Panagiotis Kriaris’ Post

It was a matter of time before Amazon decided to compete head-to-head with Temu and Shein by…copying them. Let’s take a look. Temu and Shein have been getting the world of #ecommerce by storm following a similar #strategy: 1. Low prices and a vast assortment of goods 2. A direct-fast-to-consumer approach without any middlemen 3. Direct shipping from warehouses mainly in China 4. An aggressive marketing model with a huge spend that is dependent on discounts and coupons. Temu, for example, spends $10 mn every day on Meta and Google advertising! (source: Tencent) 5. Use of #AI to identify trends on social media and run a "real-time retail" business (Shein has 600,000 items on its web-site and needs 3 days to produce new designs) 6. Customized (algorithm-based) product recommendations 7. A gamification approach driven by loyalty programs and flash sales Both companies are of Chinese origin with Shein being a fashion retailer, whereas Temu has a much broader assortment of goods. Amazon on the other hand has followed a very different model: 1. Fast (often same day) delivery 2. Easy returns 3. Own warehouses (Fulfillment by Amazon model where goods are sent to Amazon warehouses before they are dispatched to customers) 4. Build customer trust 5. Excellent customer service Now Amazon wants to beat Temu and Shein in their own game via: —  Applying the same direct model based on low prices and no middlemen —  A new section on its site dedicated to fashion, lifestyle and household products, many unbranded —  Shipping directly from China with a 9–11-day delivery —  Employing an on-demand manufacturing model with limited, quantities that increase based on demand Why is Amazon doing this? Because it clearly understands: 1) the increasing threat that these new players pose to its dominance 2) the advantages of the Temu-Shein business model: —  Rock-bottom prices —  The opportunity to offer a much broader assortment of goods via thousands of merchants —  A much faster-to-market approach that can quickly test and react to changing consumer trends Amazon’s challenges: —  Amazon currently operates an entirely different model and it will not be easy to shift gear, even if the new approach is a partial repositioning —  Temu and Shein have developed an extremely sophisticated set-up based on technology (AI), #data and the use of #socialmedia that is not easy to replicate —  Building direct relationships to thousands of merchants in China is a massive exercise —  The on-demand manufacturing model is based on (hard to build) trust —  The danger of Amazon cannibalizing its own sales and profitability by shifting existing customers to lower-priced products is real TikTok’s rise at the expense of Facebook has taught Amazon a few lessons. However, Amazon managing to beat Temu and Shein in their own game is a tall order. Opinions: my own, Graphic sources: Apptunix, Momentum Works, SketchBubble

Banking has dramatically changed over the past years, but have banks adapted? And if not, where are the opportunities? Let’s take a look. The biggest change has to do with the environment that banks operate in: —    Financial services (FS) have moved from a closed-loop set-up controlled by a few players to an open ecosystem where data and APIs are the new building blocks —    Focus has shifted from 1) products to experiences 2) infrastructure to the customer relationship —    Modern banking has moved outside of the traditional value chain and is increasingly delivered across industries and verticals that have nothing to do with #finance —    Client expectations have been re-shaped beyond industry boundaries with customers benchmarking FS against the BigTech customer experience To be able to successfully navigate to the next phase banks need to build capabilities that will allow them to compete in this new open landscape. In other words: Banks need to move away from commoditization. Which in many instances means re-inventing themselves. How? By: 1. Refocusing on their core assets and unique strengths (you cannot be everything to everybody) 2. Finding their place in the modern (and more complex) value chain 3. Expanding their reach and finding new ways of generating revenue: value-added services beyond #banking Here are a few examples who have done it: 1) BBVA launched their Open Platform in the US in 2018. It’s an API-based, white-label BaaS (Banking-as-a-Service) solution that allows third parties (i.e. fintechs) to embed BBVA banking and payments capabilities directly in their check-out. Essentially a huge business multiplier for BBVA. 2) Via its Developer Portal ING provides APIs that enable third-party developers to create applications that can interact with ING's banking services. This feeds into the ING ecosystem where novel offerings can be built, integrated, and offered to ING's customers. 3) Kazakhstan’s Kaspi bank transformed itself from a traditional bank to a digital ecosystem via a SuperApp that integrates banking, payments, and #ecommerce services across 400,000 merchants and with 65% of Kazakhstan’s population as monthly active users. A logistics’ network and SME advertising services are good examples of beyond-banking monetization. 4) Singapore’s DBS Bank has launched a property marketplace that integrates banking with real estate and a car marketplace that brings financing and insurance to car buyers. Its DBS PayLah! mobile app goes beyond #payments integrating food delivery, ride-hailing, and ticket booking. 5) Barclays became one of the first major banks in America to offer POS financing via integrating #fintech platform player Amount in 2021. The move allowed it to expand its reach and tap into the huge BNPL market. It’s not the lack of opportunity that holds (some) banks behind, but the reluctance to change their mentality. Opinions: my own, Graphic source: Software AG, Matthias Biehl

If you want to understand what’s happening in #GenAI, these are the top trends – based on CB Insights: 1. Power-hungry AI data centers —    Massive demand for high-power AI #data centers: 80% of data center power to be consumed by AI over the next 15 years —    AI data centers will drive up demand for 1) cooling technologies and 2) renewable energy —    Nuclear may be the future 2. Data scarcity —    We are running out (by 2026) of high-quality data to train LLMs —    Vendors with proprietary content to become top acquisition targets —    Cost and scarcity to drive up demand for synthetic datasets 3. GPU shortage —    GPU shortage: H100s sell at a massive mark-up —    Big tech is the biggest buyer —    Nvidia’s biggest customers (Amazon, Google, Microsoft, Meta) becoming chip rivals 4. Security —    Top LLMs face security challenges (i.e. confidential data leaks, data poisoning) —    New vendors are emerging to tackle LLM security 5. Machine Learning —    ML operations is fragmented: 130+ companies across 12 categories —    Buyers want a one-stop shop —    End-to-end platforms to dominate (via acquisitions) 6. Multimodal #AI is the new trend —    AI that can process multiple data inputs (i.e. images, video, text) for better outputs —    In its infancy but quickly evolving —    More LLM developers will follow suit to remain competitive —    To unlock new commercial opportunities – Google in the lead 7. Smaller AI models are an opportunity —    Expensive large models are not always suitable for small tasks: using GPT-4 to summarize an email is like getting a Lamborghini to deliver a pizza (Tom Dotan, WSJ) —    Models are now shrinking, making them faster to train and cheaper to run —    Sectors with sensitive data will choose small models over LLMs —    Small models will take over narrow tasks across finance, healthcare, law 8. Open-source on the rise —    Open-source LLM developers are seeing rising investor & commercial interest —    A record number of AI projects will be built on open-source models in 2024 —    Smaller open models are fast closing performance gaps vs closed LLMs 9. New architectures challenge the status quo —    Google’s research on transformers was a watershed moment in GenAI —    But: transformer-based models have a massive carbon footprint —    Users want higher token limits —    Top research firms going beyond transformers —    Challengers will gain commercial traction in 2024 10. The next iPhone —    Various players are searching for the next AI-powered iPhone —    Industry heavyweights are betting on AI-native hardware —    LLMs evolve from chatbots to operating systems 11. Autonomous agents —    LLM-powered bots that can independently reason and execute tasks —    Getting better but far from commercial launch 12. Co-pilots —    AI will trigger the next productivity boom —    Coding copilots will become standard-issue tools at enterprises Summary: Panagiotis Kriaris, Source: CB Insights

Despite the big headlines around Apple’s Pay Later exit, the truth is that not only is Apple NOT exiting but it is rather doubling down on #BNPL. Let me explain. A few days ago, Apple did announce discontinuing its Buy-Now-Pay-Later (BNPL) service. As a reminder this is: —    US offering launched in March 2023 —    Split purchases functionality into 4 interest-free installments spread over six weeks —    Loans of $50 to $1,000 for online and in-app purchases via Apple Pay —    Seamless Apple Wallet integration (all BNPL loans in one place) —    Enabled through the Mastercard Installments program with Goldman Sachs behind However, the US shutdown is, in effect, only a #strategy change that will allow Apple to significantly 1) improve, 2) streamline and 3) scale its offering in ways that the current set-up didn’t allow it to do. How will it do this? Via partners: —    in the U.S. with Affirm, Citi, Synchrony, Fiserv issuers —    in Australia with ANZ —    in Germany with Santander (Zinia) —    in Spain with CaixaBank —    in the U.K. with HSBC and Monzo In effect Apple is doing away with a model of directly managing (most) aspects of the #business and turning to the experts to do it for them, while Apple retains control. A few (good) reasons are behind the shift: 1. The US launch was always meant to act as a testbed on what works and what doesn’t. The US model was unsustainable - below is why. 2. Apple will remain fully in charge of the check-out experience and of the customer relationship with Apple Pay in the front and funding via BNPL in the background. 3. The relationship with Goldman Sachs, who was behind the set-up together with Mastercard, is being anyhow terminated. 4. Apple does not have any expertise in the business: BNPL might be a simple concept, but the set-up behind (product, #technology, underwriting, collections) can be quite complex. It now taps onto a global pool of experts that bring flexibility, size, regional expertise and speed-to-market. And Apple can always replace them.  5. Regulatory scrutiny for BNPL is tightening around the globe, affecting both costs and reporting. Managing these is not attractive. 6. Scaling the current model would require Apple to deal with operational, licensing and regulatory implications in each and every jurisdiction. 7. BNPL is, in effect, unsecured lending with strict capital and risk provisioning requirements. Balance-sheet appetite is not there for Apple. 8. The high cost of funds combined with no interest and fees mean that the current business model might have been loss-making. In most BNPL models retailers are paying the bill, but in Apple’s model they are not in the equation. BNPL is a huge fit for Apple. By taking a step back Apple can focus on what it does best, leaving the complicated job to others while it continues to strengthen its ecosystem. Knowing when (and how) to pivot is crucial. Opinions: my own, Graphic sources: Apple, WhiteSight, Panagiotis Kriaris

Germany’s attempt to build a payment champion has come to an end. What are the reasons, implications and lessons learnt? Let’s take a look. Back in 2005 a consortium of German banks decided to launch an online payment method called Giropay, which would be: —    integrated in the online banking of participating banks —    able to make #payments (direct online transfers) from bank accounts —    tailored for e-commerce (i.e. buyer protection, no sharing of data with merchants) In debt-averse Germany credit cards were never popular.  Which explains why the most successful payment methods always involved finding direct ways: —    PayPal, which allows you to mask your account or debit card behind —    Payment by invoice, which involves paying via credit transfer after receiving an invoice —    Sofort, which involved screen scraping (banking customers sharing their login credentials with 3rd parties) —    Direct debit In 2021, Giropay was acquired by Paydirekt in order to bring 3 payment methods under one roof (the Giropay brand): —    Giropay —    Paydirekt —    Kwitt (a P2P method) The idea was always to replicate the success of the girocard, Germany’s most widely used offline payment method (card) to the online world and compete with PayPal (almost 30% online market share). In practice, Giropay never managed to take off, staying at irrelevant levels (market share below 2%), which last week marked the banks’ decision to discontinue it. Several reasons are behind: 1. User experience not as easy and straightforward compared with PayPal or Klarna:  Giropay requires online #banking with a participating bank and has a more complicated checkout process. 2. Limited merchant adoption because of the complex and rather expensive structure to maintain behind. 3. Despite the high ambitions, Giropay was always one step behind the market. Even the announced discontinuation of Sofort as a stand-alone payment method and the consolidation to the Klarna App was an opportunity that Giropay didn’t succeed in taking advantage of. 4. The participating banks behind Giropay never managed to articulate a solid, fast-moving decision-making structure emulating iDEAL in the Netherlands or Twint in Switzerland. 5. Immediate rivals like Sofort managed to more successfully address key market elements under one bundle: user experience, pricing, Go-To-Market and market coverage. 6. Bad timing: Giropay bet on #ecommerce when e-commerce was still in its infancy (especially in Germany) and at the same time it failed to capitalize on the success of the Girocard. What happens next? The European Payments Initiative (EPI) wants to do with wero what Giropay failed to do not only in Germany but across Europe. Germany is one of the first and key markets. Had Giropay been successful in its #strategy, it could have been the preferred acquisition choice in place or next to the Dutsch iDEAL. Opinions: my own, Graphic sources: PayZen, Giropay, Heidenheimer Volksbank

Europe #Money2020 was a blast! Here is my selection of key topics and main takeaways: 1. Open Banking: —    Main challenge is how to get to the next level (adoption, breadth) —    Need to learn from card schemes —    Open Finance already under way with data at the center —    Different regional approaches: interoperability and regulations —    Commercial VRPs as a killer use case —    Combination with GenAI 2. #Ecommerce: —    Explosive growth in the gig and circular economy with 200+ mn new creators and sellers in 3 years —    It’s all about mass-personalization (big data, AI) —    Digital identity, AI and biometrics to address security and consumer trust —    LLMs for high-quality, multi-channel data on verticals 3. Artificial Intelligence (#AI): —    AI to enhance rather replace humans —    Predictive AI use cases already present, Gen AI on the rise —    It’s all about data and feeding the models —    GenAI is a game changer. Implementation, integration & scalability the key topics —    Europe needs to catch up vs Asia and US (i.e. Mistral AI, EU AI act) —    Intelligence at scale needs investment in AI factories 4. Digital Money: —    CBDCs and stablecoins to coexist rather than compete —    Adoption, integration and interoperability are key topics for CBDCs —    Swift as a blueprint for integrating CBDCs into existing FS infrastructure —    Public trust is a major success driver (fight skepticism and conspiracy theories) 5. #Payments: —    Real-time infrastructure is the name of the game —    Fraud a major challenge - global anti-fraud investment at $10 tr by 2025 —    Europe to compete beyond rails (Digital Euro, EPI with wero) - lots of challenges to solve (i.e. added value, adoption) —    Seamless, unified, personalized experiences that add value —    Embedded finance —    Subscription payments —    B2B models (i.e. B2B Pay Later) will win —    Digital wallets are critical infrastructure - adoption an opportunity for the EU 6. Personalization: —    One of the biggest cross-industry trends —    It’s all about mass-personalization —    Real-time (big) data analytics, cloud infrastructure and AI as key enablers —    Payments in the driving seat 7. Banking: —    Legacy systems and digital transformation are main challenges for incumbents —    Banks already rolling out GenAI use cases - 78% plan to increase AI investment —    Strategic partnerships as a key innovation enabler (i.e. chatbots) —    The biggest challenge to innovation is execution —    Sustainable financing at $2.6 tr in 2023 - projected to grow further 8. Blockchain: —    An alternative for traditional FS (speed, security, scalability) —    Ripple and Kraken examples of the evolution —    Robust infrastructure and system interoperability are key —    Regulation, data privacy and maturity of cross-chain use cases as main challenges —    Crypto becoming a legitimate asset class (i.e. ETH ETFs in the US) Analysis: Panagiotis Kriaris, Source: Money2020

Challenger banks need to pivot but their 2.0. play goes through lending, which depends on getting a #banking license. Can they compete? Let’s take a look. Branchless, digital-only, mobile-first banks were the first fintechs to challenge traditional banking via focusing on: —   gaps in niche segments —   seamless, end-to-end digital experiences Penetrating non-transparent, complex, expensive and inelastic products of incumbents built on top of outdated systems was an easy game in the early 2010s. Most players did so on the back of a growth-first strategy with ever-increasing valuations funding customer acquisition. However, the macro reality has turned the game on its head. To be able to make it to the next stage digital banks today face two main challenges: —    Shift focus from growth to profitability —    Balance increasing regulatory scrutiny and compliance requirements And while there is no one-size-fits-all playbook these are a few guidelines: —    Transition from a convenience, niche-based set-up to becoming the primary bank (get the payroll and build long-term relationships) —    Build trust (essential to become the primary bank) —    Diversify and expand revenue build-up beyond interchange, FX and subscription The last point is crucial. And it comes down to one thing: lending. Which requires a banking license. Here it's important to get one thing right: neobanks, challenger banks or digital banks are terms used interchangeably but they don’t always mean the same thing. Besides the #technology aspect, the main difference lies in the licensing piece: —   Most of these banks rely on what we call an EMI (electronic money institution) license. This is not a banking license but the next best thing. It is faster to get, more flexible and needs much less capital but has one big downside: EMIs cannot take deposits and lend them out again. —   During their 1.0. play an EMI license was enough for #fintech banks. It allowed them, among others, to operate payment accounts, deposit and withdraw money from them, acquire transactions, remit money, issue and manage e-money. However, it’s not enough to deploy their 2.0. play, which is why the biggest of these players (i.e. Revolut, Bunq) have made acquiring one a central part of their #strategy. As the Revolut saga in the UK has shown getting a banking license can be a chicken-and-egg game: getting the regulator’s approval is dependent on a strong compliance infrastructure that was never the focus or the strong point of these players. Which translates back to pivoting and building a sustainable business with more long-term KPIs than the ones that used to grow. In entering their 2.0 play, challenger banks are facing the odd situation of having to balance an #innovation, growth-first agile culture with more traditional ways of doing banking. The oxymoron? Challenging these is in their DNA.   Opinions: my own, Graphic sources: accubits, Consultancy Asia, P. Kriaris

#payments rails across the globe and the models behind them have evolved in three major (but very different) patterns and yet they are converging in certain ways. Let’s take a look. About half a century ago, magnetic-striped cards triggered a payments revolution. Swiping plastic cards at POS merchant terminals conquered the west, with Visa and Mastercard managing the rails and becoming an almost mighty duopoly. Cards made a smooth transition into the digitized #economy by embedding in smartphones (and even turning them into processors) and becoming the springboard for the rise of the #ecommerce. While the west was transitioning from old cards to chips, China was driving its own local payments revolution that erupted at the beginning of the 2000s and transformed the country from a purely cash economy to a #digital frontrunner. Starting from high smartphone penetration and bank account ownership, China essentially leapfrogged the card-based (western) model moving directly to a digital set-up built on e-wallets and QR codes and driven by two private companies (Alibaba and Tencent) that managed to build vast (2-sided consumer and merchant) ecosystems that transformed them into ubiquitous SuperApps. In parallel, a third pole had been developing in other parts of the world: —     The payments revolution in Africa was led by telecoms (being the only infrastructure available) by means of an e-#money set-up based on mobile phones. Companies such as Kenya’s M-Pesa (launched in 2007) managed to provide long needed basic financial services (saving and transferring funds, making payments or accepting government subsidies) to large swaths of the population. —     Countries like India or Brazil developed over the past few years state-sponsored real-time payments infrastructures, powering multiple bank accounts into a single app under A2A and P2P models. India’s Unified Payments Interface (UPI) has about 300 mn monthly active users holding 84% of the total e-payment transaction volumes (!), whereas Brazil’s Pix, launched only in late 2020, has managed to become the most popular payments’ method in the country with over 150 mn users. These parallel evolutionary developments could hardly have been more different: a robust decades-old, card-infrastructure in the west (monopolized by two private companies), against a digital, wallet-based closed-loop model in China (powered by 2 giant private ecosystems), versus public, state-sponsored, open, real-time rails in India and Brazil. Despite their very different origins and set-up, digitization has been acting as a huge convergence driver lately: digital wallets, super-apps, real-time payments and CBDCs (Central Bank Digital Currencies) are only some of the common underlying elements. As payments evolve to their next phase, a new digital infrastructure is in the making, fast bridging seemingly big structural gaps. Opinions: my own, Graphic sources: Credit Suisse, Alipay, Matthew Brenan, BCB, Bacancy, Alicriti

JP Morgan has been slowly but steadily doing what many traditional banks have long postponed: ride the marketplaces and platforms' wave. Let’s take a look. Financial Services (FS) are becoming a platform business. One that is shifting away from traditional #banking into a platform set-up that heavily evolves around #payments. Which means: Customers are increasingly making finance decisions in non-FS environments. And somebody needs to enable the infrastructure behind this shift, i.e. to power marketplaces to seamlessly integrate FS in their flows. It’s exactly what JP Morgan has been building. How? As early as 2019 JP Morgan was targeting big #ecommerce and gig economy players (i.e. Airbnb, Lyft, Uber Amazon, eBay) by building an e-wallet allowing these companies to offer their customers virtual bank accounts. Fast forward to today, a few examples of how this business has evolved: —    JP Morgan is behind Amazon Seller Wallet, which enables merchants manage their funds (from Amazon e-commerce sales), make payments to partners (i.e. vendors, contractors, suppliers) and transfers to their bank accounts. —    JP Morgan powers SalonCentric, L’Oréal’s B2B marketplace for professional salon products by offering embedded #banking and core treasury services in the background. —    JP Morgan helped US traditional brick-and-mortar chain-store Macy’s move online and develop an e-marketplace by addressing all issues around payments (i.e. management of funds, tokenization to mask sensitive information, third-party money complexities). Essentially JP Morgan is helping these companies to compete with banks, including JP Morgan itself. So, where’s the hack? —    JP Morgan understood early on the opportunity to sacrifice some of its traditional banking revenues for a much larger slice of the pie. —    JP Morgan has built the infrastructure that is powering modern platforms and marketplaces. It offers these players a seamless, state-of-the-art offering that enables them to keep the payment flows within their platforms and in exchange it handles all of their payment processing needs. The model bundles a few key functionalities together: 1. A white-label approach that helps marketplaces to retain and control their relationships with sellers 2. Digital onboarding 3. Pay-outs 4. Settlement 5. Management of fund flows for sellers (marketplaces don’t want to do this) 6. Value added services: i.e. bank accounts, cash management, trade finance 7.Insights JP Morgan has chosen to compete in a market that enables the players of the new #economy (e-commerce platforms, marketplaces, gig providers) to pursue their ecosystem play. It’s not alone, fintechs like Stripe or Adyen have been very successful in this game. The difference is that JP Morgan is a bank. And it wants to leverage it's infrastructure to play across the entire value chain (banking, payments, merchant acquiring). A clever bet. Opinions: my own, Graphic source: JP Morgan

It’s an oxymoron but digital ecosystems are banks’ biggest threat and at the same time one of their biggest opportunities to adapt and thrive. Let’s take a look. There are four main changes driving banks’ ecosystem strategic dilemma: —    Rapidly evolving customer preferences and the massive digital shift of consumers and businesses alike. —    The replacement of the traditional vertically integrated #banking model by an approach where various players co-exist under different layers. —    The blurring of the boundaries between industries and the surface of new roles in the value chain (i.e. marketplace orchestrators, third party ecosystem participants, open banking platforms, etc). —    Τhe shift from infrastructure ownership to control of the customer relationship as the single most important success factor. For decades it was the back-end that was defining the competitive landscape, however in today’s open economy it is the front-end that carries all the weight. Within this landscape the challenge for banks is two-fold: —    How to deliver added value in a network set-up of different layers, where banks don’t directly control all of them. —    How to turn customer interactions into personalized experiences that will drive new revenue streams.   Is it a choice for banks to adapt to this new reality? Actually not. However, there is a choice, and it has to do with the how. Unfortunately, there is no one-size-fits-all playbook, but there are some guidelines: —    It’s critical that banks get their API strategy right. Ecosystems are about interconnected products and services and APIs are the glue that brings them together. —    Banks need to be able to integrate third-party and #fintech offerings (via APIs). The commercial model is open (who is the platform behind), but the capability is a must. —    In the platform #economy mass personalization is the new name of the game and it goes through being able to combine data with new technologies like #AI. —    Understanding and navigating the opportunities arising from embedded finance, either from a front-end (delivery) perspective or from a back-end BaaS (Banking-as-a-Service) angle. —    An omni-channel first #strategy that focuses on converting customer touchpoints across the entire customer journey to business opportunities. —    Readiness to bundle their core offerings, where needed, with external capabilities as part of a unified customer proposition, which very often will extend beyond banking. —    Build on their strengths (i.e. trust, local market understanding) as a means to compete and defend their turf. —    Choose a few key strategic verticals and build, expand and monetize added-value offerings around an unparalleled level of expertise. The more effective banks are in crafting a coherent ecosystem strategy, the better they can position themselves in a digital value chain that is becoming increasingly complex to navigate. Opinions: my own, Graphic source: BCG