[ADS – Alliance Data Systems] Value Delayed, Value Revealed

In late November 2018, Alliance Data made an announcement long awaited by frustrated investors who, for years, endured missed growth expectations, operational fumbles, and a wilting valuation:

“Alliance Data Systems Corporation (NYSE: ADS), a leading provider of data-driven marketing and loyalty solutions, today announced that it is exploring strategic alternatives for its Epsilon business, including its digital media arm Conversant…This announcement is the culmination of a previously announced comprehensive internal and Board-led strategic review. Should the outcome be a divestiture, Alliance Data currently intends to apply the net proceeds to a combination of deleveraging as well as returning capital to stockholders through share repurchases and/or dividends”.

Nearly all of ADS’ value comes from its Card Services segment, but a discussion of its other two businesses provide helpful context.  

Epsilon

Housed inside Alliance Data Systems is a marketing services company called Epsilon, very similar in nature to Acxiom Marketing Services (AMS), which was recently sold to Interpublic Group and which I described in my LiveRamp post like this: 

“AMS partly helps manage the marketing operations of big brands – designing and maintaining customer databases, running custom performance analytics, aggregating third-party data, and pairing it with first party records to construct comprehensive consumer profiles for purposes of launching marketing campaigns – as part of splashy, multi-year, multi-million dollar engagements that can take up to 9 months to set up.  It is a business characterized by significant switching costs and minimal churn.  But it is also a mature enterprise with declining organic growth, struggling to keep pace in an environment where consumer attention is fragmenting across channels”.

The active ingredient to Epsilon’s business model is first party data from clients…that is, SKU level transaction information tied to clients’ customers, stored in a database built and maintained by Epsilon for the purpose of running targeted marketing campaigns.  So, Walgreen’s feeds the transaction data of its ~80mn Balance Rewards members, including yours, to Epsilon, who might supplement that information with third party data to get a fuller view of who you are and what you’re into. Epsilon will then activate a slew of third-party martech platforms to deterministically match your offline identity to your anonymized online one so that it can serve you relevant ads as you browse the web.  For instance, if you purchased shampoo during your last Walgreen’s trip, you might see a Walgreen’s ad for body wash as you read an article on GQ.com (if the process by which this happens doesn’t make sense to you, you may want to read my LiveRamp post, which gets more into the grit). 

Epsilon’s value add is that it stitches together and implements the fragmented set of technology platforms required to get from shampoo purchase to online GQ ad on behalf of CMOs at large companies who would otherwise be overwhelmed by all the newfangled products and confused about how to optimally integrate and use them.  But prior to 2014, Epsilon didn’t own its own tech.  Per the diagram below, the company was at the front end of the ad ecosystem (under services) and nowhere else.  Rather than build a proprietary stack, Epsilon seemed fine relying on “best-of-breed” technology from third party vendors. 

But with competitor Acxiom extending its reach into other parts of the ad tech value chain, in September 2014 ADS announced its $2.3bn cash/stock acquisition (~10x forward adj. EBITDA) of Conversant. Conversant is an amalgamation of several ad tech businesses¹ that can be divided into two functional categories: an affiliate marketing network that matches 3,000 advertisers with 60,000 publishers² and an onboarding business³ similar to LiveRamp⁴ (but more comprehensive in certain respects) that deterministically matches the offline identities in a client’s first party database with unique and anonymized online IDs in order to deliver targeted ads.  It can also target netizens who share the demographic or behavioral characteristics of a shopper in the retailer’s loyalty database.  Epsilon was basically accomplishing the same job for its clients as before – delivering online ads to customers as part of a high touch, “leave it all to us” value proposition – only now it owned (and could cross-sell) an onboarding service that, leveraging Conversant’s vast database of detailed digital profiles⁵, could achieve offline-to-online match rates that were superior to peers.

But management bungled the integration of Epsilon’s and Conversant’s tech platforms and failed to live up to its promise of delivering an effective end-to-end solution that linked first party client data with cross-device online IDs.  But it may not have mattered anyways.  While some parts of Epsilon -Conversant, email marketing, marketing to the automotive channel – were growing by high-single/low double-digits in 2016 and 2017, the exhaustive high touch service that Epsilon had long pitched as its key value prop and that comprised nearly 25% of its revenue, was under attack by SaaS providers, who began selling cheaper, light weight, “good enough” off-the-shelf self-serve systems directly to CMOs.  Clients were less willing to pay the same exorbitant sums and wait the same 9-12 months for Epsilon to set up their loyalty programs and build their marketing databases.  Epsilon was manufacturing customized Ferraris for clients who seemed increasingly fine driving mass produced Chevys.  A business that used to consistently generate 8%-10% growth, instead contracted, forcing management to to simplify the product and reallocate labor to Bangalore (all customer support at the Card Services segment, on the other hand, is domiciled in the US).  While this was going down, Epsilon’s traditional Mad Men style creative agency (only a single digit percent of revenue, but a perennial money loser), continued to suck, as did affiliate marketing, with both businesses’ combined revenue contracting double digits in 2018, dragging full year 2018 revenue growth, which management guided to +4%-5% growth in April of last year, down to -4%-5% ytd, subtracting all the incremental revenue from 2017.  To put these challenges in broader context, in 2013, prior to purchasing Conversant, Epsilon’s Adj. EBITDA was $290mn.  Conversant’s EBITDA at the time of its acquisition was $250mn.  The combined $540mn of EBITDA has deteriorated to $480mn LTM. 

Card Services

The private label Cards segment is the meat of this company.  While this business makes money by charging interest on credit card loans, much like a bank, it creates value by leveraging first party transactional data from private label card programs to encourage shoppers to spend more, much like a marketer. 

It is easy to confuse general purpose co-branded cards with private label cards, so let me quickly explain how they differ.  A co-branded card bears the logo of the sponsoring merchant and, like any ol’ credit card, can be used anywhere Visa or Mastercard (or whatever card scheme logo appears on the card) is accepted.  The cardholder earns merchant-specific rewards by using it, with more rewards doled out per dollar of spend when the card is used to buy stuff from the sponsoring merchant.  A private label card, on the other hand, can (typically) only be used to buy stuff from the sponsoring merchant. But in return for its limited range, the card offers the shopper a richer set of perks from that merchant (discounts on future purchases, free shipping, extended refund windows).

Here is a general purpose Visa-Cathay Pacific Airways co-branded card issued by Synchrony Bank that offers customers of Cathay Pacific 1 Mile of travel for every $1 spent on the Card in the US and 2 Miles for every $1 spent on Cathay Pacific purchases made online or in-flight.