How AI Citation Persistence and Portability Differ Across Finance, Software, Health, Legal, Home, and Consumer Markets

Answer Capsule

AI citation persistence and portability do not move together the same way across major commercial market families.

In the July through September 2026 benchmark, Finance & Insurance had the highest median Top-100 portability of the six groups compared here at 26.58%, while Business Software & Operations had the highest median persistence at 69.49%. Software also had the largest median persistence-portability gap at 46.63 percentage points. Legal & Compliance sat at the other end of both axes, with 46.52% median persistence and 16.96% median portability.

The differences become more useful when the medians are unpacked. 64.10% of Finance markets and 64.29% of Home markets reached at least 25% portability, but 8.97% of Finance markets reached 30% while none of the Home markets did. Consumer markets combined relatively high persistence with lower cross-platform transfer. Health had a wide persistence distribution and no market in the group reached 30% portability. Legal was the lowest-profile group overall, yet Immigration Law Firms reached 39.86% portability, higher than every six-group median.

Platform structure also differs by market family. Google AI Overviews and Google AI Mode were the highest-overlap platform pair in every group, but their median Top-100 overlap ranged from only 37.50% in Legal to 57.49% in Software. Microsoft Copilot was the least-connected platform in most complete markets across all six groups, but the rate ranged from 50% of complete Legal markets to more than 90% of complete Finance markets.

The practical conclusion is not that each industry needs an unrelated GEO playbook. It is that source durability, cross-platform transfer, platform-specific source maps, and benchmark expectations should be calibrated to the actual market family and then verified at the individual market level.

This article develops the industry pillar, There Is No Universal GEO Strategy: How AI Citation Persistence and Portability Differ by Industry Across 283 Markets, and should be read with Why Cross-Industry AI Citation Benchmarks Can Mislead CMOs.

The Six Commercial Market Families Do Not Have the Same Citation Profile

Questions This Section Answers

1. Which commercial market family has the highest AI citation persistence and which has the highest cross-platform portability?
2. How different are Finance, Software, Health, Legal, Home, and Consumer citation-source profiles?
3. Does the industry with the highest persistence also have the highest portability?

The six rows below are deliberately commercial labels rather than a claim that the taxonomy is perfectly symmetrical. Finance, Health, Legal, Home, and Consumer use preserved macro-sector groups. Software uses the narrower preserved Business Software & Operations detailed sector, because the broader Technology & Business macro sector also contains cybersecurity, professional services, recruiting, marketing, communications, and other non-software markets.

Three comparisons matter immediately.

First, Software leads on persistence, not portability. Its 69.49% persistence median is 4.20 points above Finance, but its 23.46% portability median is 3.13 points lower. A software source set was therefore relatively durable across the three observed months without being the most transferable across the six AI surfaces.

Second, Finance leads on portability without leading on persistence. That is an important distinction for planning. A CMO looking only at source stability could miss that Finance has the broadest cross-platform source overlap among these six commercial groupings.

Third, Legal's smaller median gap does not mean it has the easiest cross-platform source environment. Legal has the lowest median persistence and the lowest median portability. Its 30.35-point gap is smaller largely because both axes are lower.

The gap remains a difference between two measures, not a standalone performance score.

The Distribution Bands Matter More Than the Median Alone

Questions This Section Answers

1. How much do persistence and portability vary inside Finance, Software, Health, Legal, Home, and Consumer markets?
2. Which industry groups contain the widest range of AI citation-source behavior?

A sector median can conceal whether most markets cluster around the center or whether the sector contains very different source environments.

Health has the widest middle-50% persistence band among the six groups. Its interquartile range is about 16.01 points, from 53.85% to 69.85%. That means a single Health median compresses materially different source environments.

Home is much tighter on persistence. Its middle 50% spans only 61.29% to 66.67%. Yet Home's portability range still includes meaningful market differences, from Home Security Systems at 17.65% to Long Distance Moving Carriers at 29.87%.

Legal's middle 50% is low and compressed, but its total range is not. Immigration Law Firms reached 39.86% portability, while DUI Lawyers were at 10.50%. The legal-sector median of 16.96% describes the center of the group, not a ceiling.

Finance also contains strong internal spread. Business Checking Accounts reached 34.23% portability, while Debt Collection Agencies were at 12.36%. Long-Term Care Insurance reached 85.19% persistence, while Home Equity Loans were the study's exceptional negative-gap market at 18.34% persistence and 22.70% portability.

These tails are exactly why the market remains the primary unit of execution even when industry context is useful.

Finance and Home Both Reach 25% Portability Often, but Their Upper Tails Differ

Questions This Section Answers

1. Which industries most often exceed 25% or 30% AI citation portability?
2. How often do markets in each industry reach 70% persistence?

Threshold frequencies make the six groups easier to interpret than medians alone. The thresholds below are descriptive cut points, not recommended KPI targets.

Two industries can look similar at one threshold and very different at another.

Finance and Home are the clearest example. Almost the same share of markets reaches at least 25% portability: 64.10% in Finance and 64.29% in Home. But Finance has a meaningful upper tail above 30%, while Home has none.

Software shows the opposite shape. It has the highest share of markets at or above 70% persistence, 35.71%, but only 42.86% reach 25% portability. That is consistent with the group's high persistence and large persistence-portability gap.

Consumer has a strong persistence upper tail. Nearly one-third of Consumer markets reach 70% persistence, but only about one-fifth reach 25% portability and none reaches 30%.

Health has one-quarter of markets at 70% persistence but very limited portability above 25%. Legal has no markets at 70% persistence and only three at 25% portability, although its outlier Immigration Law Firms reaches 39.86%.

The useful question is therefore not simply, "Is this industry high or low?" It is, "Which axis is high, how broad is the distribution, and where does the client's market sit inside it?"

The Highest-Overlap Platform Pair Is the Same Across All Six Groups, but the Strength Is Not

Questions This Section Answers

1. Do the same AI platforms share the most citation sources in Finance, Software, Health, Legal, Home, and Consumer markets?
2. Which industries have the strongest and weakest Google AI Overviews versus Google AI Mode source overlap?
3. Which platform relationships diverge most by industry?

Across every commercial group in this comparison, Google AI Overviews and Google AI Mode were the highest-overlap Top-100 platform pair by group median. But the strength of that relationship varied substantially.

The Google sibling relationship is strongest in Software and Finance, but dramatically weaker in Legal. A 37.50% median Legal overlap between Google AI Overviews and AI Mode is almost 20 points below Software's 57.49%.

That matters because "Google AI" is not equally consolidated across industries. The corpus-wide finding that AI Overviews and AI Mode are close siblings remains true, but the amount of shared source structure changes by market family.

The weakest pair is also highly consistent. Gemini and Microsoft Copilot are the lowest-overlap median pair in five of the six groups. Home is the exception, where Google AI Overviews and Microsoft Copilot are slightly lower.

Software produces the widest median closest-to-furthest pair spread at 40.87 points. Legal produces the narrowest at 27.61 points, but again this does not mean Legal is more portable. The whole Legal overlap structure is lower.

For a deeper cross-platform treatment, see Six AI Engines, Six Different Webs.

Platform Connectedness Produces a Different Fingerprint in Each Market Family

The pair table can be summarized another way: for every market, calculate each platform's average overlap with the other five platforms, then take the group median.

This reveals two broad patterns.

First, the two Google search surfaces are the most connected surfaces in every group. They usually alternate between first and second place.

Second, Microsoft Copilot has the lowest median connectedness in all six groups. But the frequency of Copilot being the uniquely least-connected platform varies materially.

Among markets with all six Top-100 platform portfolios available, Copilot was the sole least-connected platform in:

• 67 of 73 Finance markets, 91.78%
• 12 of 14 Software markets, 85.71%
• 33 of 42 Health markets, 78.57%
• 8 of 16 Legal markets, 50.00%
• 10 of 13 Home markets, 76.92%
• 34 of 43 Consumer markets, 79.07%

The Legal result is useful because it shows that Copilot's isolation is not equally dominant everywhere. Legal is already a low-overlap ecosystem generally. Gemini was the uniquely least-connected platform in six of the 16 complete Legal markets, nearly as often as Copilot's eight.

This is another reason not to transfer a platform hierarchy from one industry into another without checking the local data.

The Middle-Tier Platform Relationships Reorder by Industry

The Google sibling pair is consistently first, but the rest of the platform hierarchy is not identical across industries. Selected Top-100 pair medians show the differences.

Software is a useful exception to a simple global ordering. In Software, ChatGPT and Copilot overlap at 26.58%, above ChatGPT and Perplexity at 24.61%. In the other five groups, ChatGPT and Perplexity are equal or more overlapping than ChatGPT and Copilot.

Legal compresses almost every relationship downward. ChatGPT and Perplexity are only 18.35%, Gemini and Copilot only 9.89%, and even Gemini's relationship with Google AI Mode is just 16.63%. The weak Legal median portability is therefore not coming from one unusually divergent pair. The whole network is comparatively fragmented.

Finance shows a different structure. Several non-Google-sibling relationships cluster near or above 30%, including Gemini with both Google search surfaces and Perplexity with Google AI Mode. That broader middle tier helps explain why Finance has the highest median portability of the six groups even though no pair approaches universal source agreement.

This kind of pair-level structure matters operationally. A single portability median can tell a CMO how fragmented a market family is on average, but it cannot identify which platform pair is most likely to share a source strategy.

Top-25 Sources Are More Persistent and More Portable, but Industry Differences Remain

Questions This Section Answers

1. Do industry differences disappear when analysis is restricted to the most prominent Top-25 citation sources?
2. Which industries show the strongest durable citation core at Top 25?

The broad persistence-portability pattern survives when the portfolio is narrowed from Top 100 to Top 25.

The most prominent source layer is more stable and more transferable in every group. But the ordering does not collapse into one common benchmark.

Legal gains nearly 14.77 points of persistence when moving from Top 100 to Top 25, but its Top-25 persistence still remains below every other group. Its Top-25 portability rises to 20.50%, also the lowest of the six.

Finance retains the highest Top-25 portability at 31.58%. Home follows at 29.89%. Software reaches 28.21%, Consumer 26.60%, Health 24.24%, and Legal 20.50%.

The Top-25 analysis therefore supports a durable-core interpretation without erasing industry structure. A CMO can reasonably care more about prominent sources, but still should not use one universal source-transfer assumption across categories.

Six Market Examples Show Why Sector Medians Need Local Context

Questions This Section Answers

1. What do individual Finance, Software, Health, Legal, Home, and Consumer markets look like relative to their group medians?
2. Which markets combine high persistence with weak portability?

A useful industry article should not end at sector statistics. The underlying markets show why those statistics require local interpretation.

Final Expense Insurance demonstrates the central persistence-portability distinction. A highly recurring source set does not imply the same sources dominate across platforms.

Natural Skincare Brand shows a similar shape in Health, while Outdoor Apparel and Technical Outfits pushes the pattern even further in Consumer markets.

Immigration Law Firms is the opposite kind of example. Its 39.86% portability is not just high for Legal. It is above every six-group median in this article. That makes it dangerous to speak of "legal portability" as though every legal market shares the same source structure.

The correct use of the group data is to form a prior expectation, then replace that expectation with market-specific evidence as soon as the market is measured.

What the Six Profiles Mean for GEO Planning

The evidence supports a layered operating model rather than six completely separate programs.

Finance: expect relatively higher cross-platform reuse than the other major commercial groups shown here, but still enough platform divergence that publisher and source analysis should remain platform-aware. Finance's median portability of 26.58% is the highest of the six, not close to universal overlap.

Software: expect strong month-to-month source stability and a relatively durable prominent core. Software's median persistence is the highest in this comparison, while platform transfer remains moderate. Source maintenance and recurring authoritative coverage may therefore be especially important to measure separately from platform expansion.

Health: expect greater market-to-market variation. Behavioral treatment, clinical services, and consumer wellness do not share one narrow source profile. A health-wide benchmark is best treated as context before moving to the actual subcategory.

Legal: expect lower source overlap on both time and platform axes in the tracked population. But do not turn that into a universal legal rule. Immigration Law Firms demonstrates that individual legal markets can behave very differently.

Home: expect a relatively compact persistence distribution and comparatively strong portability around the 25% level, but limited upper-tail portability above 30% in this sample.

Consumer: expect relatively strong persistence with lower portability. The group includes retail, pet, outdoor, cycling, and fitness markets, so local category structure remains important.

These are measurement expectations, not causal claims about why industries behave differently and not prescriptions for spend allocation. The study does not establish that regulation, buyer behavior, platform architecture, publisher economics, or any other proposed mechanism causes the observed differences.

For benchmark selection, see Why Cross-Industry AI Citation Benchmarks Can Mislead CMOs. For KPI design, see How CMOs Should Set AI Search KPIs by Industry Instead of Using One Global Benchmark.

A Practical Six-Group Comparison Framework

For an executive report, keep the comparison sequence simple:

1. Start with the market. Show the client's actual persistence, portability, platform pair structure, source depth, and recommendation metrics.
2. Add the detailed sector where available. This is especially important for Software and for heterogeneous areas such as Health and Finance.
3. Use the commercial market family as context. The six profiles in this article are useful for directional expectations, not performance targets.
4. Use the overall 283-market benchmark last. It is the broadest context, not the first comparison.
5. Keep platform relationships visible. A sector's portability median can hide whether one relationship, such as Google AI Overviews versus AI Mode, is much stronger than another.
6. Separate citation authority from recommendation authority. A stable publisher or domain source set does not prove that a company is recommended more often.

This hierarchy reduces two common errors: overgeneralizing from the full dataset and overgeneralizing from the sector median.

Methodology and Limits

The research uses the frozen July through September 2026 cohort of 283 qualifying tracked markets, 876,605 retained responses, and 8,718,629 citation occurrences resolved to registrable domains. The six surfaces are ChatGPT, Gemini, Perplexity, Microsoft Copilot, Google AI Overviews, and Google AI Mode. API copies and experimental surfaces are excluded.

Portfolio ranks count each domain once per retained response, sort by response presence, and break ties alphabetically. Eligible Top-100 comparison cells require at least 100 responses and 100 resolved domains. Thin or missing cells are excluded rather than assigned zero. Month portfolios pool platforms; platform portfolios pool completed months. Persistence and portability use Jaccard intersection divided by union, then the median eligible pair value within each market.

The six commercial groups are not a mutually exclusive taxonomy layer. Finance, Health, Legal, Home, and Consumer are preserved macro sectors. Software is the preserved Business Software & Operations detailed sector within Technology & Business. That narrower software definition is intentional and should not be interpreted as representing every Technology & Business market.

Top-25, Top-50, and Top-100 sensitivity calculations use the same market-level median pair method at each cutoff. Platform-pair fingerprints use Top-100 platform portfolios. Platform connectedness is the within-market mean overlap of a platform with the other five surfaces, then summarized across markets. Counts describing the most- or least-connected platform require all six platform portfolios for that market, which produces smaller complete-market denominators than the full sector counts.

This is observational evidence from changing samples, not a certified identical-prompt panel. Of the retained responses, 104,071 used ingestion-date proxies. Only 246 markets had all three months and all six platforms eligible in the marginal portfolios. Shared source ecosystems, retrospective taxonomy, unequal sector sizes, and candidate cross-job repetitions limit population inference.

Domain presence does not verify page quality, ownership, recommendation influence, traffic, conversions, or revenue. Higher persistence does not prove a source is better. Higher portability does not prove a strategy will transfer without modification. Sector membership does not establish a causal mechanism.

Related Research

• There Is No Universal GEO Strategy: How AI Citation Persistence and Portability Differ by Industry Across 283 Markets
• Why Cross-Industry AI Citation Benchmarks Can Mislead CMOs
• How CMOs Should Set AI Search KPIs by Industry Instead of Using One Global Benchmark
• How Publishers Can Choose Adjacent Verticals Based on Citation Stability and Platform Portability
• The Persistence-Portability Gap at Scale: 282 of 283 Markets Show Higher AI Citation Persistence Than Cross-Platform Portability Across 8.72 Million Citation Occurrences
• Six AI Engines, Six Different Webs: Comparing Citation Source Overlap Across ChatGPT, Gemini, Perplexity, Copilot, Google AI Overviews, and Google AI Mode
• Can One GEO Strategy Work Across ChatGPT, Gemini, Perplexity, Copilot, Google AI Overviews, and Google AI Mode?

The earlier Persistence-Portability framework provides conceptual context. The 2026 AI Citation Authority Study covers an earlier high-stakes sample and is a separate population.