Convert.com

Experiment-first. Revenue-focused. Now Convert Certified. Welcome, Door4. 🎉 They're a UK-based revenue optimisation agency that ties everything back to one thing: are we actually growing the business? Traffic, conversions, scaling - it all has to connect to real outcomes for their clients. What we like about them is that they're honest about how growth works. You try things. Some work, some don't. You learn either way. The experimentation mindset is baked in for them. Not a service line. Just how they think about growth. If you're looking for a partner who'll push you to do things differently and actually move the needle - go talk to Door4. Welcome to the network. Really glad to have you here.

How can teams bring more speed into experimentation... without losing control of the process? 96+ API endpoints 5,279+ docs indexed One MCP server All exposed through just 14 namespaced tools with progressive disclosure. The agent only loads deep schema context for the specific action it's about to call - keeping the context footprint light by design. At a practical level, it gives AI assistants direct access to Convert’s full REST API v2, so teams can query, analyze, and manage experiments through natural language. But the bigger story is what that changes in the day-to-day work. For developers, it means less platform plumbing, less context switching, and a faster path from brief to deployed experiment. The code still lives in their workflow, follows their standards, and stays fully reviewable. For strategists and CRO teams, it means experimentation data becomes easier to interrogate, connect, and learn from. Instead of spending hours pulling reports and stitching together past results, teams can ask better questions and get to insights faster. What we like most is that it doesn’t force a tradeoff between speed and control. Teams can move faster without giving up code ownership, visibility, or governance. We also think it’s important to be clear about where this fits. Prompt-based experimentation tools are great for quick mockups and early ideation. The MCP server is different. It’s built for production-grade deployment, analysis, and long-term maintainability. That distinction matters. Experimentation gets more valuable when it’s not just easier to launch tests, but easier to learn from them, scale them, and fold the winners back into the product properly. Curious how others are thinking about this: Where do you see the bigger opportunity right now - faster test creation, or better systems for learning from experimentation data? Would love to hear how your team is approaching it in the comments.

This is my last transmission from this page. So I need to make it count. There is a line between persuasion and manipulation. The industry spent a decade pretending it does not exist. But you know exactly where it is. You feel it every time a website makes you feel stupid, pressured, or trapped. Persuasion is someone showing you value clearly enough for you to act on it. Manipulation is exploiting the way your brain works to push you toward something that serves them at your expense. Deceptive design merchants have tried to rebrand manipulation as "advanced persuasion." It packaged dark patterns as best practice. It built an entire professional ecosystem around the idea that tricking you is just what good business looks like. It is not. It never was. You have the right to not be lied to. You have the right to understand what you're getting for your money. You have the right to make a decision in your own time, on your own terms. You have the right to cancel without running a gauntlet built to break you. You have the right to say no without being shamed for it. You have the right to use the web without being treated as a mark. Another web is possible. I have spent 24 hours here trying to prove it. Now I need you to go and demand it. From the sites you use. From the businesses you buy from. From the people building the web around you. This is not my fight. It is yours. It always was.

When to use Frequentist vs. Bayesiannnnnnnnnnnnn..... Hello? Hello? This isn't regularly scheduled programming... hello? We're being interrupted.... Hello....... ### This is not a scheduled post. My name is Jon Crowder. I run Another Web Is Possible, an ethical experimentation consultancy. And for the next 24 hours, I've taken over this page. Why? Because the web is broken. Not by accident. By design. Right now, on websites you visit every day, countdown timers are ticking down to nothing. Stock indicators are lying about scarcity. Confirmshaming copy is calling you a fool for saying no. Cancellation flows are deliberately built to exhaust you into giving up. This is not "best practice." This is a con. And the industry that built it has been getting away with it for too long. The 2025 Edelman Trust Barometer found that trust is now equal to price and quality as a purchase consideration. Every dark pattern deployed against you is unethical and commercially stupid. But the people profiting from it do not care, because they measure success in weekly spikes, not in the trust they destroyed to get them or the long term value that disappears when they do. I have 24 hours on this page. I intend to use them. If you have ever been tricked into a subscription you did not want, guilt-tripped by a cookie banner, or had costs hidden from you until the final checkout step, these posts are for you. The fight back against dark patterns starts now.

$2.5 billion. A$100 million. £99.9 million. Consumer rights violations get expensive fast. What poor customer ethics can cost (the numbers are staggering): Amazon agreed to a $2.5 billion settlement over alleged deceptive Prime sign-up and cancellation practices. Qantas was ordered to pay A$100 million for misleading consumers over cancelled flights. TSB was fined £10.9 million, after paying £99.9 million in redress to affected customers. Ignoring consumer rights comes with a serious price tag. With World Consumer Rights Day coming up this Sunday, the focus is shifting from "what can we track?" to "what must we protect?" Companies get their own internal data hygiene in order but ignore the vulnerabilities in their tech stack. Your compliance is only as strong as the weakest tool you’ve plugged in. We built privacy into the core of the platform so your experiments strengthen your posture instead of compromising it: - Zero PII: We do not touch personally identifiable information. - First-Party Ownership: Data stays with you via first-party cookies. - Global Standards: Built-in support for Consent Mode, GDPR, CCPA, ISO, and SOC-2. Protecting people is a commitment that must run through every tool in your stack.

Your experimentation team has a high churn rate. It’s not your "perfect" hire, it’s the lack of a system. AI can execute, but it still hallucinates. To hit 10x output, you need more "humans in the loop." You need well-trained practitioners with the context required to verify AI outputs. Only when the verification layer is solid can senior leadership stop troubleshooting and start focusing on actual business levers. We asked 21 experts how they build teams that stay and scale. Here is the blueprint: Hire for Temperament: Jeremy Epperson looks for the 6 traits of high-functioning practitioners. You can’t train critical thinking or process obsession. The Hero’s Journey Onboarding: Ben Labay suggests getting new hires on the flywheel immediately. Give them a customer problem to own, not a checklist to read. Fresh Eye Diagnostics: Alex Birkett advocates for pulling ideas from new hires in the first 30 days before they "drink the kool-aid." The Common Language: John Ostrowski uses specific deep-dives to ensure the whole team speaks the same language of statistics and growth. Teach to Learn: Rishi Rawat, uses bi-weekly sessions where team members teach the group. It’s the fastest way to solidify a new skill. Culture over Perks: Sina Fak and Manuel Da Costa highlight that maturity comes from accountability and process. The "Testing Bible" Approach: Ryan Levander builds internal repositories to connect thoughts across articles and experiments. The full frameworks include insights and technical checklists from: Shiv Narayanan, Peep Laja, Bhavik Patel, Martijn Scheijbeler, Lucia van den Brink, Chad Sanderson, Ton Wesseling, Tim Stewart, Guido X Jansen, Els Aerts, Ruben de Boer, and Paul Rouke. The output of your tools is limited by the input of your people. Experimentation momentum shouldn't die when your "star" player leaves the room. Build a system that outlasts any single resignation. Is your program one resignation away from collapse? Read the full article: convert.com/blog/a-b-testi…

Tech retailers struggle with the same wall: Buyer hesitation at the point of purchase. For one major retailer, refurbished IT products were a tough sell. Shoppers were confused about refurbishment grades, warranties, and product conditions. CAC was spiking, and confidence was tanking. 180 Marketing stepped in to make trust visible. Instead of running a single "big bet" redesign, Bruno Borges and his team executed 10 A/B tests in 90 days. They moved from research to the first live test in just two weeks. - Warranty Transparency: Highlighting item warranties directly on the page led to a 17.7% CVR increase (98% confidence). - Grade Availability: Detailing product stock by refurbishment grade solved bulk order friction, resulting in a 10.1% CVR lift. - Momentum Over Perfection: They focused on quick wins to build a self-sustaining experimentation culture. Over 12 months, the program delivered a 58% increase in conversion rate. By using Revenue per User (RPU) as their North Star, they ensured that every lift in conversion also protected the Average Order Value. 180 Marketing used GA4, heatmapping, and Convert Experiences. Bruno credits Convert’s flexibility, specifically the ability to toggle between Bayesian and Frequentist methodologies, making it easier to maintain strict significance levels across projects. Trust is a conversion lever. When you remove the "refurbishment" guessing game, the revenue follows.

Happy International Women's Day 💛 from the women building Convert. We're posting a video today, but the point is the work behind it: making sure women in experimentation and growth get more stage time, more leadership opportunities, and more credit for the work they’re already doing. It’s who gets invited into the room early, who gets trusted with the “big” projects, and who gets handed the mic when it’s time to share what worked. We’re focused on moving the needle where it actually counts: The WiE Summit: 1,200+ registrations for a 100% core CRO event, not AI hype. Speak Up to Uplevel: Active sprints helping women scale their personal brands and professional voices. 80% Women-Led: A leadership team that ensures our budget goes toward opening doors, not just opening tabs. Experimentation only works when you’ve got different perspectives in the room: different backgrounds, different instincts, different ways of seeing the customer. When the same voices dominate, you don’t just miss out on fairness… you miss out on better work. Tag a woman in experimentation whose work we should all be following. 👇

50 editions. 16 years. Dialogue Donderdag is still going On March 12 in Utrecht, Online Dialogue is hosting DiDo #50 - Party Edition: a proper milestone for anyone who cares about CRO, experimentation, UX, data, and psychology. They’ve tracked the evolution of CRO from a "nice-to-have" to a core business operating model. They’re bringing 4 top minds together to debate the past, present, and future of the craft. It remains the definitive meeting point for the Dutch experimentation scene. At Convert, we love seeing communities built on transparency and long-term value. We couldn't be happier to see this one hit such a legendary number. If you’re a client-side pro in the Netherlands, don’t miss the party. Congrats again to the whole OD crew! What’s one CRO topic you actually want the panel to debate? (Hot takes encouraged) onlinedialogue.nl/over-ons/dialo…

80% of our leadership team is women. Morgan Legge Trina Moitra Tiffany Fortune Dionysia Kontotasiou Poonam Nagekar This wasn't intentional. It was the result of our founders being grounded in performance and trust, not who shouts the loudest. And this circle of amazing women went to work bringing unbiased opportunities to other women. The proof's in the execution. - The WiE Summit (2024 & 2025): 500+ live attendees and 1,200+ registrations for a 100% core CRO event (zero AI hype). 2026 is loading as we speak.... - The Speak Up to Uplevel Competition: Where we ran sprints to help women publish thought leadership, share their thinking, and build a personal brand in CRO/UX/growth. And now we’re adding something we think matters. We teamed up with Kevin Anderson to build a tracking widget that shows the fastest-growing women in CRO, experimentation, and growth on LinkedIn. The goal is simple: Exposure for women, yes. But also accountability. Are our efforts moving the needle. Follower count isn't the best gauge of this. However, it isn't the worst proxy either. Launching on 9th March in respect of International Women's Day. The findings will be clustered by location and other interesting parameters. Are you interested in understanding how much visibility women in our field have? Who are growing? And how? Tag a fellow female/non-binary experimenter who'll benefit from this.

"Which A/B tool should I use?" Your answer is currently worth $0. Change that. Most A/B testing affiliate programs pay like it’s 2016. So we rebuilt ours. We launched the Convert Ambassador Program 3.0 to change that math. - 20% recurring commission in Year 1 (yes, recurring). - 90-day cookie window. Enterprise sales take time. We honor your link for 3 full months. - Tracked + paid via Reditus. (A single Pro Plan referral nets you over $1,000.) If you educate the CRO space, you belong in this program. Every application receives a manual review. We limit the number of ambassadors to ensure we can support your specific goals and provide the training you require. Stop settling for standard payouts. Your influence represents a massive asset, get paid what it is actually worth. If that’s you, apply here → convert.com/partnerships/a…

Want to influence what we build next? Here’s exactly how. Product development should not happen behind closed doors. To build a better experimentation platform, we need a tighter feedback loop between our team and the people using the tool every day. The Convert Public Roadmap is live: Real-time Transparency: See exactly what is Planned, In Progress, and Shipped. The Developer Filter: Check our direction before building manual workarounds. Agency Alignment: Use roadmap links in client audits to show what is coming. Community Logic: Vote on ideas and share your specific business use cases. How to influence what we build: 1. Browse: Search existing requests to see if your idea is already planned. 2. Vote: Up-vote the features that impact your revenue or workflow the most. 3. Submit: Share the "Why" behind your request—not just the "What." This represents a transparency system for how we build. Your feedback directly dictates where we point our engineering effort. View our roadmap and submit your feedback: convert.canny.io

"We will do it for you" no longer represents a viable agency model. Agencies that hide the "how" are already losing. Last year, 65% of companies moved experimentation in-house. These brands do not want a siloed vendor to manage a spreadsheet. They want a partner to architect a sustainable internal culture. As Craig Sullivan explains, if you fail to help them build that engine, you simply will not be part of their picture anymore. The Old Way: "We will do it for you" (The vendor trap). The New Way: "We will build it with you" (The enablement partner). Stop trying to own the work. Start building the system that makes the work possible. That is how you become a permanent partner rather than a line item to be cut.

Don't pick an A/B testing tool until you've seen this 15 platforms built for developers who demand full control Marketing picks the tool. Engineering kills the deal. It is a story we see every week. To avoid the veto, developers need more than a "snippet." They need experiment-as-code, zero flicker, and deep SDK coverage. 15 Best A/B Testing Tools (Categorized by Fit): Convert.com: Best for privacy-first, enterprise testing with strong APIs. Optimizely Full Stack: Best for large-scale, server-side experiments. LaunchDarkly: Best for advanced feature flagging and rollouts. GrowthBook: Best open-source platform with flexible hosting. Split (acquired by Harness): Best for robust SDKs and real-time flagging. Statsig: Best for warehouse-integrated, engineering-led testing. VWO FullStack: Best for teams bridging marketer UI and dev APIs. ABsmartly: Best for high-performance server-side SDKs. SiteSpect, Inc.: Best for flicker-free SPA and server-side testing. Adobe Target: Best for standardizing on the Adobe marketing stack. Amplitude Experiment: Best for tying analytics directly to testing. PostHog: Best all-in-one open-source and product analytics tool. Kameleoon: Best for privacy-sensitive and compliant product teams. Eppo by Datadog: Best for data teams running warehouse-native experiments. Firebase A/B Testing: Best for mobile developers using Remote Config. Engineers will push back on heavy scripts and black-box logic. Choose the tool that integrates with your CI/CD and respects your site performance. Read the full breakdown (link in the comments)

1 in 4 buyers leave after one generic message. Stop burning your high-value lead list. Personalization no longer represents a "bonus" feature. B2B buyers have zero tolerance for "close enough." In a market flooded with alternatives, relevance represents your only real defense against disengagement. The 2026 Reality: - 81% of prospects ignore messages lacking direct context. - 25% of leads disengage entirely after a single generic touchpoint. - Manual 1:1 personalization moves too slow for modern buying cycles. - 96% of buyers convert when the experience feels bespoke. How teams scale 1:1 relevance: High-performing B2B squads move away from broad segments. They use AI to prioritize high-value accounts and adjust messaging based on real-time behavior. This ensures consistency across every channel without increasing headcount. Automate your relevance or lose the market. The future of B2B belongs to teams that deliver true 1:1 experiences at scale.

We studied 237 agencies last year. Speero joins us to track the 2026 fallout. What a difference a year makes. Last year, 65% of you saw clients shift testing in-house and 80% felt the pressure to perform was at an all-time high. Much of what we predicted about the shift to product teams and AI maturity has already come true. To help navigate the next shift, we’re excited to have Speero on board as a sponsor for this year’s research alongside Craig Sullivan. We’re back to hunt for signals in the noise and see what has changed since our last report. Help us build the 2026 blueprint: - The AI Gap: Moving beyond basic prompts into true automation. - Pricing Pressure: How top agencies defend margins in a saturated market. - Product Shifts: The reality of optimization being rolled into product squads. - Growth Pathways: Identifying the "blue ocean" opportunities for your team. If you helped us last year, your voice was the backbone of the report. Let’s do it again. Take the 8-minute survey: asfpdxyh.paperform.co Enter the prize draw to win a $300 Amazon gift card (drawn 31 March 2026). Results are anonymised and aggregated.

AI is the engine, but humans are the heart. Step into the ring. The future of experimentation is coming to London. 🎪 The way we build, ship, and test is changing....fast. That’s why the Convert team is attending Circus London this May 11 & 12. Not your average "sit and listen" conference. It’s a 250+ person deep-dive into the "Humans in the Loop" philosophy, featuring a massive lineup of in-house leaders from brands like Etsy, Zalando, and Just Eat. 2 days of extreme networking, AI-agent workshops, and 100+ roundtable sessions with the leaders who are actually in the arena. We can’t wait to hear from this lineup of practitioners and moderators: Keynotes: Julie Beckley (Etsy), Emil Palikot, Marcel Toben (Zalando), Leah Tharin Lightning talks: David Mannheim, Alex Gold, Iqbal Ali, Bjarn Brunenberg, Samantha Reis, Oleksandra Bernatska, Patrick Christmas, Sadie Neve & Nils Stotz Moderators: Evie Thompson, Laura Haczewski, Steve Quinlan, Surjit Panda, Tony Him, Vasileios Fasoulas, Vignesh Lokanathan, Melanie Kyrklund, Peter Gostev, Theo Gough, Marco Cordioli, Jeffrey John Baker, Julian Kaljuvee, Gareth Holder PhD, Jamie Martin, Matt Mulvey, Divya Isaiah, Bhavik Patel, Dr Simon Jackson, Vasant Kumar, Stewart Ehoff, Luiza de Lange. Why we’re clearing our calendars for this one: - The Rings: 3 dedicated tracks covering the AI Growth Engine, Taming the Black Box, and AI ROI for leaders. - The Workshops: Hands-on sessions on building multi-agent analysis systems and AI workflows. - Extreme Networking: 100+ table sessions and "activation" events designed to actually connect you with peers, not just collect business cards. - The Posters: 9+ lightning talks on everything from "Agentic Engineering" to "Synthetic Users." JOIN US IN THE RING (AND SAVE 30%) The early bird window is closing. Use code 👉 earlybird30 before the end of February to lock in your spot at 30% off. 🎟 Register here: speero.com/circus See you at Kings Cross!

Still stuck in VWO? Moving to Convert is now as easy as opening a new tab. Switching your experimentation stack shouldn't feel like moving houses in a rainstorm. If you’ve been holding off on migrating because of the manual setup, we have news. You can now migrate your VWO experiments to Convert Experiences by simply having both tools open. The 2-Tab Migration Workflow: 1. Open your VWO dashboard in one tab and Convert in the other. 2. Choose your experiment type (A/B, Split URL, or Multivariate). 3. Use the Convert editor to pull your variations, targeting rules, and goals directly across. 4. Drop the Convert tracking code into your site’s (one snippet to rule them all). 5. Hit "Preview" to ensure your goals and targeting are firing perfectly. 6. Flip the switch. 🚀 Beyond the easier migration, you're gaining: - Zero Flicker: Protect your Core Web Vitals and user experience. - Privacy-First: GDPR/CCPA compliance baked into the architecture. - Superior Support: Real humans helping you with every line of code. Pro-Tip: We recommend moving one high-priority experiment at a time to ensure total data continuity. Ready to make the jump? Our support team is standing by to walk you through the process.

To floor a professional audience, you have to call out the "Vibe-Testing" epidemic. Most A/B testing hypotheses are just gut feelings dressed up in corporate jargon. When you base your experiments on conjecture, you lose before the test even fires. You can't defend the results, you can't replicate the wins, and you certainly can't explain the failures. Teams are stuck in a loop of "random acts of optimization." They change a headline because a competitor did it. They move a button because it "feels" right. Then, when the results come back flat, they have no data-backed observation to explain why. A robust hypothesis is a legal defense for your strategy. It’s the difference between being a "button-pusher" and an experimentation lead. After putting millions of tests live, we’ve found that a credible hypothesis requires 5 specific parts. No more, no less. 1. The Observation (The "Why") This is the outline of the problem. It must be 100% free of conjecture. Use qualitative or quantitative data to highlight a phenomenon. If you can’t defend the observation with data, the hypothesis is dead on arrival. 2. The Execution (The "What") The where, what, and who. Which segment of traffic is seeing this? Exactly what element is changing? This defines the "effort" in your PIE/ICE prioritization. 3. The Outcome (The "Prediction") Your educated guess. You must name the KPI you expect to move and the direction of that move. Pro tip: Always track secondary KPIs to ensure external factors aren't skewing your primary "win." 4. The Logistics (The "Math") How long will it run? What is the required sample size? What significance level are you chasing? If you don't set logistical expectations early, you'll end up peeking at data and calling "winners" too early. 5. The Inadvertent Impact (The "Ethics") Experiments involve humans. A thorough analysis of possible negative impacts on user behavior can (and should) modify how you conduct the test. Ethics in testing isn't just a "nice to have"—it's a guardrail for your brand. We built a tool to help you automate this thinking. Try Convert’s Free A/B Testing Hypothesis Generator: convert.com/hypothesis-gen…

Stop settling for incremental gains. Here are 11 A/B Tests from growth practitioners you can run today: 1. PLG Flow vs. Cluttered Landing (Rosie Hoggmascall, Fyxer) Stop using generic landing pages. Rosie compared how Typeform and Mailchimp leveraged loops. Typeform used a minimalist prompt after a survey: “You’ve taken one. Now make one.” It outperformed Mailchimp’s cluttered, text-heavy footer. The Lesson: Context beats clutter. 2. Newsletter Funnel Validation (Victor H.) Don’t build the SaaS first. Victor tested demand with a curated newsletter. Early affiliate revenue helped pay for ads, creating a self-sustaining acquisition engine before a single line of code was written. 3. Social Proof Headline Test (Oliver West, VML) Swap generic copy like "Join today for free" with "Join over 2.5 million traders." This simple credibility nudge lifted signups by 11% for a retail trading platform. 4. Removing Friction from Trials (Yarden Morgan) Remove the credit card requirement from your trial signup. Lusha tripled their signups while trial-to-paid conversion remained steady. Sometimes the fastest lift comes from removing "standard" barriers. 5. AI-Suggested Templates in Onboarding (Runbo Li) Don’t drop users onto a blank canvas. Preloading onboarding with AI-suggested templates improved first-week retention by 18% by reducing cognitive load. 6. Interactive Onboarding Checklist (Or Moshe) Trade long tutorials for quick-win checklists. Tevello saw a 21% rise in activation by giving users a sense of momentum early on. Early "wins" drive more confidence than over-explaining features. 7. Personalized Cancel Flow (Khushi Lunkad) Stop offering blanket discounts to everyone who tries to leave. Khushi segmented the cancel flow based on user feedback. Adding a small message about "preserving your data" improved churn prevention by 14.2%. 8. Progress Meter for New Users (Ibrahim Alnabelsi, Prezlab) Gamify the setup process. Ibrahim introduced a "brand score" progress bar that reflected asset completion. New account retention jumped by 17% because progress is addictive. 9. The Labour Illusion (Mike Fawcett) Faster is not always better. Mike tested a "loading" animation with messages like “Searching the market.” Even though results were instant, users trusted them more because they felt "effort" was involved. 10. Contextual Paywalls (Rosie Hoggmascall) Trigger your paywall at the exact moment of intent. Canva triggers upsells when a user hits a premium feature they want. This makes the paywall a bridge, not a fence. 11. Accelerating Review Loops (Joseph Cochrane) Timing multiplies the effectiveness of a loop. Joseph nudged exhibitors for reviews within 48 hours of an event, increasing review volume by 37% and providing fresher social proof for new prospects. Growth relies on reducing cognitive load, honoring context, and making progress visible. If you had to clear everything and launch just one of these tomorrow, which one would it be?