AI SEO: Original Source Assets

Original source asset development is the top technique in the MERIT Framework. The right asset builds AI visibility for years. The wrong asset, built with the same effort, produces a polished page nobody cites. This chapter covers the five viable asset types. It walks through the decision framework. It maps platform and industry variants. It runs the selection workshop teams use. Most teams pick their asset type in a single short meeting. The teams that earn category-defining AI visibility treat the choice as strategy work with its own outputs.

Why This Technique Matters

The default way teams pick an asset is to pick what feels most rigorous, which is usually research. Or to pick what the team already knows how to make, which is usually blog posts dressed up as opinion. Both defaults are common. Both are usually wrong. They optimize for what looks defensible in a planning meeting, not for what earns citations.

The strategic frame is different. The right question is not "what feels most rigorous" or "what can we already produce." It is "which asset earns the most sustained AI citations in our category." Asked the right way, the answer is rarely the default. The asset that converts effort into citations most efficiently in one category is often the asset a team would never have picked on instinct.

The stakes are real. Original assets are the one durable citation surface a brand controls, because they are the one place a brand can add the net-new information gain AI systems filter for. Mentions techniques depend on third parties for placement and timing. Relevance techniques only tune what you already have. If the source content is derivative, structure tweaks have a ceiling. Inclusion techniques are technical baseline work. Entity optimization and crawler access make assets findable but do not earn citations alone.

Original source asset development is the one technique where the brand controls the input, the timing, and the topic. Pick the wrong asset and you waste the strategic lift the rest of the Evidence pillar is built to amplify.

Original source asset development is the one technique where the brand controls the input, the timing, and the topic.

The cost of getting this wrong is opportunity cost. A rigorous-looking asset that earns few citations is not just a stalled asset. It is also the several other assets that never got made, because the operator time and editorial attention were already spent. One bad asset choice costs the framework, the calculators, and the opinion pieces that could have run instead.

The Five Viable Asset Types

Five asset types reliably earn AI citations when built well and promoted right. AI systems do not prefer one over another. What looks like format preference is usually about promotion quality and co-citation depth. Pick the type that fits your situation. Do not pick the type that feels most defensible.

Five cards, one per asset type. Frameworks are a named structure that supplies category vocabulary. Opinion is defensible operator reasoning under a named byline. Research is original data with a documented method. Calculators are tools that return a personal numeric answer at a unique URL. Templates are transferable artifacts the audience can copy and use. Each card names what the type is and the condition under which it wins. — Figure 1. The five viable asset types. Each type earns AI citations under a different category condition, shown on the right of each card.

Expert Frameworks and Methodologies

A framework is a named structure that organizes existing ideas. It helps people think about a problem in a new way. The MERIT Framework, Jobs-to-be-Done, Growth Loops, RACI, AIDA, the messy middle, and similar widely-cited frameworks share three traits. A memorable name. A set of components. A path for how to use it. Frameworks earn citations because they give writers, analysts, podcasters, and operators a shared vocabulary for a complex topic. Once a framework is named and adopted, it travels by reference, not by quote.

Once a framework is named and adopted, it travels by reference, not by quote.

When frameworks win. Categories where the current vocabulary is borrowed, vague, or in conflict. A clear, well-built framework replaces the borrowed words because it is easier to use. Categories where the same problem comes up across many surfaces. Think analyst reports, industry publications, podcasts, conferences. Categories where operator experience patterns repeat across many cases. Frameworks are syntheses, not new data. Strategic intent is thought leadership and category positioning, not transactional capture.

When frameworks lose. The category already has one or two named frameworks with strong analyst adoption. The team cannot sum up the framework in one sentence or one slide. The target audience is buyers in late-stage decision, not category-shapers. The brand has no named author (founder, executive, named SME) to attribute the framework to.

What it takes: operator time for synthesis, an editorial pass, diagram design (the highest-yield design work for a framework), a canonical page build, and promotional copy. The work runs in sequence: foundation first, then distribution, then depth and reinforcement.

Citation profile: slow ramp at the start. Then it compounds once the first analyst or industry publication uses the framework name. Frameworks have the most durable citation curves of any asset type. Each future piece written about the topic may cite the framework by name.

Common framework mistakes. The forgettable name. Terms like "Strategic Alignment Methodology" are not names. Frameworks need names a sharp non-expert can recall after hearing once. Over-factoring. Twelve parts in nested matrices. The limit is three to seven parts total. No application path. Definition-only frameworks without diagrams, templates, or worked examples do not travel. Borrowed framework with new words. Relabeling existing methods fools nobody. Either build new work or cite the original.

Worked example: the MERIT Framework itself. Cody C. Jensen drafted the framework from sustained patterns across Searchbloom partner engagements. He iterated on the name and the parts over multiple cycles. He designed a single diagram. The canonical whitepaper went up at searchbloom.com/merit-framework-whitepaper/. Distribution ran through three contributed pieces in industry publications, twelve LinkedIn long-form posts, and four podcast appearances. The work expanded to a per-pillar playbook in April 2026. Citation density tracked through the cadence in Chapter 13. Citation outcome: the framework name is now retrieved consistently for queries about AI SEO, AEO, and GEO.

Expert Opinion and Analysis

Opinion is defensible operator reasoning. It runs under a named expert byline. The topic is one where the operator's experience produces a non-obvious view. Most B2B teams underuse opinion because they treat it as risky. The risk (being wrong in public) is what makes opinion citable. AirOps's March 2026 analysis found AI systems cite opinion at rates close to empirical research when the source has verified expertise and the claim is specific.

When opinion wins. The operator has deep domain experience or runs a brand with visible category authority. The category has new questions where data does not yet exist. The team needs fast citation traction. The team has little production capacity beyond operator time. The category has reactive moments (platform changes, competitor moves, rule shifts) where speed and sharpness produce outsized citation density.

When opinion loses. The brand has no recognizable expert byline. The category rewards rigorous data above all else. Think hedge fund research, regulatory analysis, deep technical fields. The team is not comfortable with sharp claims. Hedged "it depends" pieces do not earn citations no matter how credible the operator.

The three traits of citable opinion. Specific. The claim is concrete enough to disagree with. "Content marketing matters" is invisible. "Most B2B blog posts published in 2026 have zero AI citation potential because they restate widely available consensus" is citable. Defensible. The opinion is backed by reasoning the reader can follow and judge, even though opinion is not a research finding. Counter to consensus. The opinion differs from the current consensus in at least one specific way. Opinions that restate consensus add no information gain and do not get cited.

What it takes: operator drafting time, an editorial pass, light design, and distribution through the operator network plus optional PR coordination. Long-form canonical pieces move from idea to publication fast. Reactive opinion publishes promptly after the trigger event. Citation traction begins reasonably quickly for sharp pieces with a credible byline.

Citation profile: spiky early with a long tail. Sharp opinions compound. Each new query on the topic may surface the original piece. Operators on a steady monthly cadence build entity-topic links that AI systems learn to weight. After sustained steady publishing, the operator becomes a citable source. Their opinion pieces earn citations at multiples of the average rate.

Common opinion mistakes. Hedged opinion. "Five things to consider" listicles or "it depends" pieces do not count as citable opinion. They are analysis without a position. Restated consensus. Opinion that matches what everyone is already saying adds no information gain. The piece may be correct and still earn no citations. Brand-byline. Opinion under the brand name with no attributed person earns lower citation rates. Use the founder, the CEO, or a named SME. Wide-topic publishing. Operators who publish across too many unrelated topics weaken the entity-topic link. Stay within the topical cluster that matches operator expertise.

Data-Driven Research

Original research earns AI citations when the method is credible and the findings come in formats AI systems can extract. The visibility upside is real. The time to citation is long. The failure modes are harsh. Under the wrong conditions, research is the most expensive asset type to get wrong. Under the right conditions, it is the most defensible competitive moat.

When research wins. The category has missing or stale data. The brand has real research infrastructure. Think panel access, statistics capacity, or partnerships with trade groups or schools. The category has analyst coverage that rewards proprietary data.

When research loses. The brand does not have research infrastructure. Underpowered work substitutes, and low sample sizes get filtered out by the analyst tier whose co-citations create AI visibility. The category lacks analyst coverage. The team lacks method rigor or partnerships.

What it takes. The options span a range of research infrastructure. DIY survey tools are the lightest option, but their sample sizes are usually too small for citable research. Managed survey platforms are the middle option: Wynter and Centiment offer B2B-validated panels, SurveyMonkey and Typeform are general-purpose with weaker B2B panel quality, and Qualtrics is enterprise-grade. Full research firms are the option for analyst-tier categories. Research belongs in the portfolio when the brand has real research infrastructure, through a managed platform, a research firm, or in-house capacity. When it does not, the brand is better served by another asset type than by underpowered work that the analyst tier will filter out.

Sample size sets whether journalists, analysts, and AI systems will cite the numbers. Minimum viable: 500-750 respondents (+/-4-5% margin of error at 95% confidence). Target: 1,000-2,000 respondents (+/-2-3% margin of error) is the sweet spot. Premium: 5,000+ respondents lets you cut the data by segment, which yields several citable claims from one survey. Stratifying for segments needs at least 200-300 respondents per segment.

Research moves slowly from design to publication, then needs further sustained time for citation accumulation. Plan for the asset to reach steady-state value only after a long ramp.

Method rigor. Sample selection. Document the recruitment method, screening rules, and response rates. Statistics. Descriptive stats, confidence intervals on key claims, and significance tests for comparisons. Limitations section. Every credible study has one. The analyst tier reads method first and findings second.

Citation profile: slow to start but durable. Statistical claims with proper attribution show up in AI citations for years if the method holds up to scrutiny. The compounding is strongest for research that becomes the "category statistic" everyone cites.

Common research mistakes. Underpowered samples. Surveys under 500 respondents for broad claims fail analyst credibility. Low-power research is worse than no research. Missing method page. Research without a documented method gets dismissed by the analyst tier. Chapter 5 covers method docs in depth. Image-locked statistics. Numbers locked inside infographics are AI-invisible. Always pair visuals with parsable HTML, covered in Chapter 5. Gated research. Gating the research behind a form means AI systems cannot index it.

Interactive Calculators and Tools

Calculators are the most underrated original source asset type. They earn citations because the URL returns a personal numeric answer. That is exactly what generative AI systems need to surface for quantitative queries. Categories full of gated PDFs and weak comparison pages can be displaced by one well-built calculator that takes far less effort than original research.

When calculators win. Buyer questions are quantitative ("how much will X cost," "what is the ROI of Y," "how big a Z do I need"). Current market answers are gated, stale, or fuzzy. The brand has domain expertise that maps to a clear method. Engineering capacity exists for the build, or no-code platforms like Convertful, Outgrow, or Wufoo can substitute.

When calculators lose. Buyer questions are qualitative. The team plans to gate the calculator behind a form. Gated calculators are AI-invisible. The method cannot be documented. Rules limit what the calculator can claim.

High-value calculator types. ROI calculators. Return on investment for a category-relevant decision. Comparison calculators. Compare options across multiple dimensions. Assessment tools. Score current state against a benchmark. Cost estimators. Estimate cost or sizing for a service or build.

What it takes: operator time for method design, design and UX work, an engineering build (or a no-code platform in place of it), benchmark data sourcing, and promotion. Calculators move from method design to launch fairly quickly, with citation traction following.

Design principles for citation. Transparent method. The calculator shows its logic, inline or via a dedicated method link. Unique result URLs. Each calculation produces a URL that returns that specific result. This is the most important design call for AI citation. Downloadable results. PDF or CSV with the method built in. Benchmark context. Show how the user's result compares to benchmarks.

The gating dilemma. The most common calculator mistake is gating results behind an email form for lead generation. The form captures a meaningful but modest share of users as leads. But AI systems cannot fill out forms. That means zero AI citations from gated calculators. The lead capture comes at the cost of the citation. The right pattern: give the basic result ungated (AI systems can index and cite it). Offer an enhanced report (extra breakdowns, exportable formats, custom benchmarks) behind an optional email gate.

Two linked panels showing a two-stage pattern. Stage one is the ungated base result that AI systems index and cite: a basic numeric result, a visible method, and a unique shareable URL, which earns the citation. An arrow labeled then leads to stage two, an optional gated enhanced report behind an email form with extra breakdowns, exportable formats, and custom benchmarks, which captures the email. — Figure 2. The ungated-base, gated-enhanced pattern. AI systems cannot fill out forms, so the base result stays ungated to earn the citation while an optional gated enhanced report captures the lead.

Common calculator mistakes. Opaque method. Calculators that show one number with no reasoning earn lower citation rates. Show the formulas inline or via a method link. No unique result URLs. Calculators that show results only on the original page, with no shareable URL per calculation, are UI features and not citable assets. Stale benchmarks. Benchmark data, pricing assumptions, and reference rates change. Quarterly refresh is the working cadence.

Worked example: Chime built a series of consumer-banking calculators (overdraft savings, ATM-fee comparison, paycheck-advance ROI) with a transparent method and unique result URLs. Each result page indexed on its own. AI Search citations grew from near-zero to single-digit shares of consumer-banking queries. Refresh velocity (quarterly benchmark updates per the cadence in Chapter 13) drove the steady citation curve. The public Chime case study reported citations tripled quickly after the refresh program standardized.

Templates and Downloadable Assets

Templates earn citations because they give the audience something to take away and use. Process checklists, planning templates, scoring rubrics, and decision matrices all serve the same role. They turn expertise into a transferable artifact. AI systems surface templates because the queries ("what is the right structure for X," "is there a template for Y") are common and the available answers are often weak.

When templates win. The category is operational, with recurring work audiences need to run well (project management, hiring, sales ops, marketing ops, finance close). The brand has tribal knowledge that can be codified into a transferable artifact. The team has little production capacity beyond operator time. Templates convert effort into citations as efficiently as any asset type. The target audience is operator-grade, not executive.

When templates lose. The category is strategic rather than operational. Strategic categories reward frameworks and opinion. The template would be seen as too generic to stand apart. The brand cannot commit to refresh cycles. Year-stamped templates need yearly refresh.

Template categories. Process templates. Workflows, SOPs, quality control. Planning templates. Project roadmaps, sprint planning, content calendars, campaign briefs. Analysis templates. SWOT matrices, competitive scorecards, decision frameworks. Assessment templates. Audit checklists, maturity models, scoring rubrics.

Format hierarchy. Google Sheets is highest citation value. It is editable, collaborative, and easy to share. The audience can copy and customize without breaking the original. Excel is high citation value. It is full-featured, offline, and familiar for enterprise audiences. PDF is lowest citation value among editable formats. It is viewable but not editable. PDFs earn fewer citations because users cannot use them without retyping.

What it takes: operator design, format conversion, a branded design header, and distribution through the operator network. Templates move from idea to publication fast. Citation traction begins reasonably quickly when promoted on community surfaces.

Common template mistakes. PDF-only. PDF templates earn far fewer citations than editable formats. The audience needs to copy and customize directly. Generic titles. "Project planning template" is generic. "Project planning template for SaaS marketing teams" earns citation when the user's query matches the specifier. No refresh cycle. Year-stamped templates decay without yearly refresh. Yearly refresh is itself a citation event. Bloated structure. Templates with too many fields or too much text do not travel. The audience wants a working tool, not a manual.

Citation Surface Yield by Asset Type

Asset selection works best when the operator can compare how efficiently each asset type converts effort into citations. The Citation Surface Yield is a Searchbloom-coined framework that compares that conversion across asset types, measured across mid-market engagements at an early mark and a later mark. The CSY makes the asset-type comparison concrete instead of intuitive. Per-asset yield tracks closely with the count of distinct, sourced insights each asset carries, which Chapter 5 measures as Information Gain Density.

Asset type	Early yield	Later yield	Citation curve
Frameworks	Moderate	Compounds strongly as the vocabulary travels	Slow start, steep compounding after the first analyst pickup
Opinion	High	Compounds modestly	Fast start, long tail when the operator sustains monthly cadence
Research	Low	Compounds the hardest for category-statistic-grade studies	Very slow start, durable plateau
Calculators	Moderate	Compounds strongly with quarterly benchmark refreshes	Medium start, strong compounding when result URLs work
Templates	High	Compounds modestly	Medium-fast start, year-stamp decay without refresh

Two patterns emerge. Opinion and templates dominate early yield. Research dominates later yield once the data starts compounding. Frameworks sit between with the best balance of early lift and durable compounding. The CSY explains why mid-market brands chasing fast traction default to opinion and templates while enterprise brands with patience default to research.

A line chart with citation yield on the vertical axis and time since publish on the horizontal axis, with no fixed durations. Opinion and templates start at high early yield and compound modestly. Research starts at low early yield and compounds the hardest, ending highest. Frameworks start at moderate yield and compound strongly. Calculators start at moderate yield and compound strongly with refreshes. — Figure 3. Citation Surface Yield by asset type. Opinion and templates lead the early window; research compounds hardest later; frameworks and calculators balance the curve. A mix of types spans more of it than one asset can.

The CSY varies by category. Wikidata-dominant categories (per the Wills correlations) lift research's yield because Wikipedia inclusion drives outsized citation share. SE-outbound-link-dominant categories lift template and calculator yield because broad listing surfaces reward utility-driven assets. Apply the CSY as the starting comparison. Adjust by category factor before locking the asset choice.

The CSY also reframes how much to make. Most teams plan one asset and put all of their effort there. The better move is to plan a mix of asset types across the year, because a spread covers more of the Citation Surface Yield curve than one large asset does. Opinion and templates carry the early window, frameworks balance the middle, and research, where the brand can produce it, anchors the durable later yield. One asset can only sit at one point on that curve.

The Decision Framework

The asset-type choice rests on the Citation Surface Yield profile above plus two questions: what you can credibly produce (Expertise Fit) and what the category's cited content lacks (Competitive Landscape). Work both questions before picking. The first question that rules an asset type out is decisive. Most teams reverse this. They start with what feels rigorous or what the team already knows how to make. Working the two questions in the right order yields very different choices.

Expertise Fit

Strong operator experience, weak data infrastructure. Frameworks and opinion.
Strong domain expertise, no research operation. Calculators or templates. Expertise becomes the method; the asset is the interface.
Strong research capability, established analyst coverage in category. Data-driven research.
Strong product or operational expertise, engineering capacity. Calculators that expose the logic of your domain.

Competitive Landscape

Category dominated by one or two named frameworks. Do not build a competing framework. Pick a different asset type.
Category with no proprietary data sources. Data-driven research has high upside. The first credible study becomes the citation default.
Category saturated with low-quality calculators. A calculator with a clear method and ungated result URLs displaces the incumbents.
Category dominated by gated PDFs. Any ungated asset has unusual citation upside.

Conflict resolution: when expertise fit and competitive landscape pull in different directions, the question that rules an asset type out wins. No research capability rules out research even if the landscape favors it. Build what you can credibly produce and what the category genuinely lacks. The two questions are AND, not OR.

A flow diagram. Two question panels feed one gate. The first panel is Expertise Fit, what the team can credibly produce. The second panel is Competitive Landscape, what the category cited content lacks. Both feed an AND gate. An asset type is built only when it passes both questions; the first question that rules a type out is decisive. — Figure 4. The decision framework. Expertise Fit and Competitive Landscape are an AND, not an OR; an asset type is built only when it clears both.

Platform-Specific Considerations

AI systems behave differently. The asset-type choice should reflect which platforms matter most to the brand's audience.

ChatGPT. Favors structured listicles, comparison tables, FAQs, and step-by-step content. AirOps March 2026 data found lists and tables appear in nearly 80% of ChatGPT citations vs 29% in Google's top results. Asset types with list-heavy structure (templates, frameworks with component breakdowns, calculators with comparison views) over-index.
Claude. Weights academic citations and source diversity. Research with a documented method, opinion with a credible author entity, and frameworks with named author attribution all over-index.
Perplexity. Heavy weight on Reddit, community discussion, and recent news. Opinion shared through community surfaces, reactive opinion on recent events, and frameworks talked about in active Reddit threads over-index.
Gemini. Pulls mostly from Google-indexed content with strong SEO signals. Because AI search is an evolution of SEO rather than a separate discipline, assets that earn organic Google rankings also earn Gemini citations; the same crawlable, authoritative work that drives a national SEO footprint feeds both.
Google AI Overviews. 97% of AI Overviews cite at least one source from top 20 organic results (seoClarity, February 2025). Asset types that rank well organically (research with backlinks, calculators with utility, in-depth frameworks with topical authority) over-index.
Microsoft Copilot. Pulls mostly from LinkedIn, Microsoft-indexed enterprise sources, and Bing-ranked content. Opinion shared through LinkedIn over-indexes.

Most teams optimize for a portfolio. The citation spread across platforms is wide enough that single-platform optimization is rare.

Industry Variants

Asset-type winners vary by industry. Ben Wills's March 2026 research (145 industries, 1,595 personas, 105,000+ LLM prompts) surfaced industry-specific signal patterns that guide asset selection.

Wikidata-dominant categories. Accounting software, baby care brands, budget hotel chains, CRM software. Reward research and frameworks because both produce the entity-level claims coded into Wikidata and Wikipedia.
SE-outbound-link-dominant categories. Agricultural equipment, B2B marketing data providers, beauty and cosmetics retail, beer brands, bottled water. Reward templates and calculators that get listed across third-party sites.
Wikipedia-citation-dominant categories. CRM software (rho=0.577). Reward research and frameworks with citation density high enough to merit Wikipedia inclusion.
Harmonic-centrality-dominant categories. Affiliate marketing networks (rho=0.577), auto insurance, brokerage and wealth management apps. Reward research with downloadable data and tools with embed-friendly result URLs.
Backlink-count-dominant categories. Car rental brands. Favor calculators and templates that earn organic backlinks through utility.
Best-search-rank-dominant categories. Most industries fall here at moderate correlation. Research and frameworks that produce ranking-friendly canonical pages are the defaults, which is why building topical authority on the core subject still does the heavy lifting.

For asset-type selection, check your category against the Wills correlations. Find the dominant signal type. Bias your asset choice toward types that produce that signal.

The Selection Workshop

Asset selection is most reliable when the team runs a structured workshop, not when it gets decided in a planning meeting. The workshop is a single focused session with the executive sponsor, the operator who will produce the content, and one or two context stakeholders.

Pre-workshop data to gather ahead of the session.

The Citation Surface Yield profile for the asset types under consideration.
Current AI citation baseline for the category, measured via Profound, Peec AI, or Semrush AI Toolkit.
Competitive map: top three competitors and the assets they get cited for.
Wills industry data for your category from the LLM Ranking Factors research.
Operator-network audit: named experts with the credibility to carry content.
An honest read on what the team can credibly produce, including whether real research infrastructure exists.

The workshop agenda, in sequence.

Review. Review the pre-workshop data. Confirm what the team can credibly produce and what the category's cited content lacks.
Work the framework. Work the decision framework. Rule out asset types the team cannot credibly produce or that the competitive landscape closes off.
Build the mix. Draft each asset (topic, format, author). Multiple candidate mixes are fine at this stage.
Pressure-test. Pressure-test each mix against three questions: does it span the Citation Surface Yield curve rather than cluster at one point; can the team credibly produce every asset in it; does it address what the category's cited content lacks.
Pick. Pick the mix. Write down the decision and the reasoning.
Assign. Assign owners and confirm the near-term execution plan.

Post-workshop deliverables, produced promptly. One-page summary with the asset list and owners. Decision document with the Expertise Fit and Competitive Landscape analysis. Calendar of asset publish dates. Measurement plan per Chapter 13.

Common Selection Mistakes

Mistake	Counter-test
1. "Let's do a survey." The most common wrong default. Research feels rigorous and produces visible deliverables.	Would a mix of opinion, frameworks, and a calculator, built with the same operator effort, produce more citation surface area? If yes, the default is wrong.
2. "Let's build a tool." The second most common wrong default. Engineering-led organizations fall into it most.	Name the specific quantitative query the calculator answers. If the answer is vague, the calculator will struggle to earn citations.
3. One big asset versus a mix. Teams default to one large asset and put all of their effort there, but a mix wins on citation density because it covers more of the curve.	Have you compared the one-big-asset plan to a mixed plan built with the same effort?
4. Building frameworks in saturated categories. Teams keen on thought leadership try to build frameworks where one or two named incumbents already exist.	Name the top one or two frameworks in your category. If you cannot clearly stand apart, pick a different asset type.
5. Choosing based on team comfort rather than audience need. Writers prefer opinion, researchers prefer research, engineers prefer calculators.	If a different team owned this decision, would the answer change?
6. Ignoring industry-specific signal patterns. Most teams pick asset types on general best practice and miss that preferences vary a lot by industry.	Have you checked the Wills industry correlations for your category and weighted your asset mix to match?
7. Skipping the promotion plan. Teams plan for asset development and forget promotion.	For every asset planned, is there matching operator time and distribution work planned for it per Chapter 3 (Third-Party Corroboration)?

The Asset Refresh Cadence Calendar

Published assets decay without refresh. The decay rate varies by asset type. The Asset Refresh Cadence Calendar gives operators a per-type schedule for when to update each asset. Programs running on a single quarterly refresh cycle waste effort on assets that do not need it and skip refreshes for assets that decay fast.

Frameworks: the longest refresh interval. Vocabulary drift is the trigger. Categories evolve their language. Framework terms become stale or misleading. The refresh updates examples, adds new sub-components if the framework's scope expanded, and replaces case studies that no longer reflect current category dynamics.
Opinion: a moderate refresh interval for evergreen pieces. Counter-claim erosion is the trigger. The contrarian view that was sharp early may become consensus later. The refresh either re-sharpens the claim against the new consensus or pivots to a new contrarian angle. Reactive opinion pieces (published promptly after a trigger event) do not refresh. They retire.
Research: an annual full refresh. Data staleness is the trigger. Survey-based research stays fresh only for a limited window before journalists, analysts, and AI systems weight it less. Annual refresh is the working cadence. Sample-sized panel research can refresh on a quarterly micro-release cycle (one new cut per quarter) without redoing the full study.
Calculators: the shortest refresh interval. Benchmark data, pricing assumptions, and reference rates change. The Chime case (citations tripled quickly after refresh program standardized, per the April 2026 public data) shows the lift quarterly refreshes drive. Skip the refresh and the citation curve flattens.
Templates: an annual refresh for year-stamped templates. Year-stamp decay is the trigger. A "Q1 2026 Marketing Plan Template" stops being useful at the end of 2026. The refresh adds the new year, updates field labels, refreshes example data, and bumps the "last updated" date. Templates without year stamps refresh on a longer cycle similar to frameworks.

The calendar drives quarterly planning. At the start of each quarter, audit which assets are due for refresh against the cadence. Schedule the operator and coordinator time. Treat refresh work as part of the Evidence program, not as discretionary work that gets pushed. Refresh velocity (the share of assets refreshed on time per quarter) is itself a tracking metric. Programs at 80%+ refresh velocity see steady citation share. Programs below 60% see citation share decline under consistent under-refresh.

The Asset Retirement Decision Framework

Not every asset deserves a refresh. Some assets should retire. The Asset Retirement Decision Framework uses four signals to distinguish a refresh-and-extend asset from a retire-and-replace asset. Continue refreshing assets that pass the framework. Retire assets that fail. Refreshing a retire-grade asset wastes operator time that should go to new asset development.

Signal 1: Citation share trend. The asset's primary topic citation share dropped 30% or more over a sustained trailing window. The drop reflects category evolution that the asset cannot recover through refresh alone.
Signal 2: Co-citation velocity decay. The Co-Citation Velocity Score per Chapter 3's framework has fallen below 0.5 of the asset's baseline. The asset is no longer being referenced by third parties at the rate that drives compounding.
Signal 3: Competitive displacement. A competitor recently published an asset with materially better data, method, or format. The displacement is visible in AI citation patterns. The competitive asset now occupies the citation slot the original asset used to hold.
Signal 4: Operator credibility gap. The operator can no longer defend the asset's claims due to category evolution. Common cases: the data assumption changed, the framework's components do not match current category dynamics, or the operator's view has matured past the asset's original framing.

The decision logic:

Three or four signals triggered. Retire the asset. Replace it with a new asset using a refreshed approach. Move the URL to a 301 redirect to the replacement if AI systems are still citing the original URL.
Two signals triggered. Run a deep refresh attempt. Add a new section that addresses the strongest of the triggered signals. Re-measure after the refresh has had time to register. If at least one signal flipped to positive, continue maintaining. If not, retire.
One signal triggered. Monitor. Address the specific signal in the next scheduled refresh.
Zero signals triggered. The asset is healthy. Continue the standard refresh cadence.

A scale that maps a count of triggered signals to a decision. Four signals are listed: citation share trend, co-citation velocity decay, competitive displacement, and operator credibility gap. Zero signals means the asset is healthy and stays on cadence. One signal means monitor. Two signals means a deep refresh attempt. Three or four signals means retire and replace. — Figure 5. The Asset Retirement Decision Framework. The count of triggered signals converts a sunk-cost judgment into a data-driven check: zero or one signal keeps the asset, two warrants a deep refresh attempt, three or four retires it.

Most programs over-maintain underperforming assets. The sunk-cost effect is strong with assets the operator built personally. The retirement framework converts the decision into a data-driven check. It frees operator time for new asset development, which has higher expected citation yield than refreshing a stalled asset.

Questions & Answers

How do I choose between a framework and a calculator if the team can produce both? Build the framework first. Frameworks earn citations by becoming vocabulary that travels by reference. Calculators earn citations by becoming utilities that earn URLs. The framework sets the category vocabulary that the calculator then puts into use.

What if our category does not have established asset types? Default to expert opinion and frameworks. Categories without set asset types are categories where vocabulary is missing. Opinion and frameworks both supply vocabulary. Research and calculators need category context that does not yet exist.

Can we mix asset types in one launch? Yes, and this is usually better than betting everything on one large asset. A mix of opinion, frameworks, calculators, and templates covers more of the Citation Surface Yield curve than a single research study, so it produces more citation surface area for the same effort.

Does the right asset type change if we are optimizing for ChatGPT vs Perplexity vs AIO? Yes. ChatGPT favors structured listicles. Perplexity weights Reddit. AIO leans on top-20 organic. Copilot pulls from LinkedIn. Most teams optimize for a portfolio. The citation spread across platforms is wide enough that single-platform optimization is rare.

How long until I see citation traction after publishing my first original asset? Opinion is the fastest to earn traction, followed by templates, then frameworks and calculators, with research the slowest. The variance is mostly a function of co-citation velocity.

Should we gate our research, calculator, or framework? No, with one exception. AI systems cannot fill out forms. Gated content is invisible to AI Search. The exception is enhanced versions of an ungated base asset. Give basic results ungated. Gate downloadable enhanced reports.

What is the smallest viable way to start an Evidence program? Start with the asset types that convert effort into citations most efficiently. A working starter cadence is one opinion piece per month plus one framework or template per quarter, all of which a credible operator can produce without research infrastructure. Consistency over several quarters matters more than the size of any single asset.

Can opinion pieces really compete with research for citations? Yes, when the operator credibility is set and the claim is specific. AirOps's March 2026 analysis found AI systems cite opinion-based content at rates close to empirical research when the source has verified expertise and the claim is specific enough to pull out as a discrete statement.