Summary
- Affinity Solutions sells an implementation-support and service-continuity account around consumer purchase intelligence, campaign measurement, purchase-based audiences, card-linked engagement and bank campaign services.
- The retention asset is support memory: knowledge of how bank-direct transaction feeds, merchant matching, privacy rules, client reporting, media exposure files and reward logic were made to work for a specific customer.
- The strongest public evidence is official and regulatory: Affinity's own pages describe bank-direct purchase data, daily feeds, clean-room, API and managed-service access, card-linked offer mechanics, financial-institution privacy boundaries, California data-broker registration and security attestations.
- Public evidence does not prove the decisive private facts. The missing proof remains grouped as economics, reliability and retention: contract values, gross margin and cost-to-serve; data-feed uptime, matching accuracy and support response; renewal rates, expansion after failed measurement tests and customer concentration.
- Network and domain records show a bounded digital operating surface, including corporate domains, AWS nameservers, CDN use and third-party hosting clues, but those records do not prove data scale, revenue, customer count or service quality.
The renewal failure that would settle the case
The private metric that would prove or disprove the Affinity Solutions thesis is the share of banks, brands and media customers that renew after a measurement dispute, a transaction-feed interruption, a privacy review, a merchant-tagging error, or a campaign result that changes budget allocation. A clean first sale is not the commercial test. The test arrives when the customer asks whether the supplier can remember enough about the prior integration to restore confidence quickly: which card data was available, which merchant names were grouped, which cohorts were allowed, which exposure file was matched, which privacy rule controlled the output, and which business question the report was meant to answer.
That moment matters because purchase-intelligence products can look simple when reduced to "transaction data." A marketer wants to know whether a campaign caused sales. A retailer wants to know why share shifted. A bank wants a cardholder offer to increase spend without damaging trust. A consulting firm wants a current view of consumer behavior. Yet the value of the answer depends on a long chain of practical decisions: data rights, feed recency, merchant enrichment, identity matching, cohort minimums, attribution logic, dashboard configuration, report delivery, support response and client education. The customer is not only buying a dataset. It is buying continuity in how that dataset is made usable.
By the third paragraph the paid unit is clear. Affinity Solutions sells an implementation-support and service-continuity account: a maintained arrangement that lets a customer plan, target, measure or reward activity using purchase behavior without rebuilding the transaction-data stack alone. The cheaper substitutes are a larger integrator, an in-house data team, a generic software-as-a-service platform, a regional or niche competitor, or delayed automation that leaves the customer using surveys, clicks, panel data or internal reports. The cost driver is support labor around bank relationships, privacy controls, data cleansing, merchant mapping, attribution, clean-room access, API support, managed reporting and customer-specific exceptions. The strongest public evidence is official company and regulatory evidence, followed by named partnership and product evidence, then domain and network-resource evidence. The missing proof must stay grouped as economics, reliability and retention: economics means contract values, margins, partner economics and cost-to-serve; reliability means data-feed continuity, matching accuracy, support response and privacy operations; retention means renewal, expansion, churn, concentration and whether customers stay after the first hard incident.
That framing also prevents overclaiming. Affinity's public website gives unusually concrete data-scale claims and product mechanics, but it does not publish revenue, customer contracts, bank-partner volumes, gross margin, service-level performance or renewal cohorts. The absence of those facts is not a minor caveat. It is the central commercial issue. A company like this earns power if it is difficult for customers to replace the working memory of how a purchase-data deployment was implemented. It loses power if customers can move the same question into their own warehouse, a large cloud marketplace, an advertising platform, a payment-network product, or a cheaper analytics supplier without meaningful disruption.
The public case is therefore a qualified retention thesis. Affinity Solutions appears to control a valuable purchase-intelligence operating surface. Its website says it sees deterministic transaction data from more than 100 million consumers in real time and helps brands and banks target, prove impact and optimize decisions: https://www.affinitysolutions.com/. Its data page says the dataset covers more than 150 million credit and debit cards, more than 100 million consumers, more than 86 billion transactions, more than $4 trillion of spend and more than 5,300 tracked brands: https://www.affinity.solutions/our-data/. Those are strong identity and product claims. They are not proof of unit economics. The question is whether customers pay repeatedly because Affinity remembers the integration well enough to make the next decision safer than the substitute.
Identity and evidence order
Affinity Solutions presents itself as a New York and San Jose company focused on consumer purchase insights. Its homepage describes purchase-based measurement, audience targeting, card-linked engagement and bank campaign services, and the footer identifies the company as Affinity Solutions, Inc. with New York and San Jose locations: https://www.affinitysolutions.com/. The about page names Jonathan Silver as chief executive and founder, lists senior leadership across operations, technology, commercial, product and growth roles, and says the company has positions in technology, analytics, client services and marketing, with offices in NYC and Silicon Valley plus field-based roles across the United States: https://www.affinity.solutions/about-us/.
The official product surface is broad but coherent. Consumer Purchase Insights is positioned as a way to understand spending behavior across brands and categories: https://www.affinity.solutions/consumer-purchase-insights/. Consumer Purchase Lift measures campaign outcomes against real purchases: https://www.affinity.solutions/consumer-purchase-lift/. Consumer Purchase Audiences builds predictive and custom audiences from purchase history: https://www.affinity.solutions/consumer-purchase-audiences/. Card-Linked Engagement supports bank-controlled cardholder offers: https://www.affinity.solutions/card-linked-engagement/. Consumer Bank Campaigns supports performance-based offers promoted to bank card customers: https://www.affinity.solutions/consumer-bank-campaigns/. This is not a single dashboard sold under many labels. It is a set of use cases built around one asset: interpreted card-spend behavior.
The evidence should be ordered by strength. First come the official company pages because they state the company's own product claims, data scale, use cases, access methods and service boundaries. Second come regulatory records and privacy notices because they verify that Affinity operates in a category where data rights, consumer requests and audit controls matter. The California data-broker registration page lists Affinity Solutions as a data broker, gives a public privacy contact, website URL and physical address, and says the registration was approved on June 30, 2020: https://oag.ca.gov/data-broker/registration/186849. Affinity's privacy notice says the company is a provider of technology, analytics, data-processing and business services, runs reward programs such as card-linked offers for payment-card holders and issuing financial institutions, and was last updated on June 29, 2026: https://www.affinity.solutions/data-privacy-notice/.
Third come named partnerships and public market signals. Affinity's retail page says tools developed with the National Retail Federation, CNBC and Pyxis by Bain & Company support retail intelligence, and its Retail Monitor page lists monthly CNBC/NRF Retail Monitor posts: https://www.affinity.solutions/retail-qsr/ and https://www.affinity.solutions/retail-monitor/. A June 2026 Comcast Advertising announcement says Comcast would integrate Affinity transaction data into its Outcomes+ audience and measurement product: https://www.affinity.solutions/newsroom/comcast-advertising-affinity-solutions-purchased-based-precisions/. A June 2026 Affinity post says Snowflake named the company a leader in a Modern Marketing Data Stack report's collaboration category: https://www.affinity.solutions/newsroom/affinity-solutions-named-a-leader-in-snowflakes-modern-marketing-data-stack-report/. These are useful credibility signals, but they are not audited financial evidence.
Last come network and domain records. Domain RDAP records show affinitysolutions.com was registered in 2000 through CSC Corporate Domains, while affinity.solutions was registered in 2014 through the same corporate registrar. DNS lookups observed for this review resolved affinitysolutions.com to 66.246.174.131, www.affinitysolutions.com to Cloudflare addresses after a CDN name, affinity.solutions to Cloudflare addresses, and www.affinity.solutions to a hosted site name before Cloudflare addresses. ARIN RDAP identified 66.246.0.0/16 as a Cologix allocation and 104.16.0.0/12 as a Cloudflare allocation. Those records matter only in a bounded way. They show an accountable public web surface and use of common infrastructure providers. They do not prove transaction volume, data-processing architecture, customer uptime, private controls or revenue.
What the customer actually buys
The customer buys a usable answer, not raw card activity. Affinity's data page says its consumer purchase data can support planning, activation and measurement, and it describes ways to match merchants, categories and cards to first-party and third-party data and digital IDs: https://www.affinity.solutions/our-data/. For a brand, the usable answer may be whether shoppers who saw media bought more than a control group. For a retailer, it may be whether lapsed customers moved spend to a competitor. For a bank, it may be which cardholders should receive which offer and how the reward should be funded. For an investor or consultant, it may be whether a company's sales trend is weakening before the official report.
The account is implementation-heavy because each customer starts from a different question. A quick-service restaurant brand may care about in-store sales after a TV or audio campaign. A grocery chain may care about share of wallet in a region. A bank may care about inactive cardholders, low-active cardholders or attritors. A media seller may care about proving that exposure produced incremental purchases. A consulting team may want a category trend across geographies. The underlying data asset may be common, but the support memory is local. Someone has to remember how that customer's categories were defined, what exclusions were agreed, what time window mattered, which campaign file was matched, and what privacy threshold controlled the report.
Consumer Purchase Insights shows the broad planning account. The product page says CPI is built on more than 86 billion credit and debit card transactions and can be accessed through data clean rooms, API delivery or managed service, with customized reports packaged according to client specifications: https://www.affinity.solutions/consumer-purchase-insights/. The access methods are economically important. A clean-room customer needs secure collaboration and query discipline. An API customer needs stable integration, field definitions, monitoring and technical support. A managed-service customer needs analysts who can translate a business question into usable reporting. Each method turns purchase data into a relationship.
Consumer Purchase Lift shows the measurement account. Affinity says the service matches media exposure files, including control groups, to purchase data; calculates spend and conversion lift for exposed and unexposed groups; and can compute incremental return on ad spend when campaign cost data is added: https://www.affinity.solutions/consumer-purchase-lift/. That description exposes the support burden. If the exposure file arrives late, the control group is weak, the customer disputes the merchant set, or the result conflicts with a platform's own metric, Affinity must defend the logic. Support memory is the difference between a reusable measurement account and a one-off report.
Consumer Purchase Audiences shows the activation account. The page says Affinity combines purchase history with predictive modeling, supports syndicated and custom audiences, offers seasonal audiences and purchase-likelihood scores, and receives purchase data daily from financial partners: https://www.affinity.solutions/consumer-purchase-audiences/. The commercial value here is not simply "more segments." It is the ability to target proven buyers, competitor customers, lapsed customers or high-value spenders while respecting data rights and partner constraints. That requires ongoing maintenance: categories change, merchants expand, data partners adjust, media platforms update, and clients want new cuts.
Card-Linked Engagement and Consumer Bank Campaigns show the bank and merchant account. The Card-Linked Engagement page describes offer development, targeting, distribution through a bank app, website or email, a qualifying purchase and then a merchant performance fee plus cardholder reward: https://www.affinity.solutions/card-linked-engagement/. The Consumer Bank Campaigns page says performance fees are assessed for activated purchases rather than impressions, and mentions recent purchase data from banks, a dashboard and qualified transaction visibility: https://www.affinity.solutions/consumer-bank-campaigns/. This is not a passive data license. It is a live operating arrangement among bank, cardholder, merchant, offer terms, transaction recognition and reward fulfillment.
Why support memory becomes retention
Support memory is the stored knowledge of how a customer's data, contracts, privacy rules and business questions were handled last time. In Affinity's case, it is likely to include merchant-brand mappings, category definitions, campaign identifiers, exposure-file formats, bank partner limits, cohort thresholds, delivery timing, customer dashboards, retailer exceptions, reward terms and the vocabulary a client uses when asking for proof. If that memory sits with Affinity rather than the customer's own staff, the customer may renew because replacement would require rediscovery.
The strongest official clue is the access-method language. Affinity repeatedly offers protected routes to data: clean rooms, API access and managed service on the data and product pages: https://www.affinity.solutions/our-data/ and https://www.affinity.solutions/consulting/. A self-service software platform can win when the customer has enough internal staff and the use case is stable. A managed-service account wins when the question changes, the result needs explanation, or the customer's team lacks the time to maintain all the mapping and privacy logic. Affinity's retention asset should be strongest where clients repeatedly ask different questions against the same transaction foundation.
Support memory also matters because purchase data is sensitive and relational. Affinity's privacy notice says that as a service provider to financial institutions, the use of personal information provided by those institutions is limited to supporting and delivering FI-branded reward program services as specified in contracts with Affinity's customers and partners: https://www.affinity.solutions/data-privacy-notice/. That means a new customer cannot simply receive every useful individual-level detail and improvise. The service must respect contractual and legal boundaries. The supplier that remembers those boundaries can move faster than a replacement that has to renegotiate or reinterpret them.
Card-linked offers are a clear retention example. Affinity's public mechanics move through offer terms, targeting, distribution, qualifying purchase, performance fee and reward: https://www.affinity.solutions/card-linked-engagement/. If a bank changes its cardholder segments or a merchant disputes which purchases qualified, the supplier needs to know the old offer rules. If a reward fails to appear or a campaign underperforms, support staff must trace the issue through targeting, distribution, merchant recognition and transaction timing. The customer does not want a generic support queue. It wants people and systems that remember how that particular offer was built.
Measurement creates a similar dependency. The Consumer Purchase Lift page explains that exposure data is matched to purchases and compared with an unexposed control group: https://www.affinity.solutions/consumer-purchase-lift/. A brand may renew if Affinity can explain why one campaign showed lift and another did not. That explanation depends on prior context: media exposure, geography, shopping cycle, merchant coverage, time lag, control-group design and sales baseline. The retention asset is not only data. It is the vendor's ability to turn repeated disputes into institutional knowledge.
This is why the cheapest substitute may be less cheap than it looks. An in-house team can buy or license data, build models and run a warehouse. A large integrator can create an analytics stack. A media platform can provide its own attribution. A cloud data provider can host a clean room. But a replacement must rebuild the customer's practical memory: what every field means, which business users trust which metric, what changed after the last campaign, and what constraints the bank or brand accepted. That reconstruction cost is the moat Affinity is trying to own.
The data asset and its cost base
Affinity's most important asset is the transaction-data network it claims to access and interpret. The data page describes more than 150 million credit and debit cards, more than 100 million consumers, more than 86 billion transactions, more than $4 trillion of spend, more than 5,300 tracked brands, more than 2,000 brands with location tagging and up to five years of historical data: https://www.affinity.solutions/our-data/. The homepage repeats the purchase-intelligence proposition and frames the dataset as bank-direct purchase data: https://www.affinitysolutions.com/. Those public facts support a serious business case because scale, recency and breadth determine whether the answer is credible enough to shape budgets.
Scale does not make the account cheap to serve. A daily feed from financial partners must be received, normalized, matched, enriched, protected, audited and turned into outputs that a customer can understand. Affinity's Comet technology page describes cleansing, transformation, unification, safeguarding, tagging and enrichment as the steps that turn purchase data into insights: https://www.affinity.solutions/comet/. Even if some steps are automated, the cost base includes engineers, data scientists, privacy staff, product managers, customer-success staff, analysts, security controls, cloud or hosted infrastructure, partner-management work and sales support.
Merchant tagging is a hidden cost driver. A card transaction rarely arrives as a perfectly clean brand, channel and location record. It may contain merchant descriptors, acquirer data, merchant category codes, locations, franchise names, online marketplace entries, refunds, partial captures and timing differences. Affinity's data page lists spend information such as dollar spend, transaction count, unique cards or shoppers, brand name, merchant category, card type and channel: https://www.affinity.solutions/our-data/. The value lies in making those fields consistent enough for business decisions. The cost lies in maintaining that consistency when merchants change names, locations, ownership, channels or payment processors.
Privacy is another cost driver. The same data that makes Affinity useful makes it regulated and sensitive. The privacy notice says Affinity processes information for customer and partner work, service delivery, website integrity, legal obligations, product quality and security; it also says financial-institution reward services may include emails with offers, offer websites, authentication, transaction processing to determine rewards, merchant analysis and offer matching: https://www.affinity.solutions/data-privacy-notice/. A company selling this service must fund privacy operations, request handling, contract review, audit response and controls that reduce the risk of misusing cardholder information.
Security controls are part of the paid unit. Affinity's privacy notice says the company has technical, administrative and physical safeguards, uses measures such as encryption and redaction, has dedicated information-security and privacy personnel, is PCI DSS Level 1 certified by independent audit, and has SOC 1 Type 2 and SOC 2 Type 2 audits by an independent third party: https://www.affinity.solutions/data-privacy-notice/. Publicly stating those controls strengthens the reliability case. It also reveals a cost burden. Payment and financial-data customers do not buy insights alone. They buy the assurance that insight delivery will survive due diligence.
The public record does not show how these costs map to revenue. Affinity does not publish prices for Consumer Purchase Insights, Purchase Lift, Audiences, Bank Campaigns or Card-Linked Engagement. It does not disclose whether revenue is mainly subscription, usage-based, managed-service, performance-fee, enterprise-contract, partner-revenue-share or mixed. The Consumer Bank Campaigns page says performance fees are assessed for activated purchases rather than impressions, and the card-linked page says merchants pay a performance fee and cardholders receive a reward: https://www.affinity.solutions/consumer-bank-campaigns/ and https://www.affinity.solutions/card-linked-engagement/. That gives one public revenue mechanism, not the whole income statement.
The economic test is therefore margin after support. A large dataset can support high gross margin if access is repeatable, customers self-serve, partner fees are stable and support effort is moderate. It can become labor-heavy if every campaign, bank program, privacy review and dashboard requires bespoke work. The public evidence supports a valuable data and service asset. It cannot prove that the asset earns software-like margins rather than consulting-like margins.
Reliability as a commercial product
Reliability in this business is not only uptime. It is whether the customer can trust the answer. A campaign measurement product may be reachable every day and still fail commercially if merchant matching is wrong, control groups are weak, data feeds lag, channel definitions move or privacy rules force aggregation so broad that the result loses usefulness. Affinity's own pages emphasize recency, deterministic data and privacy-compliant design, but outside observers cannot test the private service record.
The reliability surface starts with data receipt. Consumer Purchase Audiences says purchase data is received daily from financial partners: https://www.affinity.solutions/consumer-purchase-audiences/. Daily data is valuable because marketers and banks want near-current signals. It is also operationally demanding. A delay from a financial partner, a format change, a file-quality issue or a matching failure could affect targeting or reporting. Affinity's customers may judge the service by how quickly those issues are detected and explained.
The second reliability surface is matching and attribution. Consumer Purchase Lift says exposure files are matched to purchase data and lift is calculated against an unexposed group: https://www.affinity.solutions/consumer-purchase-lift/. That process must be reproducible enough to satisfy media buyers, brands and partners. If a result moves a budget, it will be challenged. The supplier must defend the method without exposing sensitive information or overstating what the data can show. The most reliable supplier is not the one with the simplest claim; it is the one that can explain limits before a client mistake becomes expensive.
The third reliability surface is reward execution. In card-linked engagement, the user experience reaches the cardholder. Affinity's privacy notice says reward services may include processing transaction data to determine points or cash-back awards and analyzing merchant data to identify transactions at specific retailers: https://www.affinity.solutions/data-privacy-notice/. A reward that does not post, a transaction that is not recognized, or an offer that reaches the wrong segment can damage bank and merchant trust. Reliability is therefore not a back-office metric. It is part of the bank's customer relationship.
The fourth reliability surface is regulatory and privacy response. The California data-broker registration verifies that Affinity appears in a public registration record: https://oag.ca.gov/data-broker/registration/186849. Affinity's own notice says it is registered in California and Texas, and lists 2025 privacy request metrics, including delete and opt-out request volumes and response timing: https://www.affinity.solutions/data-privacy-notice/. Those disclosures are stronger than silence, but they do not settle the reliability case. A customer still needs private evidence about audit outcomes, incident history, partner reviews, consumer complaints, and how often privacy restrictions reduce output usefulness.
The fifth surface is public web and access infrastructure. Domain and DNS records suggest Affinity uses corporate domain management, AWS nameservers and Cloudflare or hosted public web infrastructure. That is normal and not proof of internal service architecture. Public web availability matters for marketing, privacy forms and contact routes, but the real service may run through private data exchanges, clean rooms, APIs and partner environments. The responsible conclusion is narrow: public records show an accountable web presence; they do not prove production resilience.
Reliability facts that would change the judgement are precise. The case improves if Affinity can show high data-feed completion, low late-file rates, audited matching accuracy, strong incident response, low reward-dispute rates, clean privacy-review history, rapid support times and clear service levels for APIs and managed reports. The case weakens if customers see recurrent data lags, unexplained attribution swings, merchant-mapping disputes, slow support, privacy-request friction, opaque reward qualification or partner feed dependence that Affinity cannot control.
Supplier dependence and customer dependence
Affinity sits between several powerful groups. On one side are financial institutions and data partners. On another are banks, brands, retailers, media platforms, consulting firms and advertising buyers. Around them are privacy regulators, payment networks, cloud providers, identity and clean-room vendors, media sellers and merchant ecosystems. The company creates value by coordinating these dependencies. It is also exposed to them.
Financial-institution dependence is the most important upstream risk. Affinity's privacy notice says the company may receive personal information from financial-institution customers or business partners as part of providing services, and that the financial institution's privacy policy applies when the consumer has a direct relationship with that institution: https://www.affinity.solutions/data-privacy-notice/. That is commercially reassuring because the bank relationship provides a consent and customer context. It also means Affinity's service depends on partner trust, contract scope, data continuity and regulatory comfort.
Data-partner concentration is not public. Affinity says it has bank-direct purchase data and a large number of cards and consumers, but it does not publish the number of financial partners, the share of data from top partners, contract durations, termination rights, geography by partner, data latency or partner economics. Those missing economics facts are crucial. A dataset with many stable partners is more resilient than a dataset dependent on one or two large relationships. A revenue model that shares value fairly with data partners is more durable than one that partners later try to internalize.
Customer dependence is equally important. Brands and retailers may buy Affinity because they want evidence beyond clicks, surveys, panels and foot traffic. The homepage explicitly contrasts purchases with proxy signals: https://www.affinitysolutions.com/. But large customers can be demanding. They may negotiate lower prices, insist on custom reporting, ask for unusual categories, challenge measurement outcomes, require privacy reviews and move budget if the result does not align with their internal view. A high-profile customer can make the product credible; it can also increase cost-to-serve.
Media-partner dependence shows up in the Comcast announcement. Comcast said its Outcomes+ integration would combine viewership data from more than 30 million Comcast households with Affinity's transaction-level dataset, and would help advertisers identify audiences and connect exposure to purchase activity: https://www.affinity.solutions/newsroom/comcast-advertising-affinity-solutions-purchased-based-precisions/. That is a strong use-case signal because premium TV buyers want proof beyond exposure. It also means the economic value depends on Comcast's product adoption, privacy design, advertiser demand and the ability to match exposure and purchase data without creating unacceptable risk.
Retail and industry-partner dependence appears through the National Retail Federation, CNBC and Pyxis by Bain references. Affinity's retail page says the CNBC/NRF Retail Monitor is powered by Affinity Solutions and offers retail insights based on real transaction data: https://www.affinity.solutions/retail-qsr/. The Retail Monitor page lists monthly posts with CNBC/NRF branding: https://www.affinity.solutions/retail-monitor/. Those references create visibility. They do not prove the revenue contribution, renewal terms or exclusivity of the relationship.
Cloud and platform dependence is also implied. The Snowflake recognition page frames Affinity in a data-collaboration context and links to Snowflake's report: https://www.affinity.solutions/newsroom/affinity-solutions-named-a-leader-in-snowflakes-modern-marketing-data-stack-report/. That signal is useful because customers increasingly want governed collaboration where data stays in controlled environments. It also raises a competitive question: if major cloud platforms become the collaboration layer, does Affinity retain pricing power as the purchase-data and measurement specialist, or does it become one data source among many?
The dependency map points back to support memory. Affinity earns retention if it can coordinate partners better than the customer can. It is vulnerable if partners or customers decide the coordination layer is replaceable.
Competition prices the substitute
Affinity's competitors are not only companies with the same label. The first substitute is a large integrator or consulting technology team. A major consulting firm can combine payment data, loyalty data, media exposure, cloud clean rooms, data science and executive reporting. That substitute may be attractive for customers that want a broader transformation. Affinity's counter-position is specialist purchase-data depth, existing bank-direct relationships and repeatable products. The risk is that a large integrator can own the client relationship and reduce Affinity to one input.
The second substitute is the customer's in-house team. A retailer, bank or media company with strong data staff can build internal models, purchase third-party data, use its own loyalty program and run clean-room collaboration. The in-house option is attractive when the customer has scale, privacy talent and stable use cases. Affinity's advantage is that many customers do not want to build or maintain the full purchase-data, partner-management and attribution stack. Its weakness is that sophisticated customers may internalize the most valuable logic once the use case is proven.
The third substitute is a generic software-as-a-service platform. Marketing measurement, customer data, loyalty, retail media and analytics platforms all offer dashboards and integrations. Their advantage is ease of procurement, familiar interfaces and broad ecosystems. Affinity's advantage is purchase behavior tied to bank-direct transaction signals. The contest is whether verified spend data is valuable enough to justify a specialist relationship when a customer already has multiple platforms.
The fourth substitute is a media platform's own measurement. Advertising platforms prefer to prove their own effectiveness, and some have large logged-in audiences, retail-media networks, conversion data and data-clean-room partnerships. Affinity's Consumer Purchase Lift product directly challenges proxy metrics by tying exposure to purchases: https://www.affinity.solutions/consumer-purchase-lift/. That can be valuable because a third-party purchase signal may be more credible than a platform grading itself. But platforms control distribution and can bundle measurement into media buying, making independent measurement a budget fight.
The fifth substitute is payment-network, bank or issuer analytics. Financial-services companies may already see card spend and can create offers, loyalty and measurement products. Affinity's financial-services page says it helps banks increase card spend, retain customers and use retailer and media access: https://www.affinity.solutions/financial-services/. The advantage for Affinity is cross-brand and cross-partner scale. The threat is that a bank or payment network may decide to build or buy the capability directly.
The sixth substitute is delayed automation. Many marketers still rely on clicks, surveys, panel estimates, internal sales reports or delayed campaign reviews because a purchase-data implementation takes time and budget. Affinity's homepage criticizes those proxies: https://www.affinitysolutions.com/. The do-nothing option remains powerful because it is organizationally cheap. A brand may accept imperfect measurement if no one is forced to reallocate budget. Affinity wins when the pain of uncertain spend becomes larger than the implementation cost.
Competition therefore does not invalidate the thesis. It defines its price ceiling. Affinity can charge more when purchase proof changes decisions, when bank or partner relationships are hard to replicate, when privacy-safe data handling is trusted, and when support memory reduces switching risk. It has less power when customers can answer the question from internal data, when a media platform bundles an acceptable measurement tool, or when a large integrator controls the broader strategy.
Market signals and what they can prove
Market signals support credibility but should not carry the business conclusion. The newsroom page lists recent content, including a June 2026 Comcast partnership, a June 2026 Snowflake recognition post, retail-media measurement commentary, case studies and videos: https://www.affinity.solutions/newsroom/. The site also says Digiday named Affinity "Best Measurement Tool" in an April 2026 awards item. These signals show category visibility. They do not verify revenue, margin, churn, data quality or customer satisfaction.
The Comcast partnership is the strongest named market signal because it identifies a concrete integration use case. According to Affinity's June 2026 announcement, Comcast Advertising would use Affinity purchase data in an audience discovery and measurement product, with advertisers able to target verified category buyers, competitor customers, lapsed purchasers and high-value spenders, then measure post-campaign purchase activity: https://www.affinity.solutions/newsroom/comcast-advertising-affinity-solutions-purchased-based-precisions/. This is a high-value problem. TV and streaming advertisers want proof that expensive media produces sales. But the announcement does not disclose deal size, exclusivity, revenue sharing, adoption rate or whether advertisers will renew after early campaigns.
The Snowflake recognition is useful for placement in a broader data-collaboration market. Affinity says it was recognized in Snowflake's Modern Marketing Data Stack report and describes its purchase-based collaboration as privacy-governed and grounded in deterministic data from more than 150 million cards linked to more than 100 million consumers: https://www.affinity.solutions/newsroom/affinity-solutions-named-a-leader-in-snowflakes-modern-marketing-data-stack-report/. That supports the idea that Affinity is not only selling reports; it is trying to be part of enterprise data collaboration. The limit is obvious: recognition in a vendor report is not audited commercial performance.
The NRF and CNBC references create a public proof channel. Retail Monitor posts turn Affinity's data into recurring public retail commentary: https://www.affinity.solutions/retail-monitor/. That can help sales because potential customers see the data used in a monthly market context. Yet public reports can also expose the challenge. If a customer's own data diverges from the public read, support staff must explain why. Is the difference a category definition, a regional sample, a merchant-mapping issue, a cohort boundary, or a real market difference? Again, support memory becomes retention.
The careers page is a weaker but useful signal. It says Affinity hires across technology, analytics, client services and marketing: https://www.affinity.solutions/careers/. That job-family mix matches the expected cost structure of a specialist data-service company. It does not show current headcount, turnover, hiring velocity or skill depth. It should be used only to support the operating-shape inference.
Market chatter beyond official pages should remain secondary. Awards, partner posts and industry mentions can show that Affinity participates in active advertising, retail and data-collaboration markets. They cannot establish whether customers are profitable, whether the data is superior to alternatives, or whether support response is strong. The public article should not convert visibility into economics.
Network and resource evidence in proportion
Network-resource evidence belongs near the end of the assessment because it can only answer modest questions. Affinity's public domains and DNS records show an organized corporate web surface. RDAP for affinitysolutions.com shows a long-lived domain registered in 2000 through CSC Corporate Domains, with AWS nameservers and an expiration date in 2027. RDAP for affinity.solutions shows a domain registered in 2014 through the same corporate registrar, also with AWS nameservers and an expiration date in 2027. DNS lookups observed for this review showed the legacy .com domain, the www .com host, the .solutions apex and the www .solutions host using a mix of direct address, hosted-name and Cloudflare address responses.
ARIN RDAP for the 66.246.174.131 address observed for affinitysolutions.com identified the containing 66.246.0.0/16 network as a Cologix allocation: https://rdap.arin.net/registry/ip/66.246.174.131. ARIN RDAP for the Cloudflare addresses observed on www.affinitysolutions.com and affinity.solutions identified the containing 104.16.0.0/12 network as Cloudflare: https://rdap.arin.net/registry/ip/104.18.6.62 and https://rdap.arin.net/registry/ip/104.16.150.108. These records are useful for accountability and infrastructure context. They should not be used to infer Affinity's private data infrastructure, customer traffic, application architecture or operating scale.
The reason for including network evidence at all is that Affinity's service is digital and privacy-sensitive. Public web domains host product, privacy, contact, content and request pathways. Corporate registrar management and large infrastructure providers are normal signs of an organization that maintains a public commercial surface. Cloudflare use may relate to public-site delivery, protection or hosting choices. AWS nameservers point to managed DNS. These facts are bounded.
The dangerous mistake would be to treat the absence of an Affinity-owned autonomous system or proprietary public network as weakness. A purchase-intelligence company does not need to operate like a telecom provider. Its critical reliability may live in private cloud environments, partner file exchange, data warehouses, clean rooms, identity systems and secure application services that are not visible through public DNS. Public records can tell readers that the website resolves and which broad providers appear in the public path. They cannot tell readers whether a daily bank data feed is late, whether a clean-room job failed, or whether a customer-support team resolved a campaign dispute.
The public network surface also says little about privacy compliance. A privacy form, policy page and marketing site are not the same as data-protection operations. The stronger evidence remains Affinity's privacy notice, its data-broker registration and its stated audit certifications: https://www.affinity.solutions/data-privacy-notice/ and https://oag.ca.gov/data-broker/registration/186849. Even those sources remain statements and public filings rather than a full audit record. Network evidence is a small supporting layer, not the core of the business case.
The missing proof grouped correctly
The missing economics proof is contract and margin evidence. Affinity does not disclose revenue, average contract value, renewal price, discounting, performance-fee take rate, partner revenue share, bank-data costs, cloud costs, managed-service hours, customer acquisition cost, gross margin or operating profit. It does not show whether Consumer Purchase Insights is mostly subscription, whether Purchase Lift is sold per campaign, whether bank campaigns depend on performance fees, or whether consulting and managed reports consume heavy labor. Those facts would decide whether the account has software economics, service economics or a hybrid profile.
The missing reliability proof is operating evidence. Public pages say Affinity receives daily data, uses privacy-protecting processes, supports APIs, clean rooms and managed service, and holds security certifications: https://www.affinity.solutions/our-data/ and https://www.affinity.solutions/data-privacy-notice/. They do not disclose feed completion rates, data latency, merchant-match error rates, model-validation results, reward-dispute frequency, API uptime, clean-room failure rates, dashboard availability, support response times, security incidents, audit exceptions or privacy-request error rates. Those facts would decide whether customers trust Affinity after a stressful use case.
The missing retention proof is customer behavior. Affinity's pages show many use cases and named public partner contexts, but they do not disclose cohort renewal, logo retention, net revenue retention, expansion after first measurement, bank partner retention, partner concentration, churn after disputed campaigns, customer-care satisfaction or repeat use by top customers. If customers expand from one campaign to recurring measurement, from one bank campaign to cardholder engagement, or from a managed report to an API or clean-room account, the retention thesis becomes much stronger. If customers use Affinity once to test a campaign and then move on, the thesis weakens.
These three groups should not be mixed. Economics is about whether the account is profitable. Reliability is about whether the service works under pressure. Retention is about whether customers keep paying after they have learned enough to leave. A single public partnership can support all three questions only weakly. It may show demand, but it does not show margin, operational consistency or renewal.
Keeping the gaps grouped also protects the article from generic caution. The issue is not that private companies disclose less than public companies. The issue is that Affinity's main commercial value is exactly where public evidence is least visible. Support memory exists inside prior integrations, support tickets, partner reviews, campaign disputes and customer trust. A public homepage can show the claim. It cannot show whether the memory actually protects revenue.
What would change confidence
The economics case would improve first with evidence that customers move from one-off reports into recurring accounts. A disclosed mix showing annual platform contracts, repeat measurement campaigns, bank-program renewals and limited custom-service hours would support the idea that Affinity earns durable revenue without rebuilding each project from scratch. Even a private-company summary of average contract duration, renewal bands or attach rates across Consumer Purchase Insights, Purchase Lift and card-linked services would be more useful than another broad data-scale claim. The strongest economics proof would show that support memory lowers future service cost: the second and third campaign for a customer should be faster, easier to defend and more profitable than the first.
The economics case would weaken if the company depends on heavy custom work that cannot be priced. A managed-service report can be valuable, but if every customer requires bespoke merchant definitions, bespoke privacy review, bespoke media matching and bespoke executive explanation, the business may resemble consulting more than scalable software. There is nothing wrong with that if pricing captures the labor. The risk is that customers benchmark the output against software subscriptions while demanding consulting-level care. Public pages show the service shape; they do not show which side of that line dominates.
The reliability case would improve with evidence that Affinity can quantify and explain its own operating quality. Useful facts would include daily feed completion, late-file recovery, merchant-tagging revision rates, reward qualification disputes, API incident counts, clean-room execution time, privacy-request handling quality and customer-support response by severity. The best form would not be a vanity uptime number. It would connect the operating metric to the customer outcome: campaigns measured on time, rewards credited correctly, privacy reviews passed, disputed merchant groups corrected quickly and dashboards trusted after a data issue.
The reliability case would weaken if the service produces results that customers cannot reconcile with their own records. Purchase data is valuable because it sees outside the customer's walls. That same outside view creates friction when a retailer's internal sales, a media seller's delivery file or a bank's card records do not line up neatly with Affinity's output. The supplier must be able to identify whether the issue is timing, channel coverage, merchant naming, customer mix, control-group design or a real market difference. If that explanation depends on a few senior people rather than repeatable support memory, the retention asset is fragile.
The retention case would improve with evidence of expansion after stress. The cleanest signal would be customers that renew after a failed campaign, a privacy review, a reward dispute or a major budget debate because Affinity's support team resolved the issue and preserved trust. Net revenue retention, repeat use by top brands, bank-program contract extensions, expansion from managed service to API access, and customers adopting both audience and measurement products would all support the thesis. The retention case would weaken if customers mainly use the service for isolated campaign tests, public market commentary, or one-time diligence projects without embedding it into recurring decisions.
These proof points are not unreasonable demands. They are the normal diligence questions for a private data-service business whose public identity rests on a large, sensitive and valuable information asset. The public record already shows why the company matters. The missing private record decides whether the market should price Affinity as a sticky support-memory account or as a useful but replaceable data vendor.
Judgement and watchpoints
The base judgement is positive but conditional. Affinity Solutions has a credible specialist position because its official record describes a large bank-direct purchase dataset, multiple monetizable use cases, privacy and security controls, named public market channels, and a product set that turns spending behavior into planning, targeting, measurement and rewards. The company is not a generic cloud-service label. It is a purchase-intelligence service account whose value depends on implementation memory and customer trust.
The positive case is strongest where the customer needs a repeated answer that is expensive to rebuild. A brand that needs recurring campaign lift tied to sales, a retailer that needs share and cross-shopping signals, a bank that needs cardholder engagement and reward qualification, or a media seller that needs independent proof of business impact may value Affinity beyond the raw dataset. In those settings, support memory creates switching resistance. The customer does not want to reconstruct every merchant definition, data-rights boundary, campaign file and reporting explanation each time.
The negative case is that the moat may be narrower than the data-scale numbers imply. If major clouds, payment networks, banks, retail-media platforms and large measurement providers offer adequate purchase-based collaboration, Affinity may face pressure on price and ownership of the customer relationship. If customers internalize the analytic logic after first use, the supplier may become a data input rather than a retained service account. If financial partners change terms or regulators tighten deletion and opt-out rules, the data asset may become more expensive to maintain.
The customer-dependence watchpoint is concentration. The public record does not show whether Affinity's revenue depends on a few large brands, banks, media partners or data relationships. A concentrated account base can make a specialist look stronger than it is because visible partnerships create credibility, but one lost relationship can change economics. A diversified base of recurring customers would support the thesis; a few large custom accounts would make the business more fragile.
The reliability watchpoint is explainability after failure. Purchase data changes decisions. When a campaign underperforms, when a retailer disputes market share, when a bank questions reward qualification, or when a privacy team asks for limits, the vendor must explain the answer without breaching trust. That is where support memory becomes a retention asset. A supplier that can explain prior choices and repair the issue may keep the account. A supplier that can only restate a dashboard metric may lose it.
The regulatory watchpoint is consumer-data rights. Affinity's privacy notice already acknowledges data-broker registration, rights requests, California deletion obligations beginning in 2026, and security audits: https://www.affinity.solutions/data-privacy-notice/. Regulation does not automatically damage the business. It can strengthen a specialist that has already built controls and weaken thinner competitors. But it can also raise cost, limit output, increase partner scrutiny and make support more labor-intensive. The effect depends on how much Affinity has automated compliance without losing client usefulness.
The final assessment is therefore not that Affinity Solutions is valuable because it has a large dataset. It is valuable if it can keep turning that dataset into trusted, customer-specific purchase decisions after the first integration, while controlling the cost of privacy, partner coordination, data enrichment and support. The facts that would change the judgement are not abstract. They are the economics, reliability and retention facts that public evidence cannot show: contract values, margins, partner costs, feed performance, matching quality, support speed, reward accuracy, renewal, expansion and concentration. Until those are visible, the strongest fair conclusion is that Affinity has the shape of a sticky support-memory account, but the size and durability of the retention asset remain private.