The Purchase Data Playbook For Marketers

While tapping into purchases helps advertisers close the loop around the actual conversion, the use cases for purchase data have evolved far beyond measurement.

The use of transactional data is maturing in media activation, as is the ability to commingle it with other data sources like behavioral, location and even panels in channels like addressable TV.

Integrating purchase data with other data sets gives marketers more insight into consumers. In some cases, the combination of geolocation, loyalty or other offline data might indicate a high-value customer visited a store, but perhaps didn’t purchase.

Purchase data can also help optimize ad buys, and can be coupled with viewability measurements to correlate the impact of a digital ad exposure on driving sales, said Nielsen Catalina Solutions CEO Matt O’Grady.

And increasingly, advertisers want custom insights which combine online and offline purchase information with other psychographic or interest-based data sets, said Anant Mathur, global head of analytics for the media agency Essence.

And while it’s no surprise that financial services companies are making their purchase data sets available, new companies are joining the milieu – including Amazon, which integrates such data into an audience-targeting platform, and Google, which activates transaction data to connect digital ads to purchases.

Purchase and transactional data sets come in many flavors, from anonymized purchase data collected directly by credit card issuers to loyalty card information and other third-party data sets aggregated by a group of specialty providers.

AdExchanger broke down some of the big purveyors of purchase data. 

Companies That Own Transaction Data

Visa Advertising Solutions

The offering: Visa Advertising Solutions, which launched fall 2016, offers two products: Visa Audiences helps marketers target digital advertising across mobile, social, display, video and addressable TV. And Visa Ad Measurement provides ROI measurement. Visa Advertising Solutions partners with Oracle Data Cloud and other service providers to market Visa’s data segments directly to brand marketers and more than 100 big US advertisers.

Scale: Visa claims to analyze about 18 billion US financial transactions per year and $1.3 trillion in US spend.

Its secret sauce: “Because Visa processes a large proportion of transactions, we are uniquely positioned to help merchants action this information for purposes of targeting and measurement,” said Sandeep Milar, head of Visa Advertising Solutions.

For example, Visa Audiences can help marketers find new or lapsed customers to target their campaigns at scale. It also helps marketers find people who spend heavily in their category, but not with their particular brand, so they can acquire new cardholders.

MasterCard Ad Intelligence

The offering: MasterCard Ad Intelligence is a software-as-a-service tool that primarily works with marketers in the retail, restaurant and travel sectors. These clients use Ad Intelligence to tap insights based on MasterCard’s aggregated and anonymized transaction data, which can be used in campaign planning, activation, ad decisioning and post-campaign measurement.

Scale: MasterCard claims to process more than 54 billion financial transactions per year.

Its secret sauce: The real-time nature of its approach.

Unlike historic media-mix models, which measured ad effectiveness after a campaign was complete, MasterCard Ad Intelligence claims it can correlate digital ad effectiveness on offline sales throughout the life cycle of a campaign and on the fly – all in a self-serve interface.

American Express

The offering: Although American Express has been engineering a data business for years, it launched a formal data-driven business called Amex Advance in early November [read AdExchanger coverage]. Amex Advance is designed as a safe-haven environment run on Acxiom, where advertisers can match their own first-party data to Amex’s deidentified cardholder data.

Amex, which has a managed services team embedded at Acxiom, helps marketers find and target their brand's consumers based on their past purchase histories, their life stage (e.g., recently relocated) or their likelihood to purchase in the future.

Scale: Amex has access to data from more than 100 million card members representing more than $1 trillion in sales and charge volume globally each year.

Its secret sauce: Amex transaction data sets are interoperable – and can be used in combination with insights from other media and data partners. For instance, Acxiom is building “predictive intent segments” using a combination of third-party data from its own network with Amex’s anonymized cardholder segments, while publishers and TV nets also tap Amex’s segments to do predictive modeling against their own audiences.

Amazon

The offering: Although the extent of Amazon’s data offering remains a mystery, it has pulled back the curtain on its growing audience platform.

Amazon Advertiser Audiences lets marketers upload CRM data and email lists to match against Amazon’s rich set of first-party audience and ecommerce data.

For now, Amazon’s audience segments are only available to advertisers buying Amazon directly, and can’t be activated within third-party ad platforms. But rumor has it Amazon is working on enabling API access as buyers push for more performance from their Amazon ad campaigns.

Scale: About 65% of US households, or close to 90 million people, subscribe to the premium membership service Amazon Prime. And Amazon acquired Whole Foods in August, which grossed about $15.7 billion in retail sales in 2016.

Its secret sauce: Amazon can give marketers the best of both worlds. It provides both reach (because of the size of its customer base and their purchase frequency) and precision. Amazon’s degree of insight into product-level data is hard to match. But its ability to tie a media exposure back to offline sales (enter Whole Foods) could prove fruitful for marketers’ attribution models.

“Purchase insights play a big role in attribution models for clients,” said Joe Migliozzi, Shop+ lead for Mindshare North America. “Advertisers [want to see if they] sold more products when media placements were delivered in a market where the advertiser had large store displays [in-store] or that sales decreased when a small regional competitor did significant discounting for a local holiday.”

Companies That Aggregate Transaction Data

Nielsen Catalina Solutions (NCS)

The offering: As a joint venture with measurement giant Nielsen, NCS combines its own sets of CPG purchase data with Nielsen’s digital and TV panel data. NCS claims this combination gives it eyes into nearly all CPG spend at grocery, drug, mass and discount stores in the US.

Scale: NCS provides visibility into purchasing habits among 90 million households down to the universal product codes (UPC) – via Catalina’s loyalty card database spanning 18,000 stores.

Its secret sauce: Granularity. NCS refreshes its transaction data daily and provides more granularity than SKU-based purchases. UPCs are consistent across merchants, whereas SKU data applies only to single retailers.

Oracle Data Cloud (formerly Datalogix)

The offering: Oracle Data Cloud helps marketers reach current and potential customers, and to attribute campaign results back to actual sales. Through the old Datalogix offering, Oracle inherited deep data assets in verticals ranging from CPG to retail and automotive. Oracle Data Cloud also formed relationships with Facebook, YouTube, Pinterest, Twitter and Snapchat, allowing it to tie digital ad exposures back to ROI.

Scale: Oracle Data Cloud claims it accesses more than $3 trillion in annual consumer transaction data, all connected across desktop and mobile devices to more than 5 billion user profiles via Oracle's device graph.

Its secret sauce: Prior to its acquisition by Oracle, Datalogix was one of the first companies to connect the dots between digital profiles and real-world purchases. Oracle Data Cloud claims to have both scale and depth via data-sharing relationships with partners like IHS Automotive (Polk) and Visa Advertising Solutions.

Cardlytics

The offering: Cardlytics’ solution is called Cardlytics Direct. It’s powered by a network of financial services partners, who allow Cardlytics to embed its technology into their servers. In total, Cardlytics collects purchase data from more than 1,500 financial institutions. Dani Cushion, Cardlytics’ CMO, said all the processing happens on the bank partners’ premises and no personally identifiable information ever leaves the bank.

Cardlytics Direct lets marketers target anonymized customers based on purchase history and measure the actual in-store and online sales transactions.

Scale: In 2016, Cardlytics claims it analyzed 18 billion US transactions, and $1.3 trillion in US spend, across debit, credit, ACH and bill pay spend.

Its secret sauce: Cardlytics applies advanced analytics to aggregated purchase data to make it more actionable. Cardlytics’ analytics are prescriptive. For example, Cushion said, a popular apparel company knew that 61% of its sales came from 20% of their customers, which made them appear very brand-loyal.

But by looking across a greater share of the wallet, it turned out that only 5% of these customers were actual brand loyalists (defined as having two-thirds of their category spend with that retailer) while the remaining 95% of “loyal” customers were also shopping around the category and spending even more with direct competitors.

Google

Its offering: Google’s offering, Store Sales Measurement, connects mobile ads with in-store purchases. Although Google doesn’t have visibility into UPC-level data the way NCS does, Google’s login information gives marketers a deeper view into the conversion. [Note: Google would only use that information for users who have consented to associate their Web and App Activity history with their Google account.]

Scale: Google claims it has insight into about 70% of US credit and debit card transactions.

Its secret sauce: Google developed double-blind encryption technology that ensures users' data remains private, secure and anonymous. Google only reports anonymized and aggregated conversions at the campaign level. It also doesn’t report on individual transactions, but models out the value of a transaction over multiple purchases.

Although Google doesn’t identify its financial partners, it works with all of the major credit card issuers to support a double-blind match via Google's credit card-based solution, which is still in beta. And, an advertiser could also import loyalty card data or in-store transactions via a hashed ID directly into AdWords to measure in-store sales conversions via its loyalty card solution.

Worth A Mention

Commerce Signals: Commerce Signals is a startup staffed by a bunch of former bank and marketing data execs, offering a self-serve tool with “near-real-time” measurement of sales performance and an ad campaign.

Affinity Solutions: Affinity Solutions offers a “purchase-driven marketing cloud” that uses machine learning to predict consumers’ likelihood to purchase in the future based on their buying behaviors.

Kantar Shopcom: Backed by the mainstay Kantar measurement brand, Shopcom blends data assets from CPG, retail and trade marketing sources, claiming insight into 90% of all US household spend across 450 retailers and 18,500 brands.

IRI: IRI is a consumer market research and analytics firm that helps hundreds of brand manufacturers activate purchase segments in other channels like digital and TV. IRI claims its specialty is connecting disparate data sets like loyalty, social, purchase and behavioral-based data sets around media.

Correction: Google claims its Store Sales Measurement tool does not leverage location data, as the story earlier indicated.

1 Comment

  1. Wondering what would be the pros and cons of utilizing the data owner (Visa, Mastercard) versus a data aggregator (Cardlytics, Google, etc.)? Obviously it probably depends on the use case but that insight would be helpful.

    Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>