The recent Digital Maturity to AI Mastery  webinar , hosted by Commerce in partnership with The Engineer , offered valuable insights into how Agentic Commerce is reshaping the eCommerce landscape. One of the key takeaways, shared by Rupert Chappell, Key Account Director at Commerce , was how quickly product discovery is shifting from human-led search to AI-assisted journeys. A couple of stats that caught our attention with implications for retailers, particularly those in B2B, are significant: B2B online sales are expected to double in the next five years Over half 58% product discovery now begins on AI tools As Commerce highlighted, digital maturity is directly aligned to success in this brave, new world. With digital maturity defined as the alignment of technology with business goals, regular review of goals and ability to change. That message mirrors what we’ve discovered in recent client research at thinkTRIBE. Digital Transformation remains a central theme across all sectors, but the focus is evolving: from adopting new technology to ensuring digital strategy, business outcomes, and customer experience are fully aligned. It also builds on themes we’ve been exploring through our Agentic Commerce Series  - looking at how AI is redefining product discovery and purchase, and what this means for retailers preparing to embrace this next shift. 1. Digital maturity and change capability drive resilience and growth Commerce’s framework for assessing digital maturity covers six key pillars: digital objectives, tech stack, customer centricity, data, logistics and fulfilment, and culture and innovation. The results were striking: businesses at the lower end of the maturity scale were twice as likely to experience revenue decline compared to leaders. For retailers, digital maturity isn’t just about having the right tools. It’s about how well digital objectives align with business goals, and how often those goals are reviewed, refined, and adapted. As AI adoption accelerates, digital change adaptability is becoming key success. For B2B retailers, that capability is especially critical. Buyers now expect the right products faster than ever before, and their CX expectations increasingly mirror those of B2C. We explored this further in The customer experience gap in B2B eCommerce and how to close it , which highlights how B2B commerce can bridge the CX gap. Retailers that break down silos, align digital and commercial objectives, and treat technology change as a continuous capability, not a one-off project, will be best positioned to thrive as AI reshapes discovery and conversion. 2. Data is becoming your storefront As AI channels take on more of the search and discovery process, product data intelligence is emerging as the new battleground . Commerce put it succinctly: “Data will become your new storefront. That shift means retailers must ensure product data is not only visible but also verified, explainable, easily understood and optimised for each AI channel. Think of AI as your biggest referral partner. Retailers should ask: Are our products discoverable and ingestible by AI? Is our data accurate, contextual, and backed by evidence? Can AI access the right information, from reviews to specs, to make reliable recommendations? For B2B retailers, this is particularly vital. Complex pricing, contract visibility, and multi-tier product configurations make clean, structured data essential to maintaining both trust and conversion. Rather than relying on scraping, retailers should take control of how their data is surfaced, using data syndication platforms to ensure product information is consistent and AI-ready across every channel. Platforms such as Feedonomics, Commerce’s product data syndication solution , are already helping retailers optimise and distribute data more effectively across AI channels. These will become increasingly valuable for ensuring product information is consistent, discoverable, and explainable across every surface. 3. Focus AI on friction, not novelty One of the most practical messages from the webinar was clear: use AI to remove friction, not add it. For B2B retailers, this might mean: Enabling real-time pricing or stock visibility for trade accounts Streamlining repeat orders Automating “where’s my order?” or quote requests AI shouldn’t replace human interaction, it should enhance ease, speed, and add transparency. The same principle sits at the heart of great customer experience in both B2B and B2C. 4. Preparing for Agentic Commerce As we explored in our Agentic Commerce series, the rise of Agentic Commerce - where AI agents transact directly, isn’t mainstream yet, but the groundwork is being laid. We’ve already moved from buyer → retailer to agent → retailer, and agent → agent is not far away. Retailers that: Assess their digital maturity and align strategy with business goals Structure and optimise product data for AI discoverability Use AI to improve business workflows and enhance CX, not complicate it …will be best positioned to move from digital maturity to true AI readiness. Because as discovery shifts to AI, success won’t hinge on technology alone, it will depend on how well your business can adapt, align and deliver for both human and machine customers. Final thought: Treat Digital Transformation as an agile process AI adoption is accelerating, but readiness is inconsistent. Retailers that treat Digital Transformation as an agile process, continually assessing digital maturity, adapting digital objectives, refining data, and focusing on AI to improve business processes and CX, will lead the next phase of the eCommerce evolution. As Commerce summed it up: “Data will become your storefront.” The question now is how well your business is preparing to open the door? Key takeaways for retailers To prepare for AI-led discovery and the next phase of eCommerce: Measure digital maturity  across objectives, technology, data, CX and culture. Align digital and business goals  to drive agility and growth. Develop product data intelligence  to ensure AI can find, understand and ingest evidence in order to trust your product recommendations. Use AI to remove friction, not add it - take small steps focussing on specific workflows with measurable improvements. Wherever you are on your AI readiness journey, thinkTRIBE helps bridge the gap between digital transformation and real-world customer experience.

In SaaS and retail, frequent releases are now standard. Teams often deliver updates weekly or even multiple times a week. This pace helps meet customer expectations and stay competitive. As Scrum.org notes, how often you release is one of the clearest measures of true agility . However, every release brings its own pressures. There’s temporary downtime, noisy KPIs, false alerts, and the risk of new issues damaging customer experience (CX). So, how can we keep release cycles smoother and faster? Clean reporting helps, but the bigger challenge is finding methods and tools to manage release cycles effectively and fix real issues quickly. Why Frequent Releases Feel Harder Than They Should Every new release generates friction. Tight timelines often lead to technical debt. Rushed development cycles and insufficient testing are key contributors to technical debt . This makes managing releases more challenging. Skewed KPIs: Reports show errors and impacted availability, even when downtime was planned. False Alerts: DevOps teams waste time chasing noise. Hidden Regressions: Genuine CX problems and bugs get lost in the clutter. The result? Release cycles become harder, not faster. Release Monitoring to Protect CX and Performance Making release cycles easier to manage starts with visibility across the release process: Before Releases Measure baseline CX across journeys. This helps you know what “good” looks like. During Releases Accurately exclude the time between the end of one release and the start of the next. This ensures KPIs reflect reality. After Releases Easily validate updated journey CX and highlight regressions immediately. This approach to release monitoring removes unnecessary noise. It helps teams focus on what matters most: keeping releases smooth and resolving issues quickly. Shift Left: Catching Issues Earlier Easing the pressure around release cycles isn’t just about what happens in production. By shifting monitoring left into the staging environment, teams can test real customer journeys before code is deployed live. This helps spot potential regressions earlier, reduces surprises on release day, and gives DevOps teams more confidence that new features will perform as expected. How thinkTRIBE Helps Throughout the Release Cycle thinkTRIBE provides a complete release monitoring toolkit throughout the release lifecycle. From staging to production, our tools are designed to reduce friction in release management. Each feature available with our real-world CX and performance monitoring aims to help teams cut through noise and focus on fixing genuine issues faster. 1. Monitor Staging Environments (Shift Left) Monitor customer journeys during staging to uncover potential regressions before code is deployed to production. 2. Planned Maintenance Exclusions (PME) Exclude planned maintenance windows for new releases. This ensures downtime doesn’t impact KPIs or create misleading errors. 3. PME API Connect seamlessly with CI/CD pipelines and release scheduling tools like ServiceNow or Jira. This way, planned downtime is excluded automatically and accurately. Reports stay clean, and teams aren’t slowed down by false alerts. 4. AI-Enhanced Self-Healing Journeys Automatically adapt to small changes during minor releases . This eliminates false alerts and wasted effort, helping teams avoid distractions and focus on genuine regressions that impact CX. 5. Site Release Manager Receive scheduled reports at 3h, 6h, 12h, 24h, 3d, and 1w after release. These reports are benchmarked against pre-release performance, allowing you to quickly pinpoint what broke, where, and when. Why Simplifying Releases Matters Frequent releases will always carry risks. However, they don’t have to create friction and frustration. The real challenge is making release cycles easier to manage and faster to resolve when issues arise. By combining Site Release Manager, PMEs, and PME API, teams can ship with confidence. This leads to smoother releases, faster fixes, and CX kept firmly in focus. Conclusion In the fast-paced world of SaaS and retail, managing release cycles is crucial. By adopting effective monitoring and management strategies, we can enhance our ability to deliver quality updates without compromising on performance. If you want to learn more about how thinkTRIBE supports smoother, faster releases, request a demo today. Let’s work together to protect your revenue and brand reputation by proactively identifying and fixing critical digital customer experience issues.

Every January, gyms experience their busiest period of the year. New Year’s resolutions bring a flood of new members, making January the biggest month for sign-ups. In the UK, January is 28% busier than December , according to 4GLOBAL’s DataHub via ClubRight . It's a huge opportunity, but also the time when your gym's digital systems face the most pressure . When user journeys slow down, throw errors, or break under peak demand, the impact is immediate: abandoned sign-ups, lost revenue, and lost trust. If your website or apps slow down, throw errors, or break under peak demand, the impact is immediate: abandoned sign-ups, lost revenue, and damaged trust. And with heavy marketing spend driving traffic in January, any glitch means wasted investment. It’s Not Just Downtime That Matters When most people think of website risk, they think of outages. But for gyms, as with many sectors, it’s often the smaller cracks in the digital journey that cause the biggest problems and often go unnoticed until the damage is done. Slow load times → abandoned sign-ups Broken journeys (join, upgrade) → lost members Payment failures or timeouts → members unable to complete payment process Promo codes not applying correctly → drop-offs at checkout App access or class bookings stalling → frustrated members who give up Confusing error messages → wasted ad spend and brand damage Even if your site stays online, these hidden failures can stop your January campaigns from delivering results. Why Gyms Need to Test Smarter Traditional load testing only tells you whether a site can handle X number of visitors. But gyms don’t just have “visitors.” They have people trying to join, pay, upgrade memberships, book classes, and check in via apps — all at the same time. That’s why realistic load testing is so important. By simulating real member journeys under real-world traffic conditions, you see how your systems perform under true peak demand. It’s not just about keeping the lights on — it’s about protecting the customer experience. Don’t Stop at Testing — Monitor All Year Round Load testing is vital preparation, but it’s only the first step. Once January is underway, it’s just as important to monitor key customer journeys like the join journey continuously, 24/7. That way, you catch problems as soon as they appear, whether it’s in January or later in the year when new campaigns or offers go live. Ongoing monitoring keeps your join journey fast and seamless for every new member. Practical Ways to Prepare Your Gym Website for January Peak Sign-Ups Based on what we’ve seen work with major gym operators, we’ve created a simple checklist of five top tips to get your gym’s website and apps ready for January’s peak. Our work with The Gym Group, for example, improved their site capacity by 25% through managed testing. 1. Test All Layers of Your Infrastructure Make sure to test every layer of your infrastructure. This includes your servers, databases, and application layers. Each component must work seamlessly together to handle the influx of new members. 2. Include Third-Party Integrations Don't forget about third-party integrations. These can often be the weak link in your digital chain. Ensure they are tested thoroughly to avoid any disruptions during peak times. 3. Test Real Membership Scenarios Simulate real membership scenarios during testing. This helps identify potential issues that new members might face when signing up, upgrading, or booking classes. 4. Overlooked Areas: App Access and Class Bookings Pay attention to overlooked areas like app access and class bookings. These are critical for member satisfaction and retention. Ensure they function smoothly under high traffic. 5. Continuous Monitoring is Essential Finally, remember that monitoring shouldn’t stop after January. Continuous monitoring is essential to keep your join journey bulletproof all year round. Why Digital CX & Performance Matter For gyms, digital performance in January is critical: Safeguard sign-ups and revenue during your biggest month. Deliver the fast, seamless experience new members expect. Ensure your marketing spend converts into actual memberships. Free your developers from test complexity with expert support. Keep your join journey bulletproof all year round with continuous monitoring. Want to make sure your gym website is ready for January peak sign-ups? Get our top tips and see how to make January your biggest win, not your biggest risk. 👉 Download our top gym web peak preparation tips here Conclusion Preparing for January is not just about surviving the rush. It’s about thriving and making the most of the opportunities that come with it. By implementing these strategies, you can ensure your gym is ready for the influx of new members. Let’s make this January a month of growth and success!

Monitoring platforms include a planned maintenance exclusion (PME) feature to filter out expected downtime and errors during releases, keeping your KPIs clean. But for today's digital teams rolling out multiple updates, across dozens of journeys, a manual PME toggle isn’t always realistic. What’s needed is an API-driven PME, so exclusions are automated, effortless, and integrated directly into your release workflow. Here’s why. Automate exclusions to reduce risk and save time DevOps teams push updates fast, sometimes several times a week. Relying on manually setting up exclusion windows is time-consuming and means risking delays, mistakes, and frustration. With an API, DevOps pipelines can automatically open or close PME windows as part of the release process. No more chasing through lists of journeys. No more repetitive clicks. Just automated downtime exclusions that free your team’s time and energy. Integrate with existing tools and workflows An API doesn’t just save time, it fits naturally into existing workflows. The PME API doesn’t create “another tool to manage”, it connects with the systems your teams already live in. CI/CD pipelines → automatically trigger PME windows in sync with deployments. IT Service Management Platforms ( ServiceNow , Jira Service Management , Freshservice )   → change requests, maintenance logs and service support sync automatically. Observability Platforms (Splunk, Datadog) → PME windows align with deployment logs, so teams can instantly see which alerts to trust and which to ignore. This means less admin, less switching between tools, and fewer frustrations. Maintenance and release windows are logged, synced, and excluded from KPIs without extra manual effort. Cut out human error and the grind Automated exclusions prevent common mistakes. Wrong times, wrong journeys, or missed windows lead to messy KPIs. But even when done right, the manual work slows everyone down. A PME API fixes both problems: no more mistakes, no more wasted time. Adjust in real time, not by rigid schedules Releases and maintenance don't always run to plan. With an API, exclusions are opened or closed automatically based on what’s actually happening with the release, rather than being tied to predefined maintenance windows. Keep KPIs accurate and noise-free Planned downtime shouldn’t appear as checkout errors, failed logins, or SLA breaches. An exclusion API ensures those periods are filtered out automatically in real time, directly tied to release logs: Maintain accurate uptime and availability KPIs Reduce false alerts and noise for your teams Gain a clear picture of CX once releases are live Keep SLA & CX reporting clean What this means across different sectors For SaaS providers Performance/CX reports and SLAs depend on clean, accurate data . Protect SLAs and CX, avoid false downtime penalties, and simplify compliance. For Retailers CX and conversion KPIs drive boardroom conversations. A PME API provides a tamper-proof record that downtime was excluded automatically and accurately, keeping reports transparent and credible. Keep conversion, checkout success, and CX scores accurate, even during updates. In both cases, the PME API ensures monitoring reflects CX and performance reality , not your release schedule. The bottom line on automated downtime exclusions Planned downtime is unavoidable. But false downtime and the manual grind of managing PME doesn’t have to be. An API for planned maintenance exclusions makes monitoring smarter, cleaner, and fully integrated into the way you already deploy software. With a PME API, exclusions are automated, integrated, and effortless. Teams save time, reduce frustration, and trust that KPIs reflect reality, not release schedules. Further Resources Discover how thinkTRIBE’s PME API makes releases smoother and smarter. Request a demo now.

Helping teams release faster and smarter by balancing speed, risk, and customer experience Speed is the promise of CI/CD: the ability to push updates and improvements into production continuously, without bottlenecks. But speed without safeguards can quickly undermine customer experience. A single poorly performing release risks slowing journeys, introducing errors, or eroding the trust that customers place in your site. That challenge has become more pressing as CI/CD has matured. Today, the debate isn’t about whether to adopt it, that’s long settled. The real issue is how to keep CI/CD pipelines both fast and reliable in an era of AI-assisted development, heightened security threats, and rising customer expectations. For many organisations, hundreds of micro-releases now flow through production every month. That pace enables rapid iteration, but it also raises the stakes: a single unnoticed regression can ripple into downtime, lost revenue, and reputational damage. The answer lies in shifting left on monitoring, building resilience checks earlier into the pipeline, and moving from continuous delivery towards continuous verification. By doing so, teams ensure that every release not only keeps pace with business demands, but also strengthens the customer experience. Why speed alone is no longer enough Continuous deployment promised to eliminate bottlenecks by pushing new features live as soon as they were ready, while continuous delivery focused on keeping releases deployable at all times. But speed on its own doesn’t guarantee value. In fact, without visibility and guardrails , it can introduce significant risk. Key pressures shaping today’s pipelines include Security: The software supply chain has become a target, with vulnerabilities often surfacing during automated builds. AI code generation: More lines of code, shipped faster, means more opportunities for subtle performance issues to slip through. Customer experience: Deployments are judged not by volume, but by their impact on usability, reliability, and speed. Shifting left on monitoring to protect CX The principle of “shift left” has long applied to testing , moving quality checks earlier in the cycle. But leading teams shift left on monitoring as well. We’ve seen this in practice with clients who embed monitoring directly into their CI/CD workflows, rather than waiting until production to see how code behaves. User journeys run in staging environments; some even mirror those journeys in production, comparing performance side by side. Monitoring embedded in CI/CD Pipeline This proactive approach creates a safety net: if a release shows degraded performance or unexpected behaviour during staging, it can be paused or refined before it ever reaches customers. How essential is staging in today's CI/CD pipelines? Traditionally, staging environments act as a safe space to catch issues safely, before customers are affected. But newer, more risk averse release techniques are reshaping the conversation. Feature flags  let teams toggle features on or off instantly, or show them only to a small group of users. Canary deployments  gradually release code to a small subset of users or servers first, monitoring behaviour before wider rollout. Because these approaches test directly in production, some argue that staging is becoming less central. After all, production is the truest measure of performance. Others point out that staging still provides vital protection. Not everything can safely be tested live — think payment flows or large, business-critical integrations. For many teams, staging remains an important buffer, while modern techniques add flexibility on top. Wherever testing happens, one thing is clear, real-user monitoring in production is essential . Even small rollouts can affect real customers, so visibility and control must be baked into the process. From continuous delivery to continuous verification CI/CD pipelines are less about throughput and more about outcomes, teams need to measure not just: Did it deploy? But also: Did it improve resilience? Did it enhance customer experience? Did it avoid introducing hidden costs downstream? This shift, sometimes called continuous verification , places monitoring at the heart of CI/CD strategy. It ensures that the push for speed is balanced by confidence in the customer experience. Final thought: Building safer, smarter CI/CD pipelines CI/CD has matured from a development methodology into a business-critical practice. The organisations who succeed today will be those who don’t just move fast but who build pipelines with the intelligence and visibility to move fast safely. Is shifting left on monitoring the foundation for delivering change at the pace customers expect? At thinkTRIBE, we help teams strengthen that foundation by monitoring at any stage of the production cycle , from staging to live journeys, so issues are identified early and every release is backed by data-driven confidence. What next? If you’d like to explore how monitoring can strengthen your CI/CD strategy, the thinkTRIBE team is always happy to share best practices and insights. Contact us now to request a demo .

Why autoscaling alone isn’t enough to protect CX at peak A leading high street jewellery retailer re-platformed to a composable storefront with autoscaling, designed to handle seasonal peaks. Rather than assuming the setup was bulletproof, they took the smart approach, commissioning journey based Load Testing to validate real-world performance ahead of their busiest trading periods. When autoscaling creates hidden risks The test delivered exactly what was needed: peace of mind and actionable insight. It revealed a hidden autoscaling configuration problem that triggered spikes in error rates, including HTTP 500 errors, missing content, and empty product listings, every time the site scaled. Once traffic reached around 3,000 concurrent users, these errors escalated and no longer recovered between scale events. Left undiscovered, this could have left customers unable to purchase products, facing empty product listings, failed product pages and failed product filters, during the busiest sales periods. BFCM risks you won’t uncover with conventional Load Testing Conventional Load Testing often stops at measuring server response times and error rates. While those metrics are valuable, they wouldn’t have revealed the real-world customer experience issues uncovered here, problems that directly impact conversion but don’t always trigger widespread errors. Real-world journey based Load Testing exposed subtle but critical issues such as: Empty product listing pages (PLPs) where the grid loaded but no products appeared Failed product filters that stayed stuck in a loading state Missing third-party content that never populated Test and Retest Because the problem was found early, the fix was made with no customer impact. A retest proved the change worked: the site scaled cleanly to all nodes and handled traffic well beyond peak levels, maintaining stable journey delivery times without sustained errors or degradation over time. The consistency of performance under heavy load was a marked improvement over their previous SAP Hybris platform. The takeaway for any retailer preparing for BFCM? Autoscaling is a valuable safety net but it’s no substitute for proactive, real-world Load Testing to uncover CX issues, protect customer journeys, and ensure your busiest trading days run without a hitch. Composable Commerce: One Concept, Two Playbooks “ Composable commerce ” is an approach where retailers build their eCommerce stack from best-in-class components, rather than relying on a single monolithic platform. Gartner’s approach: Emphasises packaged business capabilities (PBCs) — self-contained features (like search, payments, CMS) that can be swapped in or out as needs change. MACH Alliance’s approach : Advocates for systems that are Microservices-based, API-first, Cloud-native, and Headless. The focus is on openness, flexibility, and vendor neutrality In practice, many retailers mix and match elements from both schools of thought, selecting components they like best and integrating them into a unified storefront Further resources Planning for BFCM? Your peak season will be here before you know it. Load Testing now means you can uncover and fix hidden performance risks while there’s still time to act. Whether it’s validating autoscaling, stress testing a new platform, or proving your infrastructure can handle peak traffic, our real-world journey Load Testing gives you the data and confidence you need to deliver flawless CX when it matters most. Best practices for AWS autoscaling Learn more about best practices for AWS Autoscaling .

Are you ready for AI that shops for your customers? In Part 1, we explored why Agentic Commerce might not be just another AI buzzword —and how signals from Amazon, Adobe, Shopify, and others point to a real shift in how digital commerce is evolving. So what should retailers actually do to prepare? Let’s start with the foundation: communication between systems. Enter the Model Context Protocol (MCP) To succeed in an Agent-First world, retailers need more than a traditional website. They need machine-readable ecosystems that AI agents can understand and transact with. This is where Anthropic's MCP (Model Context Protocol) comes into play. It is a developing standard that provides tools and context to AI Agents, enabling brands to communicate product, pricing, and availability data in formats that AI agents can read and act on. In simple terms, MCP helps your backend systems communicate effectively with AI Agents. MCP along with structured data, becomes the key to enabling machine-to-machine transactions and product discovery in A gent-Led journeys. The next step Agent-to-Agent Commerce The next step will be for those agents to negotiate and complete transactions directly with other Agents . That’s where  where Google's new open protocol Agent-to-Agent (A2A)  comes in. Developed with over 50 technology partners, A2A enables AI agents to communicate, share information, and coordinate actions across different platforms. For retailers, that means your own AI Agents will be able to speak directly with other Agents—negotiating prices, checking stock, and completing purchases—without human involvement. Complementing MCP, A2A provides a universal way to connect, manage, and integrate agents from different providers, paving the way for faster, smarter, and more automated commerce. So what does this mean for Online Retail? Whether you're fully convinced or still cautious about Agentic Commerce, the implications for online retail are already becoming visible. Here is what is changing and why it matters. 1. Visibility and understanding of AI Traffic becomes critical Not all AI traffic is the same. Some agents scrape your content while others generate real sales opportunities. Retailers will need to differentiate between profitable agent traffic and performance-draining bots, then optimise accordingly. 2. Discovery shifts to Agent-Led journeys As AI tools begin to replace traditional search engines, product discovery will be driven by large language models rather than search engine results pages. Retailers must begin to structure their data for machines as well as for people. MCP and structured data will become the new front doors to your storefront, enabling AI agents to interpret product information, availability, and pricing in real time. 3. Commerce moves beyond websites In the near future, customers may not visit your homepage at all. Expect more transactions to occur through Agent-to-Agent (A2A) interactions, where your backend negotiates pricing, stock, and delivery with other systems. Your site will still matter for brand and human visitors, but much of the actual commerce could shift into backend exchanges between agents, powered by MCP and structured data. 4. AI Agents raise the bar for performance While AI agents are patient, they optimise aggressively. A slow site or a broken checkout process could result in your offer being ignored or downranked in an agent’s internal selection process. This makes Load Testing and Real-user Journey Monitoring essential for maintaining performance under new conditions. Customer journeys now span across devices, bots, assistants, and digital backends. It will no longer be enough to monitor how a human travels through your site. You’ll soon need to monitor how your systems perform for both human and machine visitors alike. Where to begin (realistically) with Agentic Commerce This shift may not require immediate action from every retailer. You don’t need to overhaul your operation overnight. But ignoring this disruptor could lead to missed opportunities and reduced visibility as the landscape changes. Here are seven practical steps to get ready for AI shopping: Prepare your backend for MCP compatibility Use monitoring and analytics tools to understand AI agent traffic and conversions Identify valuable AI agents and distinguish from scraping bots Reframe your SEO strategy to prioritise AI-based discoverability Ensure product data is accurate, structured, and machine-readable Conduct load testing for increased machine-to-machine transaction capacity Consider developing your own AI agents to operate in this growing ecosystem Final thoughts Agentic Commerce is not just a buzzword or a new channel. It represents a fundamental shift in how eCommerce operates. Retailers who adapt to serve both humans and machines will unlock new revenue opportunities, reduce friction, and future-proof their customer experience—even when their customers are not the ones doing the shopping. The real question is not whether Agentic Commerce is coming. It is this: How ready will you be when it arrives? Look out for our downloadable Checklist a complete 7-step preparation guide coming soon. Further resources Read Part 1 of our blog ' Agentic Commerce Part 1: Hype or the next big eCommerce shift?' If you'd like to learn how monitoring and Load Testing can help you prepare for AI-powered commerce, get in touch .

As OpenAI rolls out its fully integrated eCommerce agent in ChatGPT, Amazon tests its Buy for Me  tool, and generative AI traffic continues to surge, a new kind of eCommerce is taking shape — one driven not just by human shoppers, but by AI agents acting on their behalf. And with Google officially launching its AI-powered Search , the shift toward AI-led product discovery is no longer theoretical — it’s here. A bold new trend is taking hold in eCommerce: Agentic Commerce . It may sound futuristic, but it’s already here. In this article, we explore whether this is just another overhyped AI story or the signs of a major shift in the Digital Commerce ecosystem. What is Agentic Commerce? Agentic Commerce refers to shopping journeys handled by AI agents on behalf of humans. These "digital shoppers", often embedded in virtual assistants, apps, and other tools, can search for products, compare prices, evaluate quality, manage delivery options, and even complete transactions leading to fully autonomous shopping experiences. Rather than browsing your site, a customer might soon just say, "Find me the best running shoes under £100 with free next-day delivery." The AI agent does the rest - evaluates options, compares retailers, and makes the purchase. What used to be a simple transaction becomes a dynamic marketplace, where agents manage the process on behalf of both buyers and sellers. In this environment, how your platform performs , and how easily machines can understand it, will directly affect whether you’re included in the shortlist or overlooked entirely. Are AI agents really set to disrupt commerce - hype or reality? Retail is no stranger to tech trends that over-promise and under-deliver. So it's fair to ask, is Agentic Commerce more buzz than business? A few reasons why scepticism is understandable: There's a lot of AI noise right now. Businesses are rushing to add "AI" onto everything. MCP (Model Context Protocol), one of the enablers of Agentic Commerce, is still evolving. Many retailers are still focused on improving human user journeys rather than preparing for autonomous AI traffic. However, it's becoming clear that this is not just hype. Signs that Agentic Commerce isn't just hype Here are three key reasons Agentic Commerce demands serious attention. 1. It’s already happening Major players are launching real production tools. Open AI introduced Operator its AI Agent designed to handle eCommerce on behalf of users. Partnering with major eCommerce brands including Ebay, Etsy and Instacart. As of July it became fully integrated into ChatGPT. Earlier this year Amazon actively tested its "Buy for Me" feature ahead of Prime Day , allowing shoppers to use agentic AI to buy products from third-party retailers from within the Amazon app. OpenAI is integrating Shopify into ChatGPT    to enable a seamless online shopping experience, this could expand to other eCommerce platforms. AI-powered search engines such as Nvidia's-backed Perplexity are already driving measurable retail traffic right now, not just in the future, and are set to increase their influence with plans to preinstall their Comet browser on mobile devices . 24%  of US adults are already comfortable with agents shopping for them and that rises to 32%  among Gen Z Source: Salesforce 2. Traffic patterns are shifting New sources are already influencing discoverability and conversion. Adobe reported a 3,300% increase YoY in retail traffic from generative AI tools , such as the large language model (LLM) ChatGPT, AI-powered search engine Perplexity and other virtual assistants and web browsers. Even if total volume is still relatively small, the trajectory is steep and real. Gartner predicts that within the next 3 years, AI agent customers will replace 20% of interactions at digital storefronts . This is not a minor trend; it is an emerging ecosystem. Agentic interfaces, LLMs and context-aware APIs could fundamentally change how, where, and when retail happens: 3. It aligns with the market - everyone's a winner AI agents help by: • Reducing friction for shoppers • Speeding up purchasing decisions • Unlocking new devices and platforms for discovery Retailers benefit by: • Increasing conversion potential • Reducing customer effort • Gaining brand exposure through agent-led journeys And this is not just a B2C story In B2B Commerce, where purchases are often repeat, low-consideration, or bulk, AI agents can streamline procurement, speed up decision-making, and reduce manual effort. Final thoughts To succeed in an agent-first world, retailers will need more than a traditional website. They’ll need machine-readable ecosystems that AI agents can understand and transact with, and infrastructure that delivers the speed, accuracy, and reliability these agents expect. That’s where emerging frameworks like the Model Context Protocol (MCP) and Google’s Agent2Agent (A2A) come in. In Part 2 , we’ll explore what MCP means, why it matters, and how digital retailers can start preparing for the Agentic Era, from data structure to performance readiness. At thinkTRIBE, we help digital teams gain the CX visibility, performance, and resilience they need to stay ahead in an AI-first world. Find out more about how we support retailers here .

GLAM (Galleries, Libraries, Archives, and Museums) institutions and publishers share a common mission: to make knowledge accessible. But as a new wave of AI scraping bots descends on their sites, that mission is being put under serious strain. The irony: open access leads to outages AI bots harvesting data are overwhelming infrastructure, disrupting user experience, and in some cases, knocking platforms offline. This GLAM-E Lab report  outlines a growing problem: What happens when demand for digital knowledge becomes so great that it threatens the very platforms designed to provide it? It’s an ironic twist we all need to consider carefully. Why AI scraping matters now Over the last year we've seen a dramatic increase in publishers and GLAM institutions struggling to manage bot traffic. Wikimedia  reports AI scraping is putting strain on its servers with a 50% increase YoY in bandwidth for multimedia content and identify that bots generate 65% of its most expensive traffic, despite being just 35% of total visits. Platforms like DiscoverLife  have seen scraping traffic surges that rendered their sites unusable. There’s been a shift from AI training crawlers toward real-time retrieval bots (RAG), which scrape data at a much faster rate. Bots often ignore protections like robots.txt, change IPs frequently, and don’t identify themselves as bot, making them nearly impossible to block with traditional tools. The result? Degraded CX: slower pages, patchy availability, frustrated real users - and ultimately outages. In many cases, the warning signs appear too late. Analytics tools often filter out bot traffic by default, which means the early signs of a problem may be hidden. A moderate spike might not raise alarms but by the time traffic surges past capacity thresholds, real users are already being affected. Blocking isn’t so simple While there are sophisticated techniques to detect and deter bots such as device fingerprinting to rate-limiting, the rapidly evolving nature of scraping means no approach remains effective or affordable for long. Should institutions block bots outright? Or try to shape how they interact? It’s not easy when many bots act more like stealth crawlers than cooperative agents. Some claim to be retrieving data “on behalf of a user,” bypassing bot rules entirely. This creates a difficult balancing act: staying open and accessible, while protecting infrastructure and budgets from being quietly drained. A broader issue: ethics, economics, and expectations This isn’t just a tech issue, it’s a governance issue. Cultural and publishing organisations can’t keep scaling servers indefinitely. But they’re also hesitant to restrict access, especially if that means turning away legitimate users. As AI use grows, open access economics are faltering. Bot traffic brings no traffic or engagement, it just extracts. As a TollBit report shared with Forbes found, AI agents send back 96% less referral traffic  than a traditional Google search . Toward a more sustainable response There’s no single fix, but several ideas are emerging: Analytics : Understanding who’s scraping and what they're scraping is step one. Fair use models : Publishers are testing tools like TollBit and Cloudflare’s permission-based paywalls . These may not suit all GLAM institutions but show how norms might evolve. Legislation & standards : Regulating bots that impersonate humans may be more impactful than trying to police content reuse. Monitoring and Load Testing : Identify CX and performance risks before they turn into outages, whether from bots or real users. Proactive CX Monitoring and Real-world Load Testing is part of digital resilience. Let’s talk — not just block AI developers, publishers, and cultural institutions all have a stake in keeping the web’s knowledge infrastructure healthy. We need shared signals. Clearer expectations. And yes possibly new legislation. But above all, we need collaboration. Because building more servers won't fix what is, at heart, a governance problem. Let’s start that conversation. Further resources If you're looking to get ahead of CX and performance issues and identify issues before they escalate, learn more about our Real-world CX Monitoring and Managed Load Testing Services .

The way consumers shop has fundamentally changed—and it’s not just about the move from bricks-and-mortar to online. Increasingly, the battle for customer experience (CX) supremacy isn’t happening on websites, but within native mobile apps. Mobile commerce is no longer just a trend—it’s the preference for millions of consumers. And while both mobile web and apps have a role to play, the numbers are drawing a clear line in the sand: native mobile apps are winning. Native Apps engage and convert at higher rates The more tailored, personalised experience delivers both increased engagement and conversion rates. Data from Glance Group shows that app users typically spend 3.5 times longer engaging with brands and complete purchases 1.5 times more frequently than website visitors . Why? Because apps offer something the mobile web simply can’t: a personalised, easier and faster experience. Speed , simplicity and stickiness Unlike mobile websites, native apps store data locally, making them inherently faster. Actions that require server calls on the web can happen almost instantly within an app. The frameworks underpinning native apps can run much faster than JavaScript-based mobile websites, meaning tasks feel near-instantaneous. For the customer, that translates into fewer delays, fewer taps—and fewer reasons to abandon a purchase. Apps also streamline customer journeys. Saved preferences, stored payment methods, and personalised recommendations remove friction, reducing the effort it takes to convert. Add in the power of push notifications to nudge shoppers back to their abandoned baskets or alert them to flash sales, and the result is a stickier, more compelling path to purchase. The 5G and satellite-enabled future of Mobile CX The technical infrastructure supporting mobile experiences is also evolving rapidly. The rollout of 5G networks has already slashed latency and boosted bandwidth, improving app responsiveness. But the next leap is even more profound: satellite-enabled mobile coverage, like Vodafone’s satellite video calls , is eliminating network dead zones. This means uninterrupted app performance—regardless of geography. Social Commerce and Live Shopping Another powerful shift is happening through the rise of social commerce. Platforms like TikTok and Instagram are rapidly becoming full-fledged shopping ecosystems. Livestream shopping—once a niche format—is now mainstream, with platforms like eBay Live and Whatnot driving billions in revenue. These immersive experiences are optimised for mobile apps, which support real-time video, embedded checkout, and interactive chat far more seamlessly than browsers. It’s a redefinition of what “shopping” means in the mobile-first age. Mobile Payments are redefining the checkout experience Native Apps are also where the latest innovations in payments are playing out. Whether it’s digital wallets, Buy Now, Pay Later (BNPL) options, or biometric security features like face ID and fingerprint login, checkout within apps is becoming faster, more secure, and more intuitive. This reduces abandonment and boosts trust—especially important as customers become increasingly sensitive to any friction in the final steps of a purchase. So why are online retailers ignoring the risks of Native App CX and performance? Despite the clear business case for mobile apps, many eCommerce brands are still flying blind when it comes to visibility of Native App CX and performance . While website monitoring is standard practice, Native App Monitoring is usually overlooked—leaving retailers exposed to glitches, slow load times, and hidden errors that quickly silently erode trust and conversions. In an era where Native Apps are the preferred storefront, ensuring they perform seamlessly isn’t optional—it’s essential. They carry the weight of your brand experience and often to your most loyal, highest-converting customers. In this context, failing to monitor your app is more than a technical oversight—it’s a commercial risk. As mobile CX becomes the new battleground for eCommerce loyalty and revenue, the ability to proactively monitor, diagnose, and fix issues in real time may be the difference between growth and churn. Start seeing what your customers experience - before it's too late Native App Monitoring Blind spots in Native App CX and performance aren’t just technical issues—they’re business risks. Our real-user monitoring service for Native Apps helps you detect issues before your customers do, so you can protect conversions, loyalty, and brand trust. Discover how thinkTRIBE’s native app monitoring can give you true visibility where it matters most.

Is your AI quietly breaking your website? AI is transforming eCommerce — powering everything from personalised product listings and targeted recommendations to AI-generated content and predictive search. It’s helping brands deliver more tailored, responsive, and optimised customer experiences. As one of the leading sectors for AI adoption, eCommerce businesses using AI-driven strategies are seeing an average revenue uplift of 10–12% . But there’s a problem few teams are prepared for: AI brings powerful capabilities — but also technical complexity and integration risks  that can lead to subtle, technical glitches, hard-to-spot errors and unexpected behaviour in your live environment. When these AI tools go wrong, they don’t fail in obvious ways. These AI risks in eCommerce fail quietly — and inconsistently. The invisible problem with AI-powered customer journeys AI enables personalised experiences at scale. That’s a strength — but also a growing risk. These real time, dynamic changes create fragmented user journeys — and with that fragmented failure modes. Whether it’s a missing checkout button, an incorrect product image, broken text, or wrong pricing, it might only affect a subset of users — while working perfectly for everyone else. Traditional QA and synthetic monitoring rarely simulate those edge cases. So your systems stay “green” — while real customers struggle. Why these issues are hard to detect These aren’t the kinds of errors that trigger alerts. They don’t appear on uptime dashboards or crash your servers. Instead, they: Subtly reduce conversion rates Increase checkout abandonment Erode trust and NPS over time They don’t cause dramatic outages — just quiet cracks that weaken CX and conversions. Such as a dynamically priced item that disappears at checkout, a predictive recommendation that leads to a 404 error or an AI-generated product image that fails to load on mobile browsers. Monitor AI risks in eCommerce from the outside in Most teams rely on a combination of internal tools, uptime monitoring, and traditional testing. But these systems don't give full visibility of your CX and often assume a single, consistent version of your site — when in reality, every customer is seeing a version shaped by hundreds of real time variables. Unless your monitoring behaves like a real user — using real browsers and customer contexts — you risk missing the most costly issues. You don’t need to simulate every possible journey — but you do need to monitor the most valuable ones, including: Add-to-cart journeys across key product categories, PDPs, and offer types Checkout flows that include AI-powered steps or upsells Mobile vs desktop journeys When monitoring reflects how actual customers behave 24/7 — you can catch hidden issues before they show up in your NPS or revenue data. The more personal the journey, the more hidden the risk Forrester recently highlighted that AI and operational resilience risks  are now core concerns for enterprise risk management (ERM) programs — a sign that these issues have become business-critical challenges. As AI continues to shape the customer journey, your monitoring and QA approach needs to evolve too. The more tailored your experience becomes, the more specific — and harder to spot — your failures might be. If your monitoring still assumes a one-size-fits-all journey, it may be time to rethink what “realistic” really looks like. What next? If you'd like to explore how true CX visibility through real-user monitoring can help surface AI-related issues, explore our DCX web intelligence monitoring. If you're looking beyond the technical risks of generative AI, Deloitte breaks it down into four key categories  — including data integrity, governance, and unintended outcomes.

The feedback every website owner wants to hear is that their customer’s journeys on their website are smooth, quick and conclude with the customer making a purchase, that’s the dream right?  So what are the nightmares? What do we need to avoid to make our website’s performance be in tip top condition so our customers feel happy and content to browse, share and buy? What are the tools for monitoring website performance? The far too long 3 seconds. It’s a known fact that if a website takes more than three seconds to load, that user will open a new window, find a new site that loads quicker and continue on their journey, buying the product or services they came to your website for, from another. Now, we know that 3 seconds in reality seems like such a short time, but in the ecommerce world of CX, that is 1 second far too long! Sales killer glitches. Another killer for your site is glitches. If a customer is using your site and they come across a button, drop down menu, or link that doesn’t work, for example it goes to the wrong page, you get an error code or absolutely nothing happens…you got it, that customer is disgruntled and gone! Crash, crash, burn. There’s nothing quite as efficient as sealing the deal of a no sale situation than a site that crashes. Whether that’s due to too many people on the site at once, or just a pathway confusion that’s landed your customer’s journey at a messy end, that wasn’t a sale, that’s another lost customer. So how do we prevent our customer journeys coming to unproductive ends? The only way to see what your customers see, is to use tools for monitoring website performance to test and monitor your site for all the issues mentioned above. Running 24/7 customer journeys gives you ongoing visibility of those customer-impacting issues. This allows you to measure and trend performance over time. Customers don’t follow a strict path on their journey through your site, they might hop between all areas, creating a huge web, interlinking different pages and media in hundreds of different ways. By dynamically following real customer journeys, your website’s performance can be measured and understood. Your site is aimed to sell products and services to customers, so it’s pretty vital that their journeys are the ones you recreate and monitor rather than testing statically with predetermined URLs. Monitor what matters to maximise your customer’s experience with Synthetic Monitoring. Here at thinkTRIBE we offer a full set of tools for monitoring website performance to help ensure your website’s performance results in customer sales rather than customer disappearance. Contact us today here at thinkTRIBE to see how we can help your website with dynamic, real time website performance monitoring.