blog/scope-of-work-complete-guide-2026) workshops so stakeholders see why test A beats pet idea B. - Fold winners into [event marketing](https://www.enrichlabs.ai/blog/event-marketing-complete-guide-2026), [webinar marketing](https://www.enrichlabs.ai/blog/webinar-marketing-complete-guide-2026), [co-marketing](https://www.enrichlabs.ai/blog/co-marketing-complete-guide-2026), and [UGC marketing](https://www.enrichlabs.ai/blog/ugc-marketing-complete-guide-2026) calendars. ### SaaS GTM - Score pricing tests, packaging pages, and onboarding changes with shared Impact definitions from [pricing strategy](https://www.enrichlabs.ai/blog/pricing-strategy-complete-guide-2026) and [product launch marketing](https://www.enrichlabs.ai/blog/product-launch-marketing-complete-guide-2026). - Keep competitive and messaging work connected to [brand positioning](https://www.enrichlabs.ai/blog/brand-positioning-complete-guide-2026) and [jobs to be done](https://www.enrichlabs.ai/blog/jobs-to-be-done-complete-guide-2026). - For AI-heavy roadmaps, score agent workflows with [AI agents](https://www.enrichlabs.ai/blog/ai-agents-complete-guide-2025) and [sales automation AI](https://www.enrichlabs.ai/blog/sales-automation-ai-complete-guide-2026) in mind so Ease reflects integration cost. * * * ## Common Mistakes - **No written scale.** Without definitions, a "7" means nothing ([Growth Method](https://growthmethod.com/ice-framework/); [ProductPlan](https://www.productplan.com/glossary/ice-scoring-model)). - **Scoring only in the room.** Anchoring pulls the group to the first loud number ([Growth Method](https://growthmethod.com/ice-framework/)). - **Inflating Impact for favorite ideas.** Cabage and Dovetail both flag subjectivity and mood ([Neal Cabage](https://nealcabage.com/ice-scoring/); [Dovetail](https://dovetail.com/product-development/ice-scoring-model/)). - **Ignoring Reach when audiences differ.** Switch to RICE ([Intercom](https://www.intercom.com/blog/rice-simple-prioritization-for-product-managers); [Growth Method](https://growthmethod.com/ice-framework/)). - **Treating ICE as customer research.** Dovetail notes ICE is not customer-centric unless you feed surveys and requests into scores ([Dovetail](https://dovetail.com/product-development/ice-scoring-model/)). - **Using ICE for the whole roadmap.** ProductPlan: better for relative shortlists and pre-work than multi-goal portfolios ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model)). - **Never revisiting scores.** Gilad's hygiene rule keeps idea banks from rotting ([Itamar Gilad](https://itamargilad.com/ice-scores/)). - **Leaving scored work in a doc.** Connect to delivery systems and agents ([Helena](https://www.enrichlabs.ai); [top marketing automation strategies](https://www.enrichlabs.ai/blog/top-marketing-automation-strategies-that-work-in-2025)). * * * ## FAQ ### What does ICE stand for in prioritization? Impact, Confidence, and Ease. Each factor is usually scored 1-10, then averaged or multiplied into one rank score ([Growth Method](https://growthmethod.com/ice-framework/); [ProductPlan](https://www.productplan.com/glossary/ice-scoring-model)). ### Who invented ICE scoring? Sean Ellis, known for coining "growth hacking" and for growth work at companies such as Dropbox and LogMeIn. Growth Method and Cabage both credit Ellis as creator/popularizer ([Growth Method](https://growthmethod.com/ice-framework/); [Neal Cabage](https://nealcabage.com/ice-scoring/); [Lenny's Newsletter interview notes](https://www.lennysnewsletter.com/p/the-original-growth-hacker-sean-ellis)). A SaaStr transcript with Ellis discusses scoring Impact, Confidence, and Ease on a 1-10 scale and treating a 10/10/10 idea as a strong test to run ([SaaStr](https://www.saastr.com/sean-ellis-founder-ceo-of-growthhackers-building-a-company-wide-growth-culture-video-transcript/)). Learning Loop’s glossary also uses the multiply form and credits Ellis ([Learning Loop](https://learningloop.io/glossary/ice-scoring-model)). ### How do you calculate an ICE score? Average method: `(Impact + Confidence + Ease) / 3` ([Growth Method](https://growthmethod.com/ice-framework/)). Multiply method: `Impact × Confidence × Ease` ([ProductPlan](https://www.productplan.com/glossary/ice-scoring-model); [Dovetail](https://dovetail.com/product-development/ice-scoring-model/)). Pick one method per backlog. ### What is a "good" ICE score? On the average scale, Growth Method treats about **7+** as high priority ([Growth Method](https://growthmethod.com/ice-framework/)). On the multiply scale, compare relative ranks inside your list rather than chasing a universal cutoff ([Itamar Gilad](https://itamargilad.com/ice-scores/)). ### ICE vs RICE: which should we use? ICE when you need speed and ideas share similar audience size. RICE when Reach differs and you want person-month Effort in the denominator ([Intercom](https://www.intercom.com/blog/rice-simple-prioritization-for-product-managers); [Growth Method](https://growthmethod.com/ice-framework/)). ### Does ICE only work for growth experiments? No. Cabage and Dovetail describe feature and backlog use. Ellis started in growth; product and marketing teams reused the same three factors ([Neal Cabage](https://nealcabage.com/ice-scoring/); [Dovetail](https://dovetail.com/product-development/ice-scoring-model/)). ### How do we stop people from gaming scores? Written rubrics, private first-pass scores, outlier discussion, and Confidence tied to evidence. Gilad's Confidence Meter concept exists to make evidence explicit ([Itamar Gilad](https://itamargilad.com/ice-scores/); [Growth Method](https://growthmethod.com/ice-framework/)). ### Should required work go through ICE? Usually no. Cabage recommends keeping required work out of the model so must-ship items are not buried by low Ease ([Neal Cabage](https://nealcabage.com/ice-scoring/)). * * * ## Conclusion ICE scoring is still the default lightweight filter for growth, product, and marketing backlogs in 2026. It is fast, shared, and good enough for weekly decisions. Ellis designed it for experiment velocity. Growth Method still documents the average formula and calibration tables. ProductPlan, Dovetail, and Gilad document multiply math, subjectivity risks, and evidence-based Confidence. Intercom's RICE model is the natural upgrade when Reach and Effort need explicit weight. Run ICE well: one goal metric, written 1-10 definitions, independent scores, outlier talks, re-scoring on learning, and a strategic track for big bets. Sort ideas. Test or ship the top of the list. Feed results back into Confidence. When the backlog is marketing, pair the framework with an operator that ships the work. [Helena](https://www.enrichlabs.ai) turns scored campaign ideas into live ads, email, and social so the ranking does not die in a spreadsheet. Start at [enrichlabs.ai](https://www.enrichlabs.ai).