Page Cro

Page CRO

Audit and improve conversion performance on Tiger Data marketing pages. Combines page review, messaging analysis, friction analysis, and optional Clarity behavioral evidence to produce prioritized recommendations: copy fixes, structural changes, and test ideas.

When to use this skill

• When the user asks for CRO, conversion advice, or landing page feedback
• When the user wants CTA improvements, hero rewrites, or friction analysis
• When the user shares a page URL, screenshot, or Clarity CSV and asks "what would you change?"
• When the user mentions bounce rates, drop-off, or conversion issues on a specific page
• When the user asks "why isn't this page converting?" or "how do I improve this page?"

Scope

Included: Homepage, product pages, feature pages, campaign landing pages, pricing pages, blog CTA strategy.

Optional inputs: Screenshots, page URLs, Clarity CSVs, notes on traffic source and conversion goal.

Excluded: Signup flow internals, product onboarding, in-app modals (unless the modal is the primary conversion mechanism on the page).

Step 0: Pre-flight check

Read REFERENCES.md from the plugin root and run the pre-flight check described there. Call list_marketing_references() to verify Tiger Den is reachable. If it fails or the tool is not found, STOP — do not continue. Follow the error handling in REFERENCES.md.

Once Tiger Den is confirmed, fetch this skill's reference docs:

get_marketing_context(slugs: ["product-marketing-context", "brand-voice-guide"])

If Clarity CSVs are provided: Before proceeding to Step 1, invoke the clarity-analyzer skill to process the CSV data. Use those findings as behavioral evidence throughout the audit.

Step 1: Classify the page

Determine the page type and primary conversion goal. Ask the user to confirm if ambiguous.

Also identify:

• Traffic source assumptions — Where are visitors likely coming from? (organic search, paid ads, email, social, direct) This shapes intent expectations.
• Visitor intent — What problem are they trying to solve? What stage of evaluation are they in?

If the user provides traffic source or goal information, use it. If not, infer from the page type and content, and state your assumptions.

Step 2: Audit the page

Work through each dimension in the CRO checklist (see references/cro-checklist.md). For each dimension, assess the current state and note specific issues.

The six audit dimensions:

1. Value proposition clarity — Is it obvious what Tiger Data does and why it matters, within 5 seconds? Is the headline specific and benefit-driven, or vague and feature-first? 2. CTA hierarchy — Is there one clear primary CTA? Are secondary CTAs visually subordinate? Is the CTA copy action-oriented and specific ("Start your free trial" not "Get started")? 3. Trust and proof — Are there customer logos, case studies, metrics, testimonials, or third-party validation? Is the proof relevant to the visitor's likely concerns? 4. Objection handling — Does the page preemptively address common hesitations (pricing, migration difficulty, vendor lock-in, performance at scale)? Are answers specific or hand-wavy? 5. Scannability and structure — Can a visitor who skims get the core message? Are headings descriptive? Is the visual hierarchy working? Is there too much text before the first CTA? 6. Friction and cognitive load — Are there unnecessary form fields, confusing navigation choices, competing CTAs, jargon without context, or dead ends? Is the path from "interested" to "acting" as short as possible?

Folding in behavioral evidence

If Clarity data was processed in Step 0, integrate those findings into the relevant dimensions:

• Scroll depth data → Scannability (where do visitors stop reading?), CTA hierarchy (is the primary CTA above the scroll drop-off?)
• Click data → CTA hierarchy (what are visitors actually clicking?), Friction (are they clicking non-clickable elements or ignoring the primary CTA?)
• Rage clicks or dead clicks → Friction (broken expectations, confusing UI)

Don't just list the behavioral data separately. Weave it into the dimensional analysis as supporting evidence.

Step 3: Assess messaging alignment

Using product-marketing-context and brand-voice-guide, check:

• Positioning match — Does the page lead with the problem Tiger Data solves, not just features? Does it align with current positioning?
• Terminology — Are product names, feature names, and technical terms correct per the glossary?
• Voice — Does the copy sound like Tiger Data? Flag AI slop, marketing fluff, passive voice, em dashes, or generic developer-tool language.
• Competitive framing — If the page references competitors, does it follow the guardrails in product-marketing-context?

Keep this section focused on issues that affect conversion, not a full brand audit. If the messaging is generally on-brand, say so and move on.

Step 4: Produce the audit report

Structure the output as follows:

Executive summary

2-3 sentences. What's the page doing well? What's the biggest conversion risk?

Top conversion blockers

The 2-3 most critical issues preventing conversions. Each should include:

• What the problem is
• Why it matters (impact on conversion)
• Where on the page it occurs

Quick wins

Changes that are low-effort and high-confidence. These should be implementable in under an hour:

• Copy tweaks (headline rewrites, CTA label changes)
• Removing friction (unnecessary fields, confusing microcopy)
• Adding missing trust signals in obvious spots

High-impact changes

Changes that require more effort but would meaningfully improve conversion:

• Structural reorganization (section reordering, adding/removing sections)
• New content blocks (case study section, comparison table, objection-handling FAQ)
• Visual hierarchy fixes (CTA prominence, above-fold content)

Test ideas

A/B or multivariate test hypotheses. Consult references/experiments.md for a menu of experiments organized by page type -- pick tests that address the specific issues found in the audit, don't just list generic tests. Each test idea should include:

• What to test (control vs. variant)
• What you expect to happen and why
• What metric to watch

Optional: Rewritten hero / CTA set

If the user asks for it, or if the hero or CTAs are the primary conversion blocker, provide:

• A rewritten headline and subheadline
• Rewritten primary CTA (button copy + supporting text)
• Brief rationale for each change

Write these in Tiger Data's brand voice using the brand-voice-guide.

Step 5: Offer next steps

After delivering the audit, ask whether the user wants:

• A deeper dive on any specific dimension
• Help rewriting specific sections (hand off to brand-voice-writer)
• A visual mockup of recommended changes (hand off to page-mockup-builder)
• To implement changes on the live site (hand off to website-content-editor)
• A re-audit after changes are made
• Clarity data analyzed if they have CSVs but didn't provide them initially (hand off to clarity-analyzer)

Dependencies

• Required: Tiger Den connector (for product-marketing-context and brand-voice-guide)
• Optional: Clarity CSV exports (processed via clarity-analyzer skill), screenshots or page URLs for visual review

CRO Copy Patterns

Reusable copy patterns for improving conversion on marketing pages. Use these as starting points when rewriting headlines, CTAs, or proof sections. Always adapt to Tiger Data's brand voice.

Headline Patterns

Problem-first headline

Lead with the pain the visitor is trying to solve.

Pattern: [Pain point] + [resolution or promise] Example: "Tired of slow queries on growing data? Tiger Data handles millions of rows without breaking a sweat." When to use: Homepage, product pages, campaign landing pages where the audience has a known pain.

Outcome-first headline

Lead with the result the visitor wants to achieve.

Pattern: [Desired outcome] + [how/with what] Example: "Ingest millions of rows per second with PostgreSQL you already know." When to use: Feature pages, product pages where the capability is the differentiator.

Audience-first headline

Lead with who this is for, creating immediate identification.

Pattern: [Audience identifier] + [what they get] Example: "For teams running time-series at scale: the PostgreSQL database that keeps up." When to use: Campaign landing pages with a specific audience, comparison pages.

Proof-first headline

Lead with a verifiable result that builds credibility.

Pattern: [Specific metric or achievement] + [context] Example: "10x faster queries. 90% less storage. Same PostgreSQL." When to use: Pages targeting evaluators, pricing pages, competitive contexts.

CTA Copy Patterns

Action + outcome

Tell the visitor what they'll do and what they'll get.

Strong: "Start your free trial" / "See Tiger Data in action" / "Try it with your data" Weak: "Get started" / "Learn more" / "Submit" / "Click here"

Reduce commitment anxiety

Lower the perceived cost of clicking.

Strong: "Start free, no credit card" / "See the demo (2 min)" / "Explore the playground" Weak: "Request a demo" (implies a sales call) / "Contact us" (implies waiting)

Match the visitor's stage

Different CTAs for different evaluation stages.

Supporting text under CTAs

A single line under the button that reduces anxiety or adds specificity.

Examples:

• "Free for 30 days. No credit card required."
• "Takes 30 seconds. Use your existing PostgreSQL tools."
• "Join 10,000+ developers building with Tiger Data."

Proof Patterns

The specific metric

Quantified results beat qualitative claims.

Strong: "Acme Corp reduced query latency by 73% after migrating to Tiger Data." Weak: "Acme Corp saw significant performance improvements."

The named testimonial

Real people with real titles at real companies.

Strong: "Tiger Data handles our 2TB/day IoT ingest without breaking a sweat." -- Jane Smith, Platform Lead at Acme Corp Weak: "Great product, highly recommend!" -- Anonymous

The "people like you" signal

Social proof that matches the visitor's identity.

Strong: "Trusted by 500+ IoT teams running PostgreSQL in production" (on a page about IoT use cases) Weak: "Trusted by thousands of companies worldwide" (on every page, no matter the context)

The migration story

For visitors evaluating a switch, proof that switching works.

Strong: "DataCo migrated 5TB from InfluxDB in a weekend. Here's how." Weak: "Easy migration from any time-series database."

Objection-Handling Patterns

The preemptive answer

Address the objection before the visitor has to ask.

Pattern: [Acknowledge the concern] + [specific answer] + [proof] Example: "Wondering about lock-in? Tiger Data is built on PostgreSQL. Your data, your tools, your queries -- they all work the same way they always have. And if you ever leave, pg_dump works just fine."

The comparison reframe

When the objection is really "why not stick with what I have?"

Pattern: [Acknowledge the current solution works] + [specific limitation it hits] + [how Tiger Data handles it differently] Example: "PostgreSQL is great for most workloads. But when your time-series tables hit billions of rows, standard Postgres starts to slow down. Tiger Data's hypertables partition automatically, so queries stay fast as your data grows."

The FAQ that earns trust

Answer real questions with real specifics.

Strong FAQ answers:

• Include numbers, architecture details, or links to docs
• Acknowledge limitations honestly
• Point to case studies or benchmarks for proof

Weak FAQ answers:

• Generic reassurances ("We take this very seriously")
• Marketing language instead of substance
• Dodge the actual question

Anti-Patterns to Flag

These patterns hurt conversion and should be called out in audits:

• The "everything wall" -- listing every feature on a single page with no hierarchy or narrative
• The invisible CTA -- CTA exists but is visually indistinguishable from surrounding content
• The premature ask -- requesting high-commitment action (demo, pricing) before establishing value
• The jargon gate -- using internal or category-specific terminology without context
• The proof desert -- making claims with zero supporting evidence on the page
• The false urgency -- countdown timers, "limited time" language that feels manipulative for a B2B product
• The navigation trap -- so many links and options that the visitor can't find the intended path
• The mobile afterthought -- page designed for desktop and merely shrunk for mobile, breaking the conversion flow

CRO Audit Checklist

Use this checklist to systematically evaluate each dimension of a page's conversion performance. Score each dimension, note specific issues, and cite behavioral evidence where available.

1. Value Proposition Clarity

What to check:

• Can a new visitor understand what the product does within 5 seconds?
• Is the headline benefit-driven (problem → solution) or feature-first?
• Does the subheadline add specificity or just repeat the headline in different words?
• Is there a clear "who this is for" signal?
• Does the above-fold content match the visitor's likely intent given the traffic source?

Common failures:

• Headline is a tagline, not a value prop ("The future of data" vs. "PostgreSQL for time-series at any scale")
• Subheadline introduces jargon without context
• Value prop assumes the visitor already knows what the product category is
• Multiple competing value props with no clear primary message

2. CTA Hierarchy

What to check:

• Is there exactly one primary CTA that's visually dominant?
• Does the primary CTA appear above the fold?
• Does the CTA copy describe the action and its outcome ("Start your free trial" not "Submit")?
• Are secondary CTAs visually subordinate (ghost buttons, text links)?
• Is there a CTA visible at every major scroll depth? (Don't overdo it, but don't make visitors scroll back up)
• Does the CTA match the visitor's likely stage? (Top-of-funnel visitors aren't ready for "Buy now")

Common failures:

• Two equally prominent CTAs competing for attention ("Start trial" and "Book a demo" side by side with equal weight)
• Primary CTA is below the fold
• CTA copy is generic ("Learn more," "Get started," "Submit")
• No CTA until the bottom of a long page
• CTA doesn't match traffic source intent (paid ad promises a demo, landing page leads with "Start trial")

3. Trust and Proof

What to check:

• Customer logos (recognizable, relevant to the target audience)
• Quantified results ("3x faster queries" backed by a case study, not just claimed)
• Testimonials with names, titles, and companies (not anonymous quotes)
• Third-party validation (awards, analyst mentions, open-source community signals)
• Social proof signals (GitHub stars, community size, adoption metrics)
• Security and compliance badges if relevant to the audience

Common failures:

• Logos present but no context (who are these companies? what did they achieve?)
• Testimonials are generic ("Great product!" — no specifics about the problem solved)
• No proof at all on a page asking for a high-commitment action (demo request, pricing inquiry)
• Proof is relevant to a different audience than the page targets
• Metrics without context ("10,000 users" — is that a lot for this category?)

4. Objection Handling

What to check:

• Does the page address the top 3-5 objections for this page type and audience?
• Are objection responses specific (data, architecture details, migration stories) or vague reassurances?
• Is there an FAQ section that covers real questions, not softball marketing prompts?
• For pricing pages: is there a clear path for "I need to talk to someone" visitors?
• For migration-heavy products: is the migration path addressed proactively?

Common objections by page type:

• Homepage: "What is this, exactly?" "Is it enterprise-ready?" "How is it different from X?"
• Pricing: "What happens when I outgrow this tier?" "Are there hidden costs?" "Can I try before I commit?"
• Feature page: "Does it actually work at my scale?" "How hard is the integration?"
• Comparison page: "Is this comparison fair?" "What are the tradeoffs you're not mentioning?"

Common failures:

• Objections acknowledged but not actually answered ("We take security seriously" with no details)
• FAQ is clearly written by marketing, not informed by actual customer questions
• Major objections completely ignored (the page pretends migration is effortless when it's a known pain point)

5. Scannability and Structure

What to check:

• Can a skimmer get the core message from headings alone?
• Are headings descriptive ("Ingest millions of rows per second") or vague ("Performance")?
• Is there a logical information hierarchy? (Problem → Solution → Proof → CTA)
• Are paragraphs short (2-4 sentences max for web)?
• Is there effective use of visual breaks (whitespace, icons, illustrations)?
• Does the page have a clear narrative arc, or does it feel like a random collection of sections?

Common failures:

• Wall of text with no visual breaks
• Headings are single-word labels ("Features," "Benefits," "Pricing") instead of persuasive statements
• Most important information is buried mid-page
• Page structure doesn't match how the visitor's decision process works
• Scroll depth data shows major drop-off before key content

6. Friction and Cognitive Load

What to check:

• Form fields: is every field necessary? Can any be deferred to later?
• Navigation: is the path from "interested" to "acting" obvious and short?
• Choices: is the visitor asked to make decisions they don't have enough context for?
• Language: is there jargon, acronyms, or assumed knowledge that could confuse a new visitor?
• Technical barriers: does the page work on mobile? Are load times acceptable?
• Dead ends: after reading a section, does the visitor know what to do next?

Common failures:

• Form asks for phone number, company size, and use case before the visitor has seen any value
• Page offers 5+ navigation paths with no clear recommended one
• Pricing tiers differ on dimensions the visitor can't evaluate yet
• Technical content assumes familiarity with the product category
• Mobile experience is a degraded afterthought
• Rage clicks on non-interactive elements (text that looks like a link, images that look clickable)

Page CRO Experiment Ideas

Comprehensive list of A/B tests and experiments organized by page type. Use this as a menu when generating test ideas in Step 4 of the audit -- pick experiments that address the specific issues found in the audit, don't just list generic tests.

Homepage Experiments

Hero Section

Trust & Social Proof

Features & Content

Navigation & UX

---

Pricing Page Experiments

Price Presentation

Pricing UX

Objection Handling

Trust Signals

---

Demo Request Page Experiments

Form Optimization

Page Content

CTA & Routing

---

Resource/Blog Page Experiments

Content CTAs

Resource Section

---

Landing Page Experiments

Message Match

Conversion Focus

Page Length

---

Feature Page Experiments

Feature Presentation

Conversion Path

---

Cross-Page Experiments

Site-Wide Tests

Navigation Tests

Page Type Frameworks

Reference frameworks for evaluating conversion performance by page type. Each framework describes the page's job, what "good" looks like, and the most common conversion failures.

Homepage

The job: Orient new visitors, build credibility, and route them to the right next step. The homepage serves multiple audiences simultaneously, so its primary challenge is clarity without oversimplification.

What good looks like:

• Above-fold hero clearly states who the product is for and what problem it solves
• One primary CTA with a clear secondary option (e.g., "Start free trial" primary, "See how it works" secondary)
• Social proof within the first scroll (logos, metrics, or a strong testimonial)
• Clear paths for different visitor types (developers, decision-makers, evaluators)
• Narrative arc from problem to solution to proof to action

Common conversion failures:

• Hero is a brand statement, not a value proposition
• Tries to say everything, communicates nothing
• No CTA above the fold
• Social proof is generic or absent
• Navigation overwhelms instead of guiding

Key metrics to watch: Scroll depth (do visitors get past the hero?), CTA click rate, bounce rate by traffic source, navigation click distribution.

Product / Feature Page

The job: Convince an interested visitor that this specific capability solves their specific problem, and move them toward trial or demo.

What good looks like:

• Opens with the problem this feature solves, not the feature itself
• Shows the feature in context (architecture diagram, code snippet, dashboard screenshot)
• Includes at least one quantified proof point specific to this feature
• Addresses the "but does it work at my scale?" objection
• CTA matches the visitor's evaluation stage

Common conversion failures:

• Feature-first framing ("We support X") instead of problem-first ("When you need to X")
• No proof that the feature works in production
• Technical details without practical context (specs without "so what")
• CTA is "Start trial" when the visitor is still evaluating
• No connection to adjacent features or the broader platform story

Key metrics to watch: Time on page (are they reading or bouncing?), CTA click rate, scroll depth to proof section.

Pricing Page

The job: Help the visitor choose a plan confidently. The pricing page is where evaluation becomes commitment, so clarity and trust are critical.

What good looks like:

• Tiers are clearly differentiated by use case, not just feature counts
• Recommended tier is visually highlighted
• Feature comparison table is scannable (not 50 rows of checkmarks)
• FAQ addresses real pricing objections (hidden costs, overage charges, scaling costs)
• Clear path for "I need to talk to someone" visitors
• Free tier or trial option reduces commitment anxiety

Common conversion failures:

• Tiers differ on dimensions the visitor can't evaluate yet ("100 hypertables" means nothing to a new visitor)
• No recommended tier, forcing the visitor to figure out which one fits
• Feature comparison is exhaustive but not useful (lists every feature, highlights none)
• No FAQ or the FAQ is softball marketing questions
• Enterprise tier is a black box ("Contact us" with no indication of what to expect)
• Annual/monthly toggle creates confusion or suspicion

Key metrics to watch: Plan selection distribution, FAQ expansion clicks, "Contact sales" vs. self-serve ratio, bounce rate.

Campaign Landing Page

The job: Convert a specific audience on a specific action. Landing pages are the most constrained page type: one audience, one goal, one CTA.

What good looks like:

• Headline directly continues the promise from the ad or email that drove the click
• One CTA, repeated at logical intervals
• Minimal navigation (or none) to prevent escape routes
• Proof is specific to the campaign audience
• Page length matches the commitment level (short for free resource, longer for demo request)

Common conversion failures:

• Message mismatch between ad/email and landing page (visitor expected one thing, got another)
• Full site navigation included, inviting visitors to wander away
• Multiple CTAs competing (sign up AND download AND watch a video)
• Generic proof not relevant to the campaign audience
• Form asks for too much too soon

Key metrics to watch: Conversion rate (primary metric), bounce rate, form abandonment rate, time to conversion.

Blog Post

The job: Earn trust, demonstrate expertise, and create a natural path to product awareness. Blog posts are top-of-funnel content, so the conversion expectation is softer.

What good looks like:

• Content delivers genuine value (the reader learns something or changes their thinking)
• CTAs are contextually relevant to the post topic, not generic
• Internal links connect to related content and relevant product pages
• Author byline builds credibility
• "What to read next" section keeps the visitor on-site

Common conversion failures:

• Hard-sell CTA jammed into educational content ("Now sign up for Tiger Data!")
• No CTA at all (the reader finishes and has nowhere to go)
• Generic sidebar CTA that has nothing to do with the post topic
• No internal links to related content or product pages
• Content is thin or AI-generated and doesn't earn trust

Key metrics to watch: Scroll depth (did they read it?), internal link clicks, CTA click rate, time on page, return visits.

Comparison Page

The job: Help an evaluator compare Tiger Data against alternatives fairly, building trust through transparency and winning on merits.

What good looks like:

• Comparison is structured and easy to scan (table format for feature comparison)
• Honest about tradeoffs (where the alternative is stronger, acknowledge it)
• Differentiators are specific and verifiable, not subjective claims
• Includes a migration path or switching cost analysis
• Proof comes from customers who switched, not just marketing claims

Common conversion failures:

• Obviously biased comparison that damages credibility
• Feature-by-feature comparison without context on what matters
• No acknowledgment of the alternative's strengths
• Comparison is out of date
• No social proof from customers who actually evaluated both options
• Doesn't address the "good enough" objection (why switch if the current solution works?)

Key metrics to watch: Scroll depth to comparison table, CTA click rate, bounce rate vs. other page types, referral source (are these visitors coming from competitive search queries?).