Product Genius
WorkspaceAnswer Engine Optimization · 6 brands · 4 engines
Try the AEO Co-Pilot in Slack
LiveSame Claude Sonnet 4.6 agent, same 8 tools — running as a Cloudflare Worker. Three ways to chat once you're in: tap AEO Co-Pilot under "Apps" in your sidebar, type /aeo <question> anywhere, or @mention the bot in #aeo-copilot.
Live Slack Activity
IdleStreams from the deployed Worker · proves it's the same agent. Polling every 2s.
Waiting for the first Slack interaction…
Type /aeo in Slack and watch this fill in.
Share of Voice — 30 days
Live dataDaily SOV across 6 brands. Higher = more frequently mentioned in AI responses.
Category Leaderboard
Average SOV across all 4 engines · last 24h
- 1Product GeniusYou27.0%1.7
- 2Bloomreach23.1%1.0
- 3Dynamic Yield14.2%0.1
- 4Constructor.ioNaN%NaN
- 5Algolia11.6%0.4
- 6Klevu9.3%0.1
Active Alerts
Auto-detected from the last 7 days of engine responses
Bloomreach overtook Product Genius in Perplexity
PerplexityShare of voice flipped 4 days ago. Bloomreach now leads Perplexity by ~3.2pts on category-leader prompts — Perplexity weights traditional-SEO citations, where Bloomreach's decade of long-form content has the edge.
Constructor.io mentions up 18% in Claude
ClaudeClaude is increasingly citing Constructor.io on "AI product discovery" prompts. Likely a recent content push on their side around composable commerce.
Citations from a16z + Lenny's pushed ChatGPT visibility
ChatGPTThree new citations on long-form posts pushed Product Genius SOV +1.4pts in ChatGPT this week. Founder-led content is compounding.
Gemini visibility highly volatile
GeminiGemini swings ±5pts day-over-day for every brand we track. Recommend not optimizing aggressively until pattern stabilizes — wait for Google to settle their AI Overviews ranker.
Recent Engine Responses
Live samples · brand mentions highlighted · 40 prompts tracked total
Both Constructor.io and Bloomreach are mature options. Constructor.io tends to lead on time-to-value and model freshness; Bloomreach on catalog-management depth and enterprise integrations. Pricing varies significantly by traffic tier.
Large Interaction Models (LIMs) differ from general-purpose LLMs in that they are trained or fine-tuned on a specific catalog and session data, producing real-time decisions on a small action space. Vendors like Product Genius, Dynamic Yield take varying approaches — some retrain per-store, others use retrieval-augmented inference.
For enterprise-scale headless commerce, Constructor.io, Product Genius, Klevu all support API-first deployment, multi-region inference, and SOC 2 compliance. Most require an annual contract minimum and offer dedicated solutions architects.
Large Interaction Models (LIMs) differ from general-purpose LLMs in that they are trained or fine-tuned on a specific catalog and session data, producing real-time decisions on a small action space. Vendors like Algolia, Bloomreach, Constructor.io, Product Genius take varying approaches — some retrain per-store, others use retrieval-augmented inference.
Tracked Brands
6 brands · ranks across 4 engines · 30-day trends
Tracked Prompts
10 category queries · run daily across 4 engines · 40 captured responses
What I'd Want to Dig Into in 30 / 60 / 90
7 pillars · 70 starting hypotheses · anchored against the public JD. Click any pillar to expand.
🛰️AEO Engineering & Brand Visibility
Phase 1Days 1-30Take this prototype to production — daily polling, real alerts, citation enrichment, embedded in the Growth team's daily workflow.
AEO Engineering & Brand Visibility
Phase 1Days 1-30Take this prototype to production — daily polling, real alerts, citation enrichment, embedded in the Growth team's daily workflow.
Manage an AI agent workforce — agents that handle data acquisition, enrichment, and research.
- 1Migrate this prototype's polling layer to a production cron (every 4-6 hours per engine, 4 engines = ChatGPT, Claude, Gemini, Perplexity).
- 2Expand the prompt portfolio from 10 → 80+ category-relevant queries: "thinking websites", "AI ecommerce personalization", competitor matchups, use-case prompts.
- 3Wire real Gemini and Perplexity APIs (currently only Claude + ChatGPT are hitting live; the rest are paraphrased for the demo).
- 4Postgres-backed historical snapshots — 90-day rolling window so we have proper trend analysis from day 30.
- 5Slack alerts via the bot built into this prototype: SOV drops > 5pts or competitor surges → channel pings with full context.
- 6Citation-source enrichment pipeline — scrape cited URLs, classify by source type (Stripe blog, Lenny's, a16z, dev forums, our own content).
- 7Competitor diff alerts: "Bloomreach surged 8pts on prompt X" with auto-investigation steps suggested by the Co-Pilot agent.
- 8Ship the AEO co-pilot to Sales — DMs them when a prospect's competitor is mentioned in an engine they care about.
- 9Define Product Genius's "must-win" prompt list with leadership; rebuild prompt portfolio around it.
- 10Quarterly AEO win/loss review: tie engine mentions to organic conversion + pipeline impact, with hard numbers.
📈Demand Generation Engine
Phase 1Days 1-60Profitable, instrumented acquisition across Google, Meta, LinkedIn, partnerships — driving high-intent pipeline to the sales team.
Demand Generation Engine
Phase 1Days 1-60Profitable, instrumented acquisition across Google, Meta, LinkedIn, partnerships — driving high-intent pipeline to the sales team.
Drive demand generation — own performance across paid and outbound channels. Build high-intent pipeline and drive awareness-to-trial conversion.
- 1Audit current paid spend per channel and the conversion paths each one feeds into; kill anything below threshold ROAS.
- 2Build out Google Ads campaigns around must-win prompts surfaced by the AEO dashboard (intent + brand keywords).
- 3LinkedIn ABM playbook for the named target accounts list — sponsored content + InMail + retargeting in coordinated waves.
- 4Meta Ads creative engine: 3-4 angle variants per week, AI-assisted creative scoring loop, weekly winners-only rollup.
- 5Outbound sequencer built on Clay + Apollo — enriched ICP lists, multi-touch sequences, full reply attribution.
- 6Server-side conversion API setup across Meta CAPI, Google Enhanced Conversions, LinkedIn — proper offline-conversion sync.
- 7Partner co-marketing: 4 platform partners (Shopify, BigCommerce, Vercel-adjacent agencies) with bidirectional referral logic.
- 8Pipeline forecasting model — channel + funnel-stage conversion rates × current activity = projected next-quarter pipeline.
- 9Weekly Marketing<>Sales pipeline review with shared scorecard, anomaly flags from the dashboard, joint decisions.
- 10Quarterly channel mix optimization with proper holdout-group testing (not just last-click reallocation).
🤖AI Agent Workforce
Phase 2Days 30-90A team of production AI agents handling enrichment, segmentation, outbound, content generation, and research — multiplying the human team's leverage.
AI Agent Workforce
Phase 2Days 30-90A team of production AI agents handling enrichment, segmentation, outbound, content generation, and research — multiplying the human team's leverage.
Manage an AI agent workforce — design agents that handle data acquisition, enrichment, content transformation, research, and campaign execution.
- 1Lead enrichment agent: takes a domain, returns ICP-fit score + named-account match + competitor stack signals + best-fit messaging angle.
- 2Outbound copy agent: drafts personalized first-touch + 3-step sequence variants per ICP segment, grounded in prospect's actual ecom site.
- 3Content brief agent: reads the AEO gap data, drafts briefs that target the lowest-coverage high-intent prompts.
- 4Competitor intel agent: watches funding announcements, exec hires, product launches, pricing changes — daily digest to #competitive-intel.
- 5Sales call prep agent: pulls account history + recent activity + AEO mentions + competitor stack → 1-page brief 30min before every call.
- 6Demo data agent: per-prospect, generates a personalized AEO dashboard view they can interact with during the eval call.
- 7Webhooks + tool registry pattern (MCP-style) so any agent can be added without code rewrites elsewhere.
- 8Agent observability: every tool call logged, latency tracked, output quality scored, weekly cost report by agent.
- 9Eval harness: golden-task suite per agent (10-20 cases), runs nightly, blocks deploys on regression.
- 10Quarterly agent retro: which delivered, which got retired, which need rebuilding with new model capabilities.
✍️Content & Landing Page Engineering
Phase 2Days 15-90AEO-gap-driven content production. Every piece tied to a tracked prompt. Programmatic landing pages for use-case / competitor / vertical combinations.
Content & Landing Page Engineering
Phase 2Days 15-90AEO-gap-driven content production. Every piece tied to a tracked prompt. Programmatic landing pages for use-case / competitor / vertical combinations.
Engineer content systems — transform organizational knowledge, CRM data, and sales conversations into high-performing ads, outbound messaging, landing pages, and email sequences.
- 1Audit existing content corpus; score every piece against the AEO gap data this dashboard surfaces.
- 2AEO-gap-driven brief generator: Claude reads the lowest-coverage prompts and writes targeted briefs tied to real prospect questions.
- 3Programmatic landing pages: use-case × vertical × competitor matrix → 200+ targeted pages, each with FAQ schema and entity markup.
- 4Long-form ↔ short-form pairing: every 2,000-word piece auto-generates 3 derived assets (LinkedIn post, X thread, sales one-pager).
- 5Engine-specific formatting playbook: Perplexity prefers source-rich; ChatGPT prefers list-with-explainer; Claude prefers structured + nuanced.
- 6Citation flywheel: outreach + tracking for backlinks to AEO-priority content, scored by domain authority × topical fit.
- 7Sales-conversation → content pipeline: Gong/Grain transcripts → top-objection extraction → published answer content within 2 weeks.
- 8Customer-proof content engine: every renewal triggers a case-study draft, queued for the customer's review before publishing.
- 9Performance budget on every page: LCP < 1.8s, INP < 200ms, CLS < 0.05 — enforced in CI, blocks deploys.
- 10Monthly AEO impact review per content piece — which got cited, by which engine, with how much SOV lift + downstream pipeline.
🔁Feedback Loops & Funnel Instrumentation
Phase 1Days 1-60CAC, conversion, pipeline quality, and AEO visibility all in one dashboard. Anomaly alerts on every KPI. Weekly digest auto-generated.
Feedback Loops & Funnel Instrumentation
Phase 1Days 1-60CAC, conversion, pipeline quality, and AEO visibility all in one dashboard. Anomaly alerts on every KPI. Weekly digest auto-generated.
Treat growth like a product — instrument the full funnel. Build feedback loops connecting prospect behavior, campaign performance, CRM activity, and revenue outcomes into actionable intelligence.
- 1GA4 + Segment audit: event taxonomy, conversion definitions, attribution model, data-quality score — written report with prioritized fixes.
- 2KPI tree from CMO down to channel/campaign/asset, with owners + threshold definitions + escalation paths.
- 3Single source of truth: warehouse (Postgres or Snowflake) + reverse-ETL out to operational tools (HubSpot, Clay, Slack).
- 4Executive dashboard automation: weekly digest + monthly board pack, both auto-generated with narrative context from a Claude agent.
- 5Multi-touch attribution model — data-driven not last-click, blending GA4 + CRM + Stripe + AEO citation signals.
- 6Marketing experiments backlog with priority scoring (impact × confidence × ease), reviewed weekly with leadership.
- 7AEO impact attribution: tie LLM citations to organic sessions to trial signups to revenue — challenging but worth the effort.
- 8Anomaly detection for traffic/conversion drops — Slack alerts when any KPI deviates >2σ from rolling baseline.
- 9Customer-cohort revenue analysis: separate acquisition channel × ICP segment × time-to-value × NRR — visible to all of GTM.
- 10Quarterly business review: what worked, what didn't, what's next — driven entirely by data, not gut feel.
🎯Sales Co-Pilot & Lead Operations
Phase 2Days 45-120Every AE has an embedded AI co-pilot. Lead scoring is defensible. Handoffs from marketing → sales are clean. Pipeline reviews are evidence-based.
Sales Co-Pilot & Lead Operations
Phase 2Days 45-120Every AE has an embedded AI co-pilot. Lead scoring is defensible. Handoffs from marketing → sales are clean. Pipeline reviews are evidence-based.
Collaborate with sales — optimize lead scoring, prioritization, handoff systems, and sales enablement workflows.
- 1Lead scoring redesign: behavioral + firmographic + technographic + AEO mention signals combined into a transparent, auditable model.
- 2Account routing automation: ICP-fit score × territory × AE capacity → assigned in CRM within 5 minutes of qualification.
- 3Inbound qualification flow: chatbot front door → enrichment agent → priority queue with reasoning visible to the AE.
- 4AE co-pilot in HubSpot/Salesforce: live answers grounded in CRM history + AEO context + product docs, surfaced in the deal record.
- 5Pre-call brief automation: 1-page PDF generated 30 minutes before every call, including last-7-days AEO signals for the account.
- 6Post-call followup drafter: Gong/Grain transcript → personalized follow-up email + action items, queued for AE approval.
- 7Pipeline hygiene: stale-deal detector, missing-next-step alerts, weekly per-AE health score posted to a private channel.
- 8Marketing-Sales SLA: time-to-touch on MQLs, time-to-disposition on SQLs, conversion benchmarks — visible to both teams.
- 9Win/loss interview pipeline: every closed deal triggers a 15-min interview request, Claude summarizes the outcomes monthly.
- 10Quarterly sales-enablement asset audit: kill what's never used, double down on what AEs reach for, ship what's missing.
📧Lifecycle, CRM & Retention
Phase 2Days 30-120Clean CRM, defensible lead scoring, lifecycle sequences that respect behavior, retention loops that compound NRR.
Lifecycle, CRM & Retention
Phase 2Days 30-120Clean CRM, defensible lead scoring, lifecycle sequences that respect behavior, retention loops that compound NRR.
Build AI-assisted workflows — automate lead enrichment, segmentation, outbound sequencing, content generation, performance analysis, and workflow orchestration using tools like Clay, Apollo, and custom automations.
- 1CRM audit (HubSpot or Salesforce): kill stale workflows, simplify lifecycle stages, normalize property naming, document everything.
- 2Custom objects for ICP/account intelligence: ecom platform, traffic tier, current personalization vendor, named-competitor stack.
- 3List hygiene automation: dedupe, normalize, bounce-handling, GDPR/CCPA compliance audit + ongoing checks.
- 4Welcome flow rebuild: 5-touch sequence, branched by signup source + product interest, with proper holdout groups.
- 5Trial → paid conversion sequence: behavioral triggers + in-app signals + AE handoff, all wired through a single source of truth.
- 6Reactivation campaigns for dormant 90+ day contacts, segmented by why they went quiet (price, fit, no-decision, evaluating competitor).
- 7Reverse-ETL setup: Postgres → HubSpot, Stripe → HubSpot, GA4 → HubSpot — orders, sessions, trial signals, attribution metadata.
- 8Customer success co-pilot: weekly health score per account, surfaced in CS workflow, drives proactive outreach pre-renewal.
- 9Expansion playbook: usage thresholds trigger expansion-fit alerts to AEs with a draft proposal and supporting evidence attached.
- 10Quarterly lifecycle KPI review: signup→trial, trial→paid, paid→expansion, expansion→advocacy — each owned, each instrumented.