Data Games – Battle of The Cloud Part 6

Understanding flows of data, and the structures in which it is controlled, provides a map of: value, power and margin. What is changing in the flow of data?

Warning.. biggest blog ever.. So I made a two page summary. 

Happy New Year! Best to you and yours. Having completed the successful sale of Commerce Signals to Verisk last year, this blog is a reflection on some of my lessons learned as well as my predictions on where I see things headed. The thoughts here are guiding my investments and launch of my next venture. I love the interaction, so please take time to write a comment on any of this. Also I ask for your pardon in advance for typos.. 

Understanding flows of data, and the structures in which it is controlled, provides a map of: value, power and margin. What is changing in the flow of data? What data is still “unique”? Where is power shifting? My past blogs referred to this dynamic as Rewiring CommerceValue Orchestration and the Transformation of Commercial Networks

Summary

  1. Primacy of Consumer [Data] combined with Ability to Deliver Value. The wide availability of data (even sensitive data) has caused the price of data [in isolation] to move toward 0. The value of data is driven by two factors: ability to combine it, and the ability to act on it. Google has indexed the world’s public information. There is much private information that remains “locked”, with limited value due to its ability to “play”. Consumer permissions flow with consumer “value”. Value should be measured across the consumer journey from intent to purchase. 
  2. Last Mile: The MOST valuable commerce data comes from direct consumer interaction. My top criterion for evaluating investments: where is direct interaction changing? Last mile specialists (ie Amazon, Adyen, Shopify, hyperWALLET/PYPL, …etc) have become specialists in MANY verticals. Specialization is beyond platform functionality as data powers customer experience. Specialists have a unique ability to “embed” in core systems, gain unique insight, and ability to sell and deliver new products (the virtuous cycle)
  3. Friction Drives Change. As discussed in 2014, Authentication in Networks, the most important battle in banking and commerce surrounds consumer identity. Bank margin is driven by ability to manage risk. KYC and financial history was the historical strength of banks.  However, bank eCommerce rates/rules (ex merchant fraud) made KYC and transaction risk management key to retailer margin. Large online merchants have created deep expertise powered by direct consumer insight and now manage fraud to less than 3bps. It is little wonder that they now use this to power their own financial services (ie Amazon, Target, Walmart, …).  
  4. Banks are losing the battle. Data flow within payment schemes was DESIGNED. For example, Merchants don’t trust banks with SKU, Banks don’t trust merchants or V/MA with consumer information, .. etc.  In other words the parties within the payment agreement knew that this data was not necessary to fulfil the core obligations of the service. These restrictions are not in place where consumer identity and actions are tracked (ie eCommerce). The bank challenge is threefold: 1) direct consumer engagement (by others not similarly constrained), 2) ability to partner within data rights and 3) backend “leakage” by retailers and data aggregators with (opaque consumer consent). 
  5. US Banks are working to structure data access. The top 6 banks have pulled their data from all marketing and “monetization” efforts as the revenue did not justify the risks of compliance. Banks are working to support well defined customer data requests to approved 3rd parties (ex US “voluntary” open banking). Bank data efforts are focused on: 1) stopping leakage, 2) changing from a model of screen scraping to directly permissioned APIs (ie Akoya) 3) gaining consumer permissions 4) getting agreements in places with 3rd parties.  While this is similar to Europe’s PSD 2 and “open banking”, it is distinctly different technically, and operationally. Banks win when they create a network with shared rules and defined access. 
  6. Discount the Value of Bank Data (outside of banking). Banks are the original data business and will continue to play an instrumental payments role in commerce (blog). Regulation and limited ability to ACT throttle the value of bank data (ex. “intersections” which banking data can play). Additionally the wide availability of proxy data (w/ no restrictions on use) limit opinions for participation of “quality” bank data. Example: Card Linked Offer companies can no longer target on individual behavior.. New competitors like DOSH use mobile data to target and deliver a mobile first experience for any card. 
  7. Invest in Retailer Data Champions. Retailers like Amazon, Target and Kroger have become new forces in data. Retailer’s key advantage? The ability to act.  Data science is no longer the realm of specialists, all major companies now treat it as core. Retailers have become data and advertising giants (Nov 11 WSJ). Amazon’s has become a solid #3 with a 51% CAGR and 2020 revenue almost $13B. Within the last 5 yrs, they have unmask virtually all card transactions tying consumer to purchase items (cards are no longer anonymous). While banks and tech platforms face enormous privacy scrutiny, retailers have thus far avoided it, making them the #1 leakage point for payment data.   
  8. Ad Dominance of Google, FB, Amazon. As a corollary to Retailers’ success in data is the abject failure of [digital] advertising industry.  Last October, Commerce Signals found that 80% digital advertising (outside of Goog/FB/Amzn) fails to produce value. The ad industry continues to push measures of success that do not matter (ie “clicks and engagement”) because they know what sales data shows. The ad industry is largely powered by data of unknown quality and source. As one CEO said “do I have a list of men over 40 that shave with Gilette? Yes. Do I know where the data came from or how old it is? No.. because I would have to pay for it.” Retailers have thus created their own data and ad teams (ex Target, Walmart, Kroger, Amazon). Agencies are quickly shrinking to creative, design and rote execution of well informed client plans.  
  9. Invest in Alternative Data (and Modeling) with clear data rights and data “chain of custody”. Each Fortune 500 company treats the management of data intersections similarly: bring your data into my environment and we will assess. You can imagine a discussion between Google and a large bank. Specialists like Axciom/LiveRamp historically played an intermediary role. To analyze data combinations in a well defined mutual objective, with anonymized results (see blog). Value can only be assigned to data where it flows. Regulated companies have challenges sharing data AND creating an economic model where value is shared (see Trust Network for more info). Alternative data does not (currently) have this challenge (ex location data). 
  10. New Bundles are accelerating. Examples are Stripe/Treasury (WSJ), Adyen and Google’s massive relaunch of Google Pay (see blog). Within payments, I’m most impressed with Adyen and Stripe’s software platform and PayPal’s ability to partner.  
  11. Rights on data are key. Regulatory hurdles are increasing fast. California is setting the national standard on privacy. Beyond CCPA, California just passed “part 2” of privacy law: California Privacy Rights Act “CPRA” on 2 Nov 2020. Before investing in ANY fintech, understand their data advantage and the rights they have on this data: the right to STORE, the Right to USE, and the right to SHARE (see blog). Regulatory efforts focused on privacy will only lock in the advantages of leading platforms and present hurdles to collaborations of smaller players. 
  12. Apple will be a data (and payments) disruptor. Within the crowd of platforms, retailers and banks, Apple stands out as unique in their approach to data: it belongs to the consumer (see Apple – Consumer Data Champion). Their direct consumer relationship, unique approach to consumer privacy, leadership in biometrics, and control of a secure platform enables them to pursue unique innovations. For example I believe Apple is working with both the networks and one large acquiring bank (?JPMC) to redesign EVERY consumer’s ability to accept payments (see Apple/Mobeewave).  They also have a unique ability to broker consumer identity with merchants, governments and healthcare providers (see blog). 
  13. V/MA and Paypal will be hubs for innovation (and my largest holdings). There are 3 drivers: 1) “ubiquity” with every consumer, merchant and bank participating, 2) defined operating model with supporting economic incentives for each participant, 3) focus, with PayPal having a big edge (Building Networks). 

Background

Today’s blog is about data. As many of you know, we sold Commerce Signals to Verisk Financial (Dec 2019). Verisk is an amazing company with deep bank agreements and even greater data. It is great to see the Commerce Signals engineering team (and platform) operate as the core of new data products.  Given my recent experience I wanted to share a little insight into the data world, particularly the flows within payments, banking, advertising and “commerce”. 

Foundational Story – Data Sharing and Value

Back in the .com days I ran a portion of Oracle’s new stuff division. As we led the rollout of B2B exchanges and Rosetta Net, the challenge of getting suppliers to share data was enormous. It wasn’t a technical issue.. It was an economic issue. As one Cisco supplier said “I’m not going to automate my Work in Process (WIP) response to Cisco.. If they don’t like what they see they will switch to another supplier in minutes” He had no incentive. Similarly in payments, retailers closely guarded all of their pricing, inventory and transactions. Agreements took years to complete.

During my time at Google Commerce in 2011, Osama directed the team to let me see everything. Paul Lee, a brilliant young engineer opened one of the few locked doors in GWC1 and showed me a screen with the real time inventory of 5 national retailers. I told him “this will never scale.. It will take you 18 months to get agreements with each retailer” (thinking that this was just a concept). He said “this is live, and we have 10 more retailers waiting to participate with NO AGREEMENTS”. I was shocked, wondering how on earth Google could get retailers to send real time store level inventory without an agreement. I asked “why are they sending this to you”? Paul responded “I don’t know, I just published the API allowing them to send it to us and I guess they see value in it”.

My internal paradigm between Data, Value, Trust and Legal Agreements was just reset. Now we know that this function powers “buy now”, shopping and local. Google uses this to help local retailers compete against amazon.. consumers discover local availability and google ships the goods to the customer’s door that day (FOR FREE). The value here is clear: retailers that shared data would allow customers searching for “wicker chair” to see inventory at the local Crate and Barrel with a buy now and shipment to your door in 4 hours. 

Google’s ability to capture intent is well known, as is Merchant’s challenge in influencing a consumer prior to sale. Google’s competition for product search is Amazon.. Retailers would never give this data to their principal competitor (ie Amazon). So how can Google use it? Variants of this data story are being repeated 100s of times a year in different verticals. The keys I look for (Tom’s investment rules of thumb):

    1. Who touches the consumer first?
    2. Who has data that can deliver value (in a new bundle)?
    3. Who can take action?
    4. Is there an economic alignment of interests between parties?
    5. Do I trust the parties involved?
    6. Does the consumer trust the party taking action?
    7. Legal/Regulatory constraints?

What is the role of “payments” in the above example? It is an afterthought.. It doesn’t matter. Payments work. This is why I laugh at pundits thinking Google, Facebook or Amazon want to be a bank. Specialists have created tremendous value bundles which have embedded payments.  Just like connectivity, or power.. payments are always on and work (an infrastructure service).  The good news for payment players is that embedding payments leads to increased volume. V/MA are the only ubiquitous and profitable payment services (stand alone). 

The three best companies to research as case studies in embedded payments are: Adyen, Stripe and Shopify. As I related in Embedding Payments, Adyen has redefined the economic model of “payments” company. When Adyen shows up at a merchant, its as if they are a software platform company. They talk about acceptance, and POS .. but they also talk about payroll, AR/AP integration, supply chain. Adyen acts as a multi-specialist solving multi “last mile” business problems with a team of enterprise software specialists. Their direct business relationships place them to act as a financial services switch (on Treasury and Retail). Their competition? Paypal has a team that talks about a buy button with a sales head from carrier billing (my biggest ding on my largest investment).

When I look for new value bundles and vertical specialists, it is beyond platform functionality as data powers customer experience. What radical new customer experiences are being created? Specialists have a unique ability to “embed” in core systems, gain unique insight, and a channel to sell/bundle new products (the virtuous cycle)

In my view, tremendous success of Salesforce was due to a combination of fantastic team and the MOST important data for any company: sales. They were the first enterprise platform to take advantage of a virtuous cycle and execute in a pay by the drink cloud platform model. Their own payments platform evolved from invoicing, to enabling the direct channel (selling Salesforce+Stripe cloud payments).  

For established payment players FIS and GPN have the best software platform teams. GPN has an edge here as they have numerous vertical specific merchant solutions and are more nimble. 

Friction as Driver of Change

For V/MA ubiquity comes with a cost – time to change and rigidity. Effective networks (see blog) are wonders of business and social interaction that largely reinforce an existing pattern, product, or social structure. Networks are resilient to change as they create value to all those connected… and this value expands as the network grows (network effects). The reinforcing nature of networks has proven effective in insulating participants from being impacted by change and keeping disruptive forces at bay.  As discussed earlier this year in Network of Networks, trust is domain specific, thus current networks are constrained by BOTH the rigidity that comes with scale and by trust extension. 

As discussed in 2014, Authentication in Networks, the most important battle in banking and commerce surrounds consumer identity. Bank margin is driven by ability to manage risk. KYC and risk decisioning based upon financial history was the historical strength of banks.  Direct consumer interaction by merchants and platforms combined with the friction of CNP rates/rules (ex merchant fraud) made KYC and transaction risk management key to retailer margin. Large online merchants thus created deep expertise powered by consumer insight (managing fraud to less than 3bps). The ability to manage CNP risk thus resides outside of banking. It is little wonder that retailers use data to power both customer experience and their own financial services (ie Amazon, Target, Walmart, Stripe,  …).  

Illustrative Story – Amazon Business Lending powered by Marketplace data

We are just beginning to see the disruption that will take place here (see Dec 30 WSJ – Buy Now Pay Later). LOOK FOR FRICTION in experience, time and cost. A fantastic start up along these lines is Payzer focused on retail contractors. A company building a house or installing a new HVAC has a montage of payment, invoicing and financing challenges. Payzer allows a coordinator to issue job specific debit cards to all employees.. Turn them on/off, set purchase limits and assign costs to jobs. They also allow both home owner and contractor to finance. Everyone can see bills of materials, time, expense.. Brilliant industry focused solution in a $1.7T industry (US).  

Banking/Payments

In 2013, Battle of the Cloud Part 5 I painted a picture of large issuers working the innovate outside of V/MA networks and the challenge they face: there is no other payment network that will deliver higher margins or better economics. The V/MA network is the least common denominator, and the ONLY entity with agreements across all “invested” parties: Merchants, Consumers, Banks. The issue is not technical, it is that networks require connections (ex Chase Pay). I am adamant here, banks are best served by innovating with existing networks. There is no other model with better economics or better adoption. 

P2P story

Much has been made of the battle between Venmo and Zelle, the majority of both P2P schemes use V/MA rails (OCT/AFT). For Zelle, only 6 banks which have the necessary EWS and TCH integration to settle directly (TCH RTP) use another scheme. Zelle is the best bank payment innovation in the last 20 yrs.. They prove that banks can win with a fantastic product. The issue is how they can use it beyond P2P.  For PayPal we see pilots with QR codes in CVS. Walmart asked Zelle to work on acceptance.. They said “no” our banks won’t let us..  So we have the largest merchant in the world asking how they can accept your new payment instrument.. And you can’t move?   Banks put people, products and technologies in nice little boxes. Expansion creates channel conflict and cannibalization…. Now you know why bankers have grey hair. 

While banks are viewed as the original data business, they DO NOT externalize a virtuous cycle. Consumer insights and opportunities are locked up (or thrown away). Banks have limited ability to act themselves, or partner with those that can act. This combined with the eCommerce loss above has led to a rapid atrophy of bank data (see Banks as a Data Business and Changing Economics of Payments). 

Banks win when they create a network with shared rules and defined access. Top US Banks are working to structure data access: 1) stop leakage, 2) change from a model of screen scraping to directly permissioned APIs (ie Akoya) 3) gain consumer permissions 4) put in place agreements with 3rd parties.  While this is similar to Europe’s PSD 2 and “open banking”, it is distinctly different technically, and operationally (see Open Banking). Banks win when they create a network with shared rules and defined access. 

There is an obvious “race” to build a bank data network (I know.. I tried). V/Plaid and MA/finicity and Akoya are the lead US contenders. I am rooting for Akoya.. It is the only one that will have clear consumer permissions. The banks will also push all screen scraping through them, so why would a start up want to use Plaid if they can get the same data directly from a bank? Speed of agreement is possibly the driver.. But Akoya may be able to manage this. 

My view is that the DOJ’s effort to block Visa’s acquisition of Plaid may be a huge blessing in disguise for Visa. Plaid has neither the consumer agreements or the bank agreements that drive data transparency, privacy and future value. The card will be a central hub for value delivery regardless of who the enablers are. Plaid is seen by banks as an “end run” on their efforts to control data flow. While Visa’s sales team did a great job checking with the Bank’s to ensure there were no show stoppers before they made the announcement, the banks purposefully announced Akoya consortium 5 days after Visa’s $5B Plaid announcement.  Obviously not a healthy sign. 

This is my biggest hit on Plaid, while they claim not to store any data, they actually store the MOST important piece: bank/issuer log in credentials. I doubt if most consumers realize that giving a banking password out during a Paypal account verification meant allowing Plaid to access your bank account for up to 7 yrs. While banks, mobile operators and the tech platforms have faced enormous scrutiny over data use.. Retailers have not.. And will be the next ones to face a reckoning (in 2021). 

Discount the Value of Bank Data (outside of banking). Banks are the original data business and will continue to play an instrumental payments role in commerce (blog). Regulation and limited ability to ACT throttle the value of bank data (ex. “intersections” which banking data can play). Additionally the wide availability of proxy data (w/ no restrictions on use) limit opinions for participation of “quality” bank data.  Recently the top banks have pulled their data from all marketing and “monetization” efforts (revenue did not justify the risks of compliance). Today most Card Linked Offer companies can no longer target on individual behavior based upon bank data. New entrants like DOSH use mobile location data to target and deliver a mobile first experience for any card. 

Other notable Bank data news over last 12 months

Retailer Data

Retailers like Amazon, Target and Kroger have become new forces in data. Retailer’s key advantage? The ability to act.  Data science is no longer the realm of specialists, all major companies now treat it as core. Retailers have become data and advertising giants (Nov 11 WSJ). Amazon’s has become a solid #3 with a 51% CAGR and 2020 revenue almost $13B

This is an amazing stat.. Amazon’s ad revenue is now the size of Google’s in 2011.. For a division less than 3 yrs old. 

Retailers recognize they are the players best suited to ACT within their own virtuous cycle. Google, Facebook and the agencies are their primary partners. One of their principle advantages (currently) is the degree to which they escape regulatory scrutiny. If regulators knew top 20 retailers send real time sku level purchase information to Google and FB .. I believe they would throw a gasket (as would consumers). What make this possible? Companies like LiveRamp, Epsilon, Neustar and others have mapped cards to consumer IDs. There is no longer purchase anonymity (see blog). 

Last October, Commerce Signals 80% analysis showed of digital advertising fails to produce value (excluding Google and Facebook). No wonder the agency business is failing and targeted television advertising is taking off (see WSJ Article). Retailers must take the lead in data as their partners have proven to be “wanting”. 

Part 2 coming.. Tomorrow

4 thoughts on “Data Games – Battle of The Cloud Part 6”

Leave a Reply

Your email address will not be published. Required fields are marked *