Definitions of the terms that show up across the platform — Signal, MIND Engine, Living PRD, Compounding Intelligence, RICE, and the rest. Linked inline wherever they appear elsewhere on the site.
A discrete unit of customer voice captured from any channel — a Slack message, an App Store review, a portal vote, a widget submission, a web scrape, an email. The atomic unit of a product intelligence layer. Every signal carries channel attribution, account context, timestamp, sentiment, and a link to the source. Signals are what the MIND Engine classifies, deduplicates, and links into clusters.
The discipline of synthesizing every customer signal — across channels, time, and source — into a persistent layer that informs product decisions with evidence. The category that replaces feedback management tools, ad-hoc spreadsheets, and PM memory as the source of truth for what to build. The full pillar definition lives at What is product intelligence?
Software that captures customer signals from multiple channels, classifies and links them automatically, and exposes the resulting intelligence as Living PRDs, evidence-backed RICE scores, competitive gap analyses, and natural-language queries answered with citations. A Product Intelligence Platform is the operational substrate of the product intelligence discipline.
The property of a product intelligence layer where every signal captured today increases the value of every signal captured tomorrow. Compounding comes from four mechanisms: deduplication gets smarter as signal history grows, sentiment trends become real as evidence accumulates, account-level reach densifies, and competitive context builds. A workspace 2,000 signals in produces categorically sharper artifacts than the same workspace at signal one.
The classification and linking pipeline at the heart of MindBacklog. Every signal that arrives runs through the MIND Engine in under a second — classified by signal type, linked to existing features, scored for sentiment, deduplicated against prior signals, and attached to workspace and account metadata. Without this layer, raw signals are noise; with it, they are intelligence.
The end-to-end flow from "signal arrives from a channel" to "signal is searchable, cited, and counted in every artifact." Stages include ingestion, classification (MIND Engine), feature linking, sentiment scoring, deduplication, account attribution, and storage. The pipeline runs continuously — most signals are classified and indexed within a second of capture.
A product requirements document generated from the signal layer, with citation chips on every paragraph linking to source channels. Unlike a static Confluence or Notion PRD, a Living PRD refreshes as new evidence arrives — citation counts update, sentiment trends shift, competitive context appears. The defining artifact of product intelligence. See the PRD generator tool for the full mechanism.
The central view in MindBacklog showing what is trending in the signal layer — features gaining signal volume, sentiment shifts, account reach changes, and new clusters surfacing. Replaces what other tools call a "feedback inbox" — the difference is that the Hub shows synthesis (clusters and trends), not unread items. See the feature page.
The persistent side panel available on every page in MindBacklog — a natural-language interface to the signal layer. Type "what are customers saying about export?" or "should we ship this quarter?" and Ask Mind returns a cited answer drawn from the full signal history. A few thousand signals in, Ask Mind answers with depth no general-purpose AI can match, because it has processed every signal your customers have sent.
A prioritization approach where every score input (reach, impact, confidence, effort in RICE; or business value, time criticality, risk reduction, and job size in WSJF) traces back to specific signals in the layer. The score is auditable: click reach, see the 47 signals that produced it. Distinguishes Product Intelligence Platforms from roadmap tools, which typically capture scores as PM gut numbers.
A prioritization framework: Reach × Impact × Confidence ÷ Effort. The default scoring framework in MindBacklog. Each input is populated from the signal layer — reach from unique accounts, impact from sentiment delta, confidence from signal density and recency, effort from PM input. Click any input on a MindBacklog RICE card to see the evidence behind the number.
Weighted Shortest Job First — a SAFe-derived prioritization framework: (Business Value + Time Criticality + Risk Reduction) ÷ Job Size. Available in MindBacklog as an alternative to RICE under advanced scoring settings. RICE is the default because it maps cleanly to signal-layer inputs; WSJF is supported for teams already running it.
The dashboard inside MindBacklog that mines competitor public reviews, changelogs, and feature pages — surfacing sentiment shifts after competitor releases, feature gaps your customers also complain about, and a gap matrix showing what they ship that you do not (and vice versa). Most product intelligence is internal; competitive intelligence completes the picture. See the feature page.
An automated email summarizing what changed in the signal layer overnight — new signals, trending clusters, sentiment shifts, competitor activity. Pushed to PM inboxes before the workday starts. The Monday-morning Slack-scrolling ritual replaced by 30 seconds of reading. Configurable cadence (daily, weekly) and filtering (by feature, sentiment direction, account tier).
The integrations that capture customer signals. MindBacklog supports 9+ direct channels — Slack, Email, App Store, Google Play, Public Portal, Feedback Widget, Web Scraper, Web Form, Reddit — plus unlimited via REST API, Zapier, and Make. Each channel preserves source metadata (timestamps, author, account, URL) so citations remain auditable. Jira is a sync target, not an inbound signal channel.
A single product's intelligence layer in MindBacklog. Each workspace has its own channels, signals, Living PRDs, roadmap, and competitive intelligence — fully isolated from other workspaces in the same account. Multi-product companies typically run one workspace per product. Workspaces are also the unit of access control: who can see which signals.
A startup phase of MindBacklog usage — used before a product has live customers and inbound signal volume. The system generates artifacts (PRDs, prioritization, competitive analysis) from product description and competitor research alone. As real signals start flowing, the workspace transitions to Active Mode and the artifacts begin compounding citations.
The chasm between captured customer feedback and a confident roadmap decision. The data is in the system (Slack, support widget, App Store, portal, sales notes). The connections between data points are not. The PM is doing the synthesis in human memory — inconsistently, partially, forgetting half by Friday. A Product Intelligence Platform is the answer to the synthesis gap. See the pillar page section on the synthesis gap.
The full category definition — what product intelligence is, the synthesis gap it solves, and why it compounds with every signal processed. About an 8-minute read.
Read pillar →The three-stage flow from a signal arriving in a channel to a Living PRD in engineering's hands. Capture, Compound, Prove.
See pipeline →The live demo — a mature workspace with real signals, Living PRDs, RICE scoring, Ask Mind, and the Competitive Intelligence dashboard. No signup.
Open demo →