Data Licensing Boom in India: Why Publishers Are Monetizing Consumer Insights Beyond Traditional Ad Networks
SuprContentIn shortThe data licensing boom in India describes an emerging shift in which digital publishers turn first-party audience information, aggregated behavioral patterns, content-interest signals, regional consumption data, contextual intelligence, and market research into paid business products beyond ordinary advertising inventory.
On this page14 sections
Publishers can package these assets as audience segments, research reports, data feeds, dashboards, analytical products, privacy-controlled matching services, or licensed datasets. The model matters because India's digital audience and advertising markets continue to grow while many publishers receive only a limited share of that value. Data licensing gives news companies, regional publishers, specialist websites, entertainment businesses, and digital content companies another way to earn from the consumer understanding created through direct audience relationships.
The phrase "data licensing boom" needs one qualification. Public research does not yet provide a reliable nationwide figure showing the size of publisher consumer-data licensing revenue in India. The shift is better understood as a developing monetization category supported by first-party data, privacy changes, regional audience depth, advertiser demand for better consumer signals, and publisher pressure to reduce dependence on programmatic advertising.
India's Digital Growth Has Exposed a Publisher Monetization Gap
India has reached enormous digital audience scale, yet audience growth and publisher revenue have not increased at the same rate. Digital advertising has expanded quickly, but much of that spending flows through large search, social, commerce, video, and advertising technology systems rather than directly to independent publishers.
A 2024 study of India's digital news publishing market estimated total digital news revenue at about ₹2,556 crore in 2023. Advertising generated approximately ₹2,345 crore, or 92% of the total, while subscriptions generated about ₹212 crore. Digital news publishers received only 5.7% of the ₹41,469 crore digital advertising market covered by that study. The same research found that 56% of surveyed large digital news operations were profitable.
The wider digital market continued expanding after that measurement period.
India's media and entertainment sector reached ₹2.78 trillion in 2025. Digital media crossed ₹1 trillion in annual revenue for the first time. Digital advertising increased 26% to ₹947 billion, or ₹94,700 crore, and accounted for 63% of total advertising revenue.
The difference between these figures matters. A large digital advertising market does not mean that every publisher receives comparable financial benefits.
Industry reporting in 2025 found many digital news businesses recording relatively modest growth even while national digital advertising expenditure was rising. Individual publisher performance varied widely, showing how traffic scale, direct sales, subscriptions, branded content, programmatic yield, and audience quality can produce very different results.
By August 2026, research covering five major Indian newspaper groups found that non-print businesses were expected to contribute around one-quarter of revenue during the fiscal year, compared with roughly 13% in fiscal 2019. Those non-print businesses include digital operations, events, and out-of-home activities. The finding does not measure data licensing specifically, but it confirms the broader move toward diversified revenue.
The commercial pressure behind consumer insight monetization is therefore straightforward. Publishers need to earn more from the audience relationships, contextual knowledge, regional expertise, and data generated by their own products.
How Publisher Monetization Moved From Ad Inventory to Audience Intelligence
Publisher monetization has moved through several stages. The earliest digital publishing model focused heavily on selling page space. Advertising networks later automated access to that inventory. Programmatic systems introduced real-time auctions, behavioral targeting, audience identifiers, and data-based campaign optimization. The current stage gives publishers stronger incentives to treat their own audience intelligence as a separate commercial asset.
During the early web period, publishers mainly sold banners, sponsorships, and other placements directly to advertisers. Revenue depended heavily on traffic and available inventory.
Programmatic advertising expanded during the 2010s. Publishers gained access to more advertiser demand, but automation also moved pricing and audience decision-making deeper into advertising technology systems.
Third-party cookies and cross-site identifiers supported much of behavioral advertising. Publishers frequently supplied the page, content, and audience while external systems handled identity, targeting, bidding, and measurement.
Privacy regulation, browser restrictions, mobile privacy controls, consent requirements, and growing concern about unauthorized data use began changing those economics. Direct audience relationships became more valuable.
Commercial data businesses now show how first-party publisher signals can support products built from identifiers, device information, user engagement, content consumption, pattern recognition, clustering, recency scoring, and predictive audience classification.
The Indian regulatory timeline adds another important layer.
The Digital Personal Data Protection Act became law in August 2023. The Digital Personal Data Protection Rules were notified in November 2025, with implementation scheduled through a phased transition. The framework makes purpose, permission, transparency, security, and individual rights central considerations when personal data is processed.
By 2026, the commercial discussion had moved beyond pure audience acquisition. Current industry forecasts increasingly describe India's next media phase as one focused on converting scale and engagement into sustainable revenue. An August 2026 outlook projects India's entertainment and media market rising from US$25.7 billion in 2025 to US$36.7 billion by 2030 and specifically identifies monetization as a major next-stage priority.
Data licensing fits directly into that transition.
What Indian Publishers Can Actually Monetize as Data Products
Publisher data licensing can cover several products, but they do not all carry the same privacy, commercial, or technical requirements. The safest and often most scalable products are frequently aggregated insights rather than exports of identifiable consumer records.
Aggregated audience intelligence
Publishers can analyze large groups of users to show how content interests differ across categories, regions, periods, devices, or audience groups.
A report might show rising aggregate attention to electric vehicles, overseas education, residential property, personal finance, cinema, travel, or consumer technology.
Contextual intelligence
Contextual data describes the content environment rather than the identity of the person reading it.
Publishers possess detailed knowledge of article topics, categories, keywords, language, geography, sentiment, content format, and editorial context. This information can support contextual advertising, content research, category analysis, and campaign planning.
Declared preference data
Registered users can explicitly provide interests, language preferences, location choices, newsletter topics, or other preferences.
Declared data can be more interpretable than inferred behavior because the consumer directly supplied the preference. The publisher still needs an appropriate purpose and permission for commercial processing where personal data is involved.
Behavioral engagement data
Page visits, article categories, repeat visits, content depth, video interactions, search activity, newsletter engagement, and other events can describe audience behavior.
Behavioral signals require careful interpretation. Reading an article about home loans does not prove that a person intends to apply for a mortgage.
Intent signals
A collection of repeated, recent, and high-quality interactions can indicate possible commercial interest.
Intent data is most useful when publishers distinguish observed interest from verified purchase intent.
Research and survey products
Publishers with trusted audience communities can conduct surveys or panels and sell aggregated findings to businesses.
Survey products become especially useful when methodology, sample size, dates, respondent criteria, and limitations are disclosed clearly.
Data feeds and APIs
Publishers can provide recurring access to approved trend information through downloadable datasets or application programming interfaces.
Examples include topic demand indexes, category trends, regional interest patterns, content metadata, pricing information, or other non-personal market intelligence.
Audience activation products
A publisher can create defined audience groups and make them available for approved advertising activation without necessarily giving the advertiser a raw user database.
Content licensing
Editorial archives, metadata, structured content, and other intellectual property can also be licensed.
Content licensing and consumer data licensing are not the same business. Ownership of an article does not automatically give a publisher unrestricted rights to commercialize personal information about people who read that article.
Regional and Vernacular Audiences Give Indian Publishers a Distinct Data Advantage
India's linguistic and regional diversity creates consumer intelligence that broad national datasets can struggle to describe with enough detail. Regional publishers often possess direct knowledge of language preferences, local topics, cultural consumption, city-level interests, and state-specific consumer behavior.
A 2025 consumer internet study characterized a large share of India's market as mobile-first and vernacular-first. It also highlighted the commercial importance of regional language content and noted significant differences in digital behavior between higher-income consumers and the broader mass market.
The same research highlighted more than five Indian languages with over 100 million native speakers and described publishing activity across more than 22 languages. It also identified regional expansion as a major strategy across digital content categories.
These conditions give regional publishers several possible data advantages.
A Telugu financial publisher can observe aggregate interest across savings, loans, gold, property, taxation, or local business topics.
A Tamil education publisher can measure changing reader attention across examinations, overseas education, professional courses, employment categories, and scholarships.
A Marathi automobile publisher can compare aggregate interest in motorcycles, electric vehicles, used cars, financing, and specific vehicle categories.
A Hindi agriculture publication can study content demand around crops, machinery, weather, irrigation, commodity markets, and government programmes.
These examples describe possible products, not verified commercial results.
Regional data becomes valuable when publishers preserve distinctions between language, location, topic interest, declared preference, observed behavior, and actual purchasing activity.
A reader who views several health articles should not automatically be classified as having a medical condition.
A reader comparing smartphones is showing product interest, but the activity does not confirm a planned purchase.
Good consumer intelligence preserves those boundaries.
The Data Licensing Process Starts With Rights, Not With Selling
A publisher should begin a data licensing programme by identifying what data exists, how it was collected, what rights apply, and whether the proposed commercial purpose is permitted. Product design comes after that assessment.
A practical workflow includes the following stages.
Create a data inventory
The publisher records its available data sources.
These can include website analytics, application analytics, registrations, subscriptions, newsletters, surveys, content taxonomies, search activity, events, transactions, and editorial metadata.
Classify the information
Each dataset should be classified as personal data, anonymous information, aggregate information, content metadata, commercial data, or another relevant category.
Pseudonymized data should not automatically be treated as anonymous data.
Map collection purpose and permission
The publisher records why information was collected and which permissions apply.
Data collected for account administration should not automatically become available for unrelated commercial licensing.
Assess commercial usefulness
High-volume information is not automatically valuable.
Buyers generally need data that helps answer a business problem, such as category demand, geographic differences, audience overlap, market timing, content preferences, or campaign planning.
Clean and standardize the dataset
Invalid traffic, bots, duplicate records, incomplete events, inconsistent categories, and technical errors can reduce data quality.
Definitions should remain consistent across reporting periods.
Aggregate or minimize information
The product should contain only the information necessary for its defined purpose.
Aggregation can reduce exposure while still supporting market analysis.
Define licensing rights
Contracts should specify permitted use, access period, retention, redistribution restrictions, security requirements, deletion duties, audit rights, and restrictions on re-identification.
Control delivery
Reports, APIs, dashboards, secure file transfers, and controlled analytical environments provide different levels of access.
The delivery method should match the sensitivity of the information.
Monitor usage
Publishers need records showing who accessed licensed information, what product was used, and whether contractual restrictions were followed.
Data licensing therefore operates as a governed information business, not a simple export from an analytics system.
Consumer Insight Products Can Serve More Buyers Than Advertisers
Publisher audience intelligence can have value beyond media buying because many business functions need information about consumer demand, interests, timing, geography, and category movement.
Advertising teams can use aggregated publisher insights to identify relevant audiences and contextual environments.
Brand teams can study changing category interest and regional differences.
Market research teams can use publisher panels or aggregate behavioral information as one input to wider research.
Product teams can examine consumer interest in features, categories, and problems when the underlying information is suitable for that purpose.
Retail and commerce businesses can use approved category-demand signals to understand which topics attract attention before a transaction occurs.
Financial services companies can study aggregate interest in permitted financial topics, subject to strict controls around personal and sensitive information.
Real estate businesses can use broad geographic and property-interest patterns without requiring individual reader identities.
Entertainment companies can examine aggregate demand by genre, language, actor, release format, or content category.
Public affairs and policy researchers can use aggregated information to understand attention to civic topics when the dataset is appropriate and privacy controls are applied.
The commercial advantage comes from specificity.
Generic information that buyers can obtain free of charge has limited licensing value. Proprietary audience access, local context, frequent updates, consistent methodology, rare categories, or high-quality declared preferences can make a publisher's product more distinctive.
Data Quality Determines Whether Consumer Insights Have Commercial Value
A consumer dataset becomes useful when buyers can understand what each variable means, how frequently information is updated, how representative the audience is, and what limitations affect interpretation.
Publishers should measure data products with metrics that describe both commercial performance and information quality.
Coverage measures the number of valid users, events, categories, regions, or records represented.
Freshness measures how recently the underlying activity occurred.
Completeness shows how much of the expected data is present.
Consistency measures whether definitions and collection methods remain comparable across reporting periods.
Match rate can describe how successfully permitted identifiers are connected within approved matching processes.
Segment size describes the number of records or users qualifying for an audience category.
Renewal rate measures how many licensing customers continue paying after the initial contract.
Usage rate shows whether customers actually access reports, dashboards, feeds, or APIs.
Revenue per data customer helps publishers evaluate the commercial contribution of licensing relationships.
Data processing cost includes collection, storage, engineering, analytics, security, governance, and delivery expenses.
Publishers also need to document sampling bias.
A publisher's audience is not automatically representative of India's population.
A technology website may overrepresent people interested in devices and software. A financial publication may overrepresent users interested in markets. A regional entertainment website may contain strong language and genre preferences.
Those characteristics can make a dataset valuable, but only when buyers understand what population the information represents.
Pricing Data Requires a Different Model From Pricing Advertising Inventory
Advertising inventory is commonly priced around impressions, clicks, actions, or campaign outcomes. Data products are more often priced according to access, exclusivity, update frequency, coverage, analytical depth, permitted uses, and delivery method.
A recurring consumer trend report can use a subscription model.
A continuously updated API can use access tiers based on volume or frequency.
A custom research dataset can use project pricing.
A privacy-controlled audience activation agreement can use licensing or usage fees.
An exclusive dataset can command different economics from information licensed to multiple customers.
Publishers should calculate the full cost of producing the product before setting a price.
Engineering, data storage, analytics, legal review, consent management, security, account management, product development, quality control, and customer support all affect margin.
Revenue alone is therefore a weak measure of success.
A licensing product generating significant sales but requiring expensive custom work for every customer can have poor economics.
The stronger model is repeatable. The same governed dataset or analytical process should serve multiple approved customer needs without rebuilding the product each time.
India's DPDP Framework Changes the Economics of Consumer Data Licensing
India's Digital Personal Data Protection framework means publishers cannot treat identifiable audience information as unrestricted commercial inventory. Publishers processing digital personal data need to consider lawful purpose, consent or other permitted grounds, notice, security, individual rights, retention, and accountability.
The Digital Personal Data Protection Act states that personal data can be processed for a lawful purpose when the individual has provided consent or when the processing falls within specified legitimate uses. A consent request must be accompanied or preceded by notice explaining the personal data involved and the purpose of processing.
The Digital Personal Data Protection Rules were notified in November 2025. The implementation model provides an 18-month phased compliance period. Government guidance also states that consent notices should be clear and separate and should explain the specific purpose for which personal data is collected and used.
The framework gives individuals rights relating to access, correction, updating, erasure, and nomination. Data fiduciaries also face security and breach-related responsibilities.
For publishers, the commercial lesson is significant.
A privacy notice stating that data is used to provide a website service does not automatically create unrestricted permission for every future licensing purpose.
Publishers need to know whether a proposed dataset contains personal data, whether the intended processing is covered by the applicable purpose, what disclosures were made, which third parties receive information, and how withdrawal or deletion requests affect downstream processing.
Children's data requires additional care.
Sensitive subject areas also deserve conservative treatment even when publishers believe a technical identifier is pseudonymous.
Legal compliance should be assessed for the specific data product and use case. Data licensing agreements cannot substitute for obligations imposed by applicable law.
Anonymous, Aggregate, and Pseudonymous Data Must Not Be Confused
The distinction between anonymous, aggregated, and pseudonymous information affects both privacy risk and product design.
Anonymous data is information that cannot reasonably be connected to an identifiable individual under the applicable circumstances.
Aggregated data combines information across groups. Aggregation can reduce individual exposure, but small groups or unusually detailed combinations can still create identification risks.
Pseudonymous data replaces direct identifiers with another value, such as a code or token. Pseudonymization can reduce risk, but the information can remain personal data when re-identification is possible using additional information.
Identifiable personal data relates to an individual who is identifiable by or in relation to that information.
The practical mistake is assuming that deleting a name or email address automatically makes information anonymous.
A dataset containing precise behavior, device attributes, account history, location information, and persistent identifiers can still relate closely to a person.
Publishers building licensing products should therefore apply data minimization at product level.
If an advertiser only needs an aggregate regional demand index, individual-level browsing histories are unnecessary.
If a research customer needs category trends, the publisher can deliver category trends rather than exposing raw event records.
Technology Is Turning Publisher Data Into Repeatable Products
A scalable publisher data business needs technology that can collect, classify, govern, analyze, and distribute information consistently. The exact system varies by publisher size and product complexity.
A customer data platform can unify approved first-party audience events.
A consent system can record permissions and preference changes.
A content taxonomy can connect articles and videos with consistent subject classifications.
Analytics systems can calculate engagement and category trends.
Data warehouses can store structured events and analytical outputs.
Identity systems can connect permitted identifiers across approved interactions.
Machine learning models can group patterns or estimate audience propensities when the methodology and purpose are appropriate.
APIs can distribute recurring information products.
Data clean rooms can support controlled matching or analysis between parties while limiting direct exposure of underlying records.
Access control systems can restrict who is permitted to view a dataset.
Audit logs can record access and changes.
The technology does not determine whether licensing is lawful or commercially useful.
Good governance, useful data, clear customer demand, and valid rights remain necessary regardless of the software used.
The Best Publisher Data Products Sell Answers, Not Raw Records
The long-term commercial opportunity for publishers lies in converting data into understandable business intelligence. A buyer usually values an answer to a business problem more than millions of unorganized events.
Consider the difference between two products.
One product delivers a file containing millions of page-view records.
Another shows how consumer attention to electric two-wheelers changed across selected Indian states, which content categories drove that interest, how recently the activity occurred, which language groups were represented, and what limitations apply to the dataset.
The second product is easier for a buyer to understand and use.
Publishers can therefore move through a value chain:
- Raw events become standardized data.
- Standardized data becomes classified audience information.
- Classified information becomes aggregated patterns.
- Patterns become consumer insights.
- Consumer insights become recurring commercial products.
- Recurring products become a separate revenue stream.
The value comes from interpretation, context, quality, frequency, and rights.
This approach also reduces pressure to sell excessive user-level information.
Publishers can retain greater control over their intellectual and audience assets by selling defined outputs rather than unrestricted copies of underlying databases.
Data Licensing Will Sit Beside Advertising, Subscriptions, Commerce, and Events
Data licensing is unlikely to replace advertising for most Indian publishers. The more realistic future is a diversified revenue portfolio where different audience assets support different products.
Advertising monetizes attention.
Subscriptions monetize willingness to pay for access or utility.
Events monetize community and participation.
Commerce and affiliate models monetize transactions or referrals.
Syndication monetizes editorial content rights.
Research products monetize knowledge.
Data licensing monetizes permitted information and analytical value.
These revenue types can reinforce each other when managed carefully.
Registration can improve subscription experiences while producing declared audience preferences.
Editorial taxonomies can improve content discovery while supporting contextual intelligence products.
Research surveys can create editorial material while producing aggregate market datasets.
Direct advertiser relationships can identify demand for new audience products.
The direction of the Indian media sector supports greater diversification. Digital media became India's largest media and entertainment segment in 2024 at ₹802 billion. Advertising reached ₹1.28 trillion that year, with new media accounting for 56% of total advertising.
By 2025, digital media had crossed ₹1 trillion and digital advertising reached ₹947 billion.
The opportunity is therefore not a shortage of digital activity.
The business challenge is converting that activity into publisher-controlled revenue.
What Will Define the Next Phase of Publisher Data Monetization in India
The next phase of Indian publisher monetization is likely to focus on controlled first-party relationships, regional intelligence, direct commercial products, contextual data, recurring research, and privacy-aware audience services.
Several forces support that direction.
Digital advertising continues expanding.
Regional content remains central to India's internet market.
Publishers need greater revenue diversity.
Advertisers want measurable audience information.
Consumer privacy requirements are becoming more explicit.
Data engineering and analytical tools are becoming more accessible.
Businesses increasingly expect recurring insights rather than static media plans.
The strongest publishers will not necessarily be those with the largest raw datasets.
Commercial value depends on whether a publisher possesses information that is difficult to reproduce, current enough to matter, permitted for the intended use, clearly defined, and useful for a buyer's decision.
A specialist publisher with a smaller but highly relevant audience can therefore possess more valuable intelligence for a particular category than a much larger general-interest website.
Indian publishers also have an opportunity to build products around regional demand patterns that global or national datasets describe poorly.
The boundaries will matter as much as the opportunities.
Publishers that overclassify users, infer sensitive characteristics from weak signals, retain unnecessary records, or license information for purposes outside user expectations create legal and commercial risk.
Publishers that focus on aggregate intelligence, declared preferences, clear purposes, limited access, documented methodology, and useful analytical outputs can create a more durable model.
India has already moved from limited internet publishing to mass mobile consumption, programmatic advertising, subscription products, regional digital services, first-party audience strategies, and privacy regulation.
The 2026 market is entering a further stage in which audience knowledge itself can become a product.
Data licensing is one expression of that shift.
The opportunity for Indian publishers is not simply selling more data. It is building trusted, repeatable information products from consumer relationships they already understand better than many external advertising systems.
India’s publisher data licensing opportunity is growing from a broader shift toward first-party data, regional audience intelligence, privacy-aware analytics, and revenue diversification. Digital advertising continues to expand, but independent publishers do not always capture a proportionate share of that growth. Consumer insight products give publishers another way to earn from the knowledge created through direct audience relationships.
The strongest opportunities are likely to come from aggregated audience intelligence, contextual data, regional and language-specific trends, declared preferences, research products, APIs, and controlled audience activation. Commercial value depends on data quality, freshness, clear methodology, lawful processing, and whether the information helps buyers make better decisions.
India’s Digital Personal Data Protection framework also changes how publishers can commercialize audience information. Personal data cannot be treated as unrestricted inventory. Publishers need clear purposes, appropriate permissions, strong security, controlled access, and careful separation between anonymous, aggregated, pseudonymous, and identifiable information.
The future of publisher monetization will therefore be broader than advertising alone. Advertising, subscriptions, research, events, commerce, content licensing, and data products can work as separate but connected revenue streams. For Indian publishers, the larger opportunity is not selling more raw consumer records. It is creating trusted, repeatable, privacy-conscious intelligence products that turn audience knowledge into sustainable business value.
What Is Data Licensing for Indian Publishers?
Data licensing is a revenue model in which publishers provide approved access to aggregated audience insights, contextual data, research findings, audience segments, data feeds, or analytical products under defined commercial terms. It allows publishers to earn from audience intelligence beyond traditional advertising inventory.
Why Are Indian Publishers Looking Beyond Traditional Ad Networks?
Indian publishers are seeking additional revenue because digital advertising growth does not always translate into proportional publisher income. Programmatic advertising depends heavily on traffic, auction pricing, advertiser demand, and external advertising technology systems. Data products give publishers another way to monetize direct audience relationships and market knowledge.
What Types of Data Can Publishers Monetize?
Publishers can create products from aggregated audience trends, contextual information, declared preferences, behavioral engagement patterns, research surveys, content metadata, intent signals, and approved audience segments. Personal data requires stronger legal and privacy controls than anonymous or aggregated information.
Why Is First-Party Data Important for Publishers?
First-party data comes directly from a publisher's websites, apps, newsletters, subscriptions, surveys, registrations, and other audience interactions. It can help publishers understand content interests, regional behavior, engagement patterns, and declared preferences without relying entirely on third-party advertising identifiers.
How Can Regional Publishers Benefit From Data Licensing?
Regional publishers can develop intelligence around language, geography, local interests, category demand, and consumer behavior. Telugu, Tamil, Hindi, Bengali, Marathi, Malayalam, Kannada, and other language publishers can offer insights that broad national datasets may not describe with the same level of regional detail.
What Is the Difference Between Audience Data and Consumer Intent Data?
Audience data describes characteristics or behavior associated with a group of readers. Consumer intent data attempts to identify signals suggesting interest in a product, service, or category. Reading related content can indicate interest, but it should not automatically be treated as proof of a planned purchase.
How Do Publishers Turn Audience Data Into Commercial Products?
Publishers usually begin by identifying available data, classifying its sensitivity, checking permissions, cleaning and standardizing the information, creating aggregated insights, defining licensing rights, and delivering the product through reports, dashboards, APIs, secure files, or controlled analytical environments.
How Does India's Digital Personal Data Protection Framework Affect Data Licensing?
India's Digital Personal Data Protection framework places responsibilities on organizations processing digital personal data. Publishers need to consider purpose, consent or other permitted grounds, notices, security, retention, user rights, and access controls before using personal data for commercial licensing.
What Is the Difference Between Anonymous, Aggregated, and Pseudonymous Data?
Anonymous data cannot reasonably be connected to an identifiable individual. Aggregated data combines information across groups, although very small or highly detailed groups can still create privacy risks. Pseudonymous data replaces direct identifiers with codes or tokens, but it can still remain personal data when re-identification is possible.
Will Data Licensing Replace Advertising Revenue for Indian Publishers?
Data licensing is more likely to become an additional revenue stream than a replacement for advertising. Publishers can combine advertising, subscriptions, research, events, commerce, content licensing, and data products. The strongest data businesses will focus on useful, repeatable, privacy-conscious intelligence rather than simply selling raw consumer records.