HubSpot Digital Marketing cheat sheet

A contact record is the unique object keyed by email address; a company record is a separate object linked to contacts, so dedupe rules differ for each.

Lifecycle stage is a single-select property with fixed default values (subscriber, lead, marketing qualified lead, sales qualified lead, opportunity, customer, evangelist, other); it moves forward by definition, so use it for funnel position, not for recency.

Lead status is a separate multi-value property describing sales follow-up (new, open, in progress, connected, unqualified); lifecycle stage answers where the contact is, lead status answers what sales is doing.

List membership is dynamic (filter-based, recalculated on property change) or static (snapshot, changed only by manual add or workflow); a contact can appear in many lists, so lists do not replace a segmentation property.

Contact properties have field types (single-line text, number, date picker, enumeration, calculation) that constrain filtering, workflows, and personalization tokens; changing a property type after data exists can invalidate dependent assets.

A calculated contact property derives its value from other properties using formulas and does not store manually editable input; it recalculates when inputs change, so it is unsuitable for values needing audit history.

Merging two contact records keeps the most recent value for most properties and preserves the surviving record ID; list memberships and activity history from both records are combined, so merges are effectively irreversible.

Consent for communication is tracked per subscription type (opt-in, opt-out, not specified) with a lawful basis and timestamp; a global unsubscribe suppresses all marketing email regardless of other list membership.

Property history and change source (form, import, workflow, integration, manual) are recorded per property; this is the primary audit trail for debugging why a contact received a specific message.

Content pillars map to buyer-stage intent: awareness content answers problem questions, consideration content compares approaches, decision content compares vendors; a single asset usually serves one stage well.

The buyer's journey stages (awareness, consideration, decision) determine content format: educational blog and video for awareness, guides and case studies for consideration, demos, pricing pages, and trials for decision.

Content clusters pair one comprehensive pillar page with multiple supporting posts that all link to the pillar; internal links pass authority and signal topical relevance to search engines.

Topic clusters outperform isolated posts because search engines evaluate site-level topical coverage, not just single-page keyword density.

Content repurposing follows a hub-and-spoke model: one long-form asset becomes short posts, email snippets, video scripts, and social variants, multiplying reach without new research.

Content audit criteria: traffic, conversions/assists, backlinks, and freshness; pages with low traffic and no conversions are candidates for consolidation, redirect, or removal.

A content calendar specifies owner, publish date, target keyword, persona, funnel stage, and CTA per asset; missing CTA assignment is the most common cause of traffic without leads.

Content ROI compares cost of production and promotion against attributed leads and revenue; assisted conversions must be counted or long-cycle content appears unprofitable.

Gated versus ungated is a trade-off: gating captures contact data but reduces reach and backlinks, so gate only assets with high perceived value at consideration and decision stages.

Deliverability depends on authentication (SPF, DKIM, DMARC), list hygiene, and engagement; authentication proves sender identity but does not guarantee inbox placement.

Hard bounce = permanent failure (invalid address) and should be suppressed immediately; soft bounce = temporary failure (full mailbox, server issue) and is retried before suppression.

Open rate is unreliable as a primary metric because image blocking and privacy features (Apple Mail Privacy Protection) pre-fetch pixels and inflate opens; prefer click and conversion metrics.

Click-through rate = unique clicks / delivered; click-to-open rate = unique clicks / unique opens and measures content relevance among engaged readers only.

List segmentation by lifecycle stage, engagement recency, and purchase history typically lifts click rates more than subject-line testing alone.

A/B testing requires one variable per test and enough sample size for significance; testing subject line, send time, and content simultaneously makes results uninterpretable.

Workflow (automation) emails triggered by behavior outperform batch blasts on click rate because timing matches expressed intent; re-engagement sequences usually suppress non-openers after a defined window.

Personalization tokens pull contact property values into subject lines and body; a missing or empty property renders a default value, so fallback text must be set before send.

CAN-SPAM and GDPR require a working unsubscribe in every marketing email and honored opt-outs within the statutory window; transactional emails are exempt from most marketing consent rules but not from identification requirements.

Conversion rate = conversions / visitors; form conversion rate = submissions / form views, so a low form conversion rate with high traffic points to form friction, not traffic quality.

Form field count trades data quality against completion: each added field reduces completion, so request only fields needed for the immediate routing decision and enrich the rest later.

Progressive profiling collects additional fields across repeat visits instead of all at once, raising completion while building the profile over time.

Progressive profiling depends on cookie or contact recognition; anonymous visitors on a new device will see the full first-time form again.

Lead scoring combines explicit fit (job title, company size, industry) with implicit engagement (page views, email clicks, form fills) to rank readiness; negative scores handle unqualified fits.

A lead is marketing qualified when it meets a score or fit threshold; it becomes sales qualified only when sales accepts it, so MQL and SQL counts must be reported separately.

Landing page CRO levers: single conversion goal, no navigation, benefit-led headline, social proof, and short form; removing site navigation is the highest-leverage single change.

Offer-to-audience match determines lead quality: a top-funnel offer (checklist) generates volume with low intent, while a bottom-funnel offer (demo) generates fewer, higher-intent leads.

Lead routing assigns new leads by territory, company size, or round-robin; routing rules must be tested because a misassigned lead stalls in the funnel without appearing in anyone's queue.

Keyword metrics: search volume estimates demand, keyword difficulty estimates link authority needed, and cost-per-click is a proxy for commercial intent.

Search intent classification (informational, navigational, commercial, transactional) must match page type; a transactional page targeting an informational query will not rank regardless of optimization.

On-page elements: title tag, meta description (not a ranking factor but affects click-through), H1, URL, internal links, and image alt text; the title tag is the strongest on-page text signal.

Technical crawlability requires robots.txt rules, XML sitemaps, canonical tags, and no accidental noindex; a canonical tag consolidates duplicate URLs but does not block crawling.

Page speed and Core Web Vitals (LCP, INP, CLS) are ranking inputs; image compression, caching, and reduced third-party scripts are the usual fixes.

Backlinks are evaluated by relevance, authority, and anchor text; one relevant editorial link typically outweighs many low-quality directory links, and disavow is rarely needed for normal sites.

Rank tracking must be segmented by device and location because results are personalized and localized; a single national average hides local ranking gaps.

Internal linking distributes authority from high-traffic pages to priority pages; orphan pages (no internal links) are crawled rarely and rank poorly.

Structured data (schema markup) enables rich results and improves click-through but is not a direct ranking factor; invalid markup can suppress the rich result entirely.

Organic reach measures unique users who saw a post without paid distribution; it is capped by platform ranking algorithms that favor early engagement.

Engagement rate = engagements / reach (or / impressions); choosing reach as the denominator makes rates comparable across posts with different follower counts.

Platform ranking signals weight recency, early engagement velocity, and relationship history; posting when the audience is active raises initial velocity.

Paid social targeting options: core audience (demographics, interests), custom audience (CRM list upload, site visitors via pixel), and lookalike audience built from a seed custom audience.

A social pixel enables retargeting and conversion tracking; without it, paid social can only report clicks and impressions, not downstream conversions.

Social listening tracks brand mentions and competitor share of voice; it measures sentiment and volume, not attribution to revenue.

UGC (user-generated content) and employee advocacy extend reach because audiences engage more with people than brands, but usage rights must be secured in writing.

Attribution for social is usually last-touch-poor: social often appears as an assist earlier in the journey, so multi-touch or first-touch models reveal its contribution.

Community management response time is a service metric; unanswered comments and DMs reduce both engagement and trust, and platform algorithms penalize low response rates on some networks.

The inbound flywheel has three stages: attract (content, SEO, social), engage (forms, email, chat, sales conversations), and delight (support, onboarding, advocacy); momentum compounds because delighted customers feed referrals back into attract.

Inbound replaces the funnel metaphor because the flywheel retains and recycles contacts; the funnel treats the customer as an endpoint and loses post-purchase momentum.

Buyer personas define goals, pain points, and objections for a segment; content and messaging decisions should trace to a specific persona rather than a generic audience.

The attract stage is measured by traffic and new contacts; engage by MQLs, conversion rates, and sales-accepted leads; delight by retention, NPS, and referrals.

Inbound is pull-based: the prospect finds the brand through search, content, or referral. Outbound is push-based: interruption via ads, cold email, or calls. Most programs need both, but the metrics differ.

Content offers must align to funnel stage; a mismatch between ad copy, landing page, and follow-up email breaks the promise and depresses conversion at each step.

The flywheel's force is friction reduction: every removed step (fewer form fields, faster reply, clearer pricing) increases the rate at which contacts move forward.

Inbound requires a closed loop between marketing and sales: shared definitions of MQL and SQL, feedback on lead quality, and SLA on follow-up time are the operational core.

Attribution models: first-touch credits the first interaction, last-touch credits the final one, linear splits evenly, and time-decay weights recent touches; the model choice changes which channel looks effective.

Sessions and contacts are different units: sessions count visits (one contact can generate many), contacts count unique people; mixing them in one chart overstates reach.

Conversion rate = conversions / sessions; when comparing channels, confirm the denominator is consistent, because some reports use sessions and others use contacts.

Campaign reporting ties assets to a campaign parameter (UTM or campaign property); without consistent tagging, traffic appears as direct or referral and attribution breaks.

A/B test validity requires sufficient sample size and a single variable; stopping a test early on a favorable result inflates false positives.

Cohort analysis groups contacts by acquisition month and tracks retention or revenue over time; it separates growth from churn, which aggregate totals hide.

Funnel reports show drop-off between stages (visit to lead to MQL to customer); the largest percentage drop identifies the highest-leverage optimization target.

Dashboards should separate leading indicators (traffic, MQLs, pipeline created) from lagging indicators (revenue, retention) because leading metrics allow corrective action before results are locked.

Data hygiene affects reporting: duplicate contacts, missing lifecycle values, and untagged campaigns create gaps that look like performance changes but are measurement errors.

Lifecycle stage defaults run subscriber, lead, MQL, SQL, opportunity, customer, evangelist, other; the value reflects funnel position, not engagement recency.

Lead status is a separate property describing sales follow-up state such as new, open, in progress, connected, or unqualified.

Dynamic lists recalculate automatically when a contact's properties change; static lists are snapshots changed only by manual edits or workflows.

Merging contacts keeps the surviving record ID and most recent property values, and combines list memberships and activity history; the merge cannot be undone.

A calculated contact property is computed from other properties and is not manually editable, so it cannot store values requiring an audit trail.

A global unsubscribe suppresses all marketing email for a contact regardless of which lists they belong to.

Property change history records the source of each edit, such as form, import, workflow, integration, or manual entry.

Awareness content answers problem questions; consideration content compares approaches; decision content compares vendors, demos, and pricing.

A pillar page covers a broad topic and links to supporting cluster posts, which all link back to the pillar to consolidate topical authority.

A content audit scores pages on traffic, conversions or assists, backlinks, and freshness to decide keep, consolidate, redirect, or remove.

Gating an asset captures contact data but reduces reach and backlinks, so gate only high-value consideration and decision assets.

Every content asset needs an assigned persona, funnel stage, target keyword, and CTA; a missing CTA is the usual cause of traffic without leads.

Content ROI must include assisted conversions, or long sales-cycle content appears unprofitable under last-touch measurement.

SPF, DKIM, and DMARC authenticate a sending domain; they prove sender identity but do not by themselves guarantee inbox placement.

Hard bounces are permanent failures and should be suppressed immediately; soft bounces are temporary and are retried before suppression.

Open rate is inflated by pixel pre-fetching and image blocking, so click and conversion metrics are more reliable.

Click-to-open rate equals unique clicks divided by unique opens and measures content relevance among engaged readers.

A/B tests should change one variable at a time; simultaneous changes to subject line, send time, and content make results uninterpretable.

Personalization tokens render fallback text when a property is empty, so defaults must be set before sending.

Form conversion rate equals submissions divided by form views, isolating form friction from traffic quality.

Each additional form field lowers completion, so request only fields needed for immediate routing and enrich later.

Progressive profiling collects new fields across repeat visits, raising completion while building the profile over time.

Lead scoring combines explicit fit attributes with implicit engagement signals, and may include negative scores for poor fits.

MQL counts and SQL counts must be reported separately because SQL requires sales acceptance of the lead.

Removing site navigation from a landing page is typically the highest-leverage single conversion optimization change.

Top-funnel offers generate volume with low intent; bottom-funnel offers generate fewer leads with higher intent.

Search volume estimates demand, keyword difficulty estimates required link authority, and cost-per-click is a proxy for commercial intent.

Search intent must match page type; a transactional page cannot rank well for an informational query.

The title tag is the strongest on-page text signal; the meta description affects click-through but is not a ranking factor.

A canonical tag consolidates duplicate URLs for indexing but does not block crawling; use robots.txt or noindex to block.

Core Web Vitals LCP, INP, and CLS are ranking inputs improved by image compression, caching, and fewer third-party scripts.

Backlinks are judged on relevance, authority, and anchor text; one relevant editorial link often outweighs many directory links.

Orphan pages with no internal links are crawled rarely and rank poorly, so internal linking is a technical priority.

Organic reach counts unique users who saw a post without paid distribution and is limited by platform ranking algorithms.

Engagement rate should use a consistent denominator such as reach or impressions so posts with different follower counts stay comparable.

Custom audiences are built from CRM list uploads or site visitors tracked by a pixel; lookalike audiences are modeled from a seed custom audience.

Without a social pixel, paid social can report clicks and impressions but not downstream conversions.

Social listening measures mention volume and sentiment, not revenue attribution.

Social often appears as an assist earlier in the journey, so multi-touch or first-touch models reveal contributions that last-touch hides.

The flywheel stages are attract, engage, and delight; delighted customers feed referrals back into attract, so momentum compounds.

Attract is measured by traffic and new contacts, engage by MQLs and conversion rates, and delight by retention, NPS, and referrals.

Inbound is pull-based through search, content, and referral; outbound is push-based through ads, cold email, and calls.

Buyer personas define goals, pain points, and objections for a segment, and every content decision should trace to one persona.

Friction reduction is the flywheel's force: fewer form fields, faster replies, and clearer pricing increase the rate contacts advance.

A closed loop requires shared MQL and SQL definitions, lead-quality feedback, and a follow-up SLA between marketing and sales.

First-touch credits the first interaction, last-touch the final one, linear splits credit evenly, and time-decay weights recent touches more heavily.

Sessions count visits and contacts count unique people; one contact can generate many sessions, so mixing the units overstates reach.

Funnel reports expose percentage drop-off between stages; the largest drop identifies the highest-leverage optimization target.

Cohort analysis groups contacts by acquisition period and tracks retention or revenue over time, separating growth from churn.

Leading indicators such as traffic, MQLs, and pipeline created allow correction before lagging revenue results are locked.

Duplicate contacts, missing lifecycle values, and untagged campaigns create reporting gaps that look like performance changes but are measurement errors.

Lifecycle stage is a single-select property with fixed default values from subscriber through evangelist, and represents funnel position rather than engagement recency.

Lead status is a separate property describing sales follow-up state, such as new, open, in progress, connected, or unqualified.

Dynamic lists recalculate automatically on property change, while static lists change only through manual edits or workflows.

Merging two contacts keeps the surviving record ID and most recent property values, combines activity history and list memberships, and cannot be undone.

A calculated contact property is computed from other properties and is not manually editable, so it cannot hold values needing an audit trail.

A global unsubscribe suppresses all marketing email for a contact regardless of which lists they belong to.

Property change history records the source of each edit, such as form, import, workflow, integration, or manual entry.

Awareness content answers problem questions, consideration content compares approaches, and decision content compares vendors, demos, and pricing.

A pillar page covers a broad topic and links to supporting cluster posts, which each link back to consolidate topical authority.

A content audit scores pages on traffic, conversions or assists, backlinks, and freshness to decide keep, consolidate, redirect, or remove.

Gating an asset captures contact data but reduces reach and backlinks, so gate only high-value consideration and decision assets.

Every asset needs an assigned persona, funnel stage, target keyword, and CTA; a missing CTA is the usual cause of traffic without leads.

Content ROI must include assisted conversions, or long sales-cycle content appears unprofitable under last-touch measurement.

SPF, DKIM, and DMARC authenticate a sending domain; they prove sender identity but do not by themselves guarantee inbox placement.

Hard bounces are permanent failures and should be suppressed immediately, while soft bounces are temporary and retried before suppression.

Open rate is inflated by pixel pre-fetching and image blocking, so click and conversion metrics are more reliable indicators.

Click-to-open rate equals unique clicks divided by unique opens and measures content relevance among engaged readers.

A/B tests should change one variable at a time; simultaneous changes to subject line, send time, and content make results uninterpretable.

Personalization tokens render fallback text when a property is empty, so defaults must be configured before sending.

Form conversion rate equals submissions divided by form views, isolating form friction from traffic quality.

Each additional form field lowers completion, so request only fields needed for immediate routing and enrich later.

Progressive profiling collects new fields across repeat visits, raising completion while building the profile over time.

Lead scoring combines explicit fit attributes with implicit engagement signals and may include negative scores for poor fits.

MQL counts and SQL counts must be reported separately because SQL requires sales acceptance of the lead.

Removing site navigation from a landing page is typically the highest-leverage single conversion optimization change.

Top-funnel offers generate volume with low intent, while bottom-funnel offers generate fewer leads with higher intent.

Search volume estimates demand, keyword difficulty estimates required link authority, and cost-per-click is a proxy for commercial intent.

Search intent must match page type; a transactional page cannot rank well for an informational query.

The title tag is the strongest on-page text signal, while the meta description affects click-through but is not a ranking factor.

A canonical tag consolidates duplicate URLs for indexing but does not block crawling; robots.txt or noindex blocks crawling.

Core Web Vitals LCP, INP, and CLS are ranking inputs improved by image compression, caching, and fewer third-party scripts.

Backlinks are judged on relevance, authority, and anchor text; one relevant editorial link often outweighs many directory links.

Orphan pages with no internal links are crawled rarely and rank poorly, so internal linking is a technical priority.

Organic reach counts unique users who saw a post without paid distribution and is limited by platform ranking algorithms.

Engagement rate should use a consistent denominator such as reach or impressions so posts with different follower counts stay comparable.

Custom audiences are built from CRM list uploads or site visitors tracked by a pixel, while lookalike audiences are modeled from a seed custom audience.

Without a social pixel, paid social can report clicks and impressions but not downstream conversions.

Social listening measures mention volume and sentiment, not revenue attribution.

Social often appears as an assist earlier in the journey, so multi-touch or first-touch models reveal contributions that last-touch hides.

The flywheel stages are attract, engage, and delight; delighted customers feed referrals back into attract, so momentum compounds.

Attract is measured by traffic and new contacts, engage by MQLs and conversion rates, and delight by retention, NPS, and referrals.

Inbound is pull-based through search, content, and referral, while outbound is push-based through ads, cold email, and calls.

Buyer personas define goals, pain points, and objections for a segment, and every content decision should trace to one persona.

Friction reduction is the flywheel's force: fewer form fields, faster replies, and clearer pricing increase the rate contacts advance.

A closed loop requires shared MQL and SQL definitions, lead-quality feedback, and a follow-up SLA between marketing and sales.

First-touch credits the first interaction, last-touch the final one, linear splits credit evenly, and time-decay weights recent touches more heavily.

Sessions count visits and contacts count unique people; one contact can generate many sessions, so mixing the units overstates reach.

Funnel reports expose percentage drop-off between stages, and the largest drop identifies the highest-leverage optimization target.

Cohort analysis groups contacts by acquisition period and tracks retention or revenue over time, separating growth from churn.

Leading indicators such as traffic, MQLs, and pipeline created allow correction before lagging revenue results are locked.

Duplicate contacts, missing lifecycle values, and untagged campaigns create reporting gaps that look like performance changes but are measurement errors.