Conversion Rate Data Interpretation

Are Open Banking providers misleading merchants with claims of 99.8%+ conversion? No — but merchants may be asking the *wrong question*. #Conversion in Open Banking isn't standardized like card authorization rates. Providers often use a range of definitions for the conversion metric. To get the full picture and ensure comparability, I recommend merchants request the following metrics in RFPs: 🏦 Technical Completion Rate 🏦 This measures how many payments are successfully completed after a customer finalises the initiation process (SCA in the bank app). It reflects bank API performance more than the provider's, but since it's commonly used, it's a useful starting point. 👆Conversion from Bank Selection👆 A critical point in Open Banking payments is selecting the bank for authentication. This metric shows how well the provider technically redirects customers to their bank for SCA. Furthermore, if the provider doesn't explain the process effectively you may see higher abandonment rates. 🛒 Conversion from Checkout 🛒 This is the most important, though hardest to measure. It tracks how many users complete a payment after choosing Open Banking as their method. It is most effective for providers who redirect customers to their own hosted payment page to select a bank and complete the payment. High abandonment at this stage can indicate poor communication or user confusion. Data splits you may want to request: New vs. Returning Users: Returning users typically convert better, but most merchants focus on new users. (If they can't provide this data then it gives insight into their data analytics) Bank-Specific Conversion: A provider’s performance varies because of their underlying bank mix. If one has more HSBC than Barclays the aggregate number is not comparable Value Bands: conversion of payments in the following bands 0-100,101-5000 and 5001+ can offer insight into a provider’s ability to handle different transaction sizes. Device & Channel: Providers must perform well in both mobile browsers and apps, particularly if you have an app integration. Finally, you may want to specify a customer conversion session - i.e. 20 minutes, 2 hours, 1 day. My experience is this is harder for providers to show consistently but could be valuable. If you’d like to discuss these metrics further, feel free to reach out! #Openbanking #openfinance #fintech #payments #checkout

𝗜𝗺𝗮𝗴𝗶𝗻𝗲 𝘁𝗵𝗶𝘀: You’re the head of marketing, and your CEO asks, “𝗪𝗵𝗮𝘁’𝘀 𝗼𝘂𝗿 𝘄𝗲𝗯𝘀𝗶𝘁𝗲 𝗰𝗼𝗻𝘃𝗲𝗿𝘀𝗶𝗼𝗻 𝗿𝗮𝘁𝗲?” Pause. Breathe. And never give a single, blended number. Here’s why: Blended conversion rates lump together traffic sources with different goals and behaviors. It’s the fastest way to mislead your CEO—and derail your strategy. Instead, here’s how you should answer: “Great question! We have multiple traffic sources, each serving different purposes. Which one would you like to dive into?” 𝗪𝗵𝗲𝗻 𝘁𝗵𝗲𝘆 𝗶𝗻𝗲𝘃𝗶𝘁𝗮𝗯𝗹𝘆 𝗮𝘀𝗸 𝗮𝗯𝗼𝘂𝘁 𝘁𝗵𝗲 𝘀𝗼𝘂𝗿𝗰𝗲𝘀, 𝗯𝗿𝗲𝗮𝗸 𝗶𝘁 𝗱𝗼𝘄𝗻 𝗹𝗶𝗸𝗲 𝘁𝗵𝗶𝘀: 1. Demand Capture → Paid Search / Affiliates → Pricing Page / Demo Request 2. Education / Exploratory → Main Website Pages → Blog & Resources Each source has unique intent—and requires a tailored measurement approach. 𝗙𝗼𝗿 𝗗𝗲𝗺𝗮𝗻𝗱 𝗖𝗮𝗽𝘁𝘂𝗿𝗲, 𝗯𝘂𝘆𝗶𝗻𝗴 𝗶𝗻𝘁𝗲𝗻𝘁 𝗶𝘀 𝘀𝘁𝗿𝗼𝗻𝗴𝗲𝗿. 𝗖𝗼𝗻𝘃𝗲𝗿𝘀𝗶𝗼𝗻 𝗿𝗮𝘁𝗲𝘀 𝗮𝘃𝗲𝗿𝗮𝗴𝗲 𝗮𝗿𝗼𝘂𝗻𝗱 𝟱%. 𝗠𝗲𝘁𝗿𝗶𝗰𝘀 𝘁𝗼 𝘁𝗿𝗮𝗰𝗸: → Landing Page Conversion Rates → Conversions to Opportunity → Opportunity to Revenue 𝗙𝗼𝗿 𝗘𝗱𝘂𝗰𝗮𝘁𝗶𝗼𝗻 / 𝗘𝘅𝗽𝗹𝗼𝗿𝗮𝘁𝗼𝗿𝘆, 𝗶𝗻𝘁𝗲𝗻𝘁 𝗶𝘀 𝗹𝗼𝘄𝗲𝗿. 𝗖𝗼𝗻𝘃𝗲𝗿𝘀𝗶𝗼𝗻 𝗿𝗮𝘁𝗲𝘀 𝗮𝗿𝗲 𝘁𝘆𝗽𝗶𝗰𝗮𝗹𝗹𝘆 𝘂𝗻𝗱𝗲𝗿 𝟭%. 𝗦𝘂𝗰𝗰𝗲𝘀𝘀 𝗶𝘀 𝗺𝗲𝗮𝘀𝘂𝗿𝗲𝗱 𝗯𝘆 𝗹𝗲𝗮𝗱𝗶𝗻𝗴 𝗶𝗻𝗱𝗶𝗰𝗮𝘁𝗼𝗿𝘀 𝗹𝗶𝗸𝗲: → Assisted Conversions → Chat Engagements → Avg. Session Duration → New Visitors vs. Returning Visitors → Keyword Rankings → Brand vs. Non-Brand Clicks The key takeaway? Blended metrics hide insights that drive action. Specificity isn’t just better—it’s essential. Friends don’t let friends give blended conversion rates to CEOs. Let’s keep marketing data meaningful. 🚀 Have you faced this situation before? How did you handle it?

founder learnings! part 8. A/B test math interpretation - I love stuff like this: Two members of our team (Fletcher Ehlers and Marie-Louise Brunet) - ran a test recently that decreased click-through rate (CTR) by over 10% - they added a warning telling users they’d need to log in if they clicked. However - instead of hurting conversions like you’d think, it actually increased them. As in - Fewer users clicked through, but overall, more users ended up finishing the flow. Why? Selection bias & signal vs. noise. By adding friction, we filtered out low-intent users—those who would have clicked but bounced at the next step. The ones who still clicked knew what they were getting into, making them far more likely to convert. Fewer clicks, but higher quality clicks. Here's a visual representation of the A/B test results. You can see how the click-through rate (CTR) dropped after adding friction (fewer clicks), but the total number of conversions increased. This highlights the power of understanding selection bias—removing low-intent users improved the quality of clicks, leading to better overall results.

Two years ago, while working on marketing analytics, I faced a challenge in optimizing ad spend for a digital campaign. The marketing team was running social media ads, but despite high traffic, the conversion rate remained low. Instead of increasing the budget, we turned to SQL and data analysis to identify inefficiencies. Breaking Down the Problem with SQL 1️⃣ Finding the Best & Worst Performing Ads We analyzed click-through rates (CTR) and conversion rates for each ad campaign. SELECT campaign_id, ad_id, COUNT(DISTINCT user_id) AS clicks, COUNT(DISTINCT CASE WHEN purchase = 1 THEN user_id END) AS conversions, COUNT(DISTINCT CASE WHEN purchase = 1 THEN user_id END) * 100.0 / COUNT(DISTINCT user_id) AS conversion_rate FROM ad_clicks GROUP BY campaign_id, ad_id ORDER BY conversion_rate DESC; 🔹 Insight: Some ads had a high CTR but low conversions, meaning they attracted traffic but failed to convert. 2️⃣ Identifying Wasted Ad Spend We checked if ads were targeting low-value customers who rarely made purchases. SELECT ad_id, COUNT(DISTINCT user_id) AS total_clicks, COUNT(DISTINCT CASE WHEN customer_lifetime_value < 50 THEN user_id END) AS low_value_clicks FROM ad_clicks ac JOIN customers c ON ac.user_id = c.customer_id GROUP BY ad_id ORDER BY low_value_clicks DESC; 🔹 Insight: A large portion of the budget was spent on users with low lifetime value, leading to poor ROI. 3️⃣ Finding the Best Audience Segments To optimize targeting, we analyzed which customer segments converted best. SELECT age_group, location, COUNT(DISTINCT user_id) AS total_visitors, COUNT(DISTINCT CASE WHEN purchase = 1 THEN user_id END) AS conversions, ROUND(COUNT(DISTINCT CASE WHEN purchase = 1 THEN user_id END) * 100.0 / COUNT(DISTINCT user_id), 2) AS conversion_rate FROM customer_data GROUP BY age_group, location ORDER BY conversion_rate DESC; 🔹 Insight: The highest converting customers were from specific age groups and cities, which weren’t the primary ad targets. Challenges Faced Data Volume Issues: The dataset contained millions of ad clicks, so I used indexed filtering to improve performance. Attribution Problems: Some users converted days after clicking the ad, so we used attribution modeling instead of last-click conversions. Budget Reallocation Resistance: Marketing teams were hesitant, so we presented data-backed ROI projections. Business Impact ✔ 20% decrease in ad spend waste by cutting low-value audiences. ✔ 15% increase in conversion rate after retargeting the right audience. ✔ Better marketing decisions through data-driven campaign optimization. Key Takeaway: SQL isn’t just for reporting—it helps businesses make smarter marketing decisions and maximize ROI. Have you used SQL to optimize marketing campaigns? Let’s discuss!

50 people toured our properties last month. Zero signed leases. Our marketing team said the problem was pricing. "We're getting leads and tours. The rents must be too high." I pulled the numbers. In seven years of multifamily, I've tracked one ratio religiously: 70% of leads should become tours, 30% of tours should become applications. Leads → tours? Great. Leads were flowing. Tours were scheduled. Tours → applications? Broken. Here's what I found: → Same unit. Same price. Same location. → 50 prospects walked through knowing the rent before they showed up. → Zero applications. If pricing were the problem, they wouldn't have scheduled the tour. They already saw the number. They came anyway. The problem wasn't what we were charging. The problem was what happened after they walked in the door. Follow-up timing. Tour experience. Objection handling. Application process friction. We fixed the tours → applications problem. Simplified the application. Shortened response times after tours from 48 hours to 4 hours. Added immediate follow-up calls instead of email-only. The math on getting this right: One vacant unit at $1,350/month costs $45/day. Fix the conversion problem 30 days faster, save $1,350. Across 12 units sitting vacant, that's $16,200 in recovered NOI. At a 5% cap rate, $324,000 in protected property value. For our investors, diagnosing conversion problems correctly means we stop giving away rent when the issue is execution. Everyone wants to blame pricing. Data separates pricing problems from process problems. One costs you margin. The other costs you nothing to fix. Comment CONVERT if you track tour-to-application rates. Our newsletter documents how we diagnose conversion problems and the exact fixes that work. Subscribe for the systems we use.

Amazon’s new "Views" metric in the SERP has been generating a lot of buzz—and for good reason. Just take a look at this example from my personal Amazon page (yes, Dexter the Mini Schnauzer has taken hold of my credit card). For JustFoodForDogs, Amazon shows "600+ views" and "500+ bought" in the past month. On the surface, that looks like an amazing conversion rate. But when you start diving into the numbers, things get a bit more complicated. Let’s break it down with data from three of my products that I analyzed : Product A: Amazon Display: 2K+ views, 300+ bought Actual Data: 4,405 sessions, 5,749 page views, 1,774 PPC clicks Actual Units Sold: 353 Conversion Rates: Implied (15%), Actual Overall (8.5%), Paid Traffic (8.01%) Product B: Amazon Display: 600+ views, 50+ bought Actual Data: 1,345 sessions, 1,634 page views, 233 PPC clicks Actual Units Sold: 137 Conversion Rates: Implied (8.33%), Actual Overall (7.81%), Paid Traffic (7.30%) Product C: Amazon Display: 2K+ views, 600+ bought Actual Data: 6,192 sessions, 9,719 page views, 2,443 PPC clicks Actual Units Sold: 718 Conversion Rates: Implied (30%), Actual Overall (12.05%), Paid Traffic (13.1%) Across these examples, there’s a clear pattern: the "Views" metric doesn’t tell the full story. If these views are mostly coming from paid traffic, the actual conversion rates could be much lower than what Amazon’s "X+ viewed" and "X+ sold" numbers imply. While most consumers won’t be calculating CVRs, sellers might think that a competitor’s product is performing better than it actually is, leading to potentially costly decisions. It’s also worth noting that not all traffic is created equal. If you’re driving unqualified clicks to your listing, it can drag down your conversion rate, signaling to Amazon that your product might be irrelevant and potentially hurting your rankings.

I run a CRO agency, but years ago, we stopped looking at conversion rates as the key metric. The reason? A 2.3% conversion rate doesn't tell you what to fix. It merely hints that something needs investigation. When you treat conversion rate as a goal rather than a diagnostic signal, you miss the deeper insights that drive real growth. Think of it this way: A fever doesn't tell the doctor what's wrong with you. It just indicates that something needs attention. Similarly, your conversion rate is a symptom, not the diagnosis. The real value comes from understanding the why behind the number: → Why do 97.7% of visitors leave without converting?  → Why do certain segments convert at 2x the average rate?  → Why do returning visitors behave differently than new ones? When you reframe conversion rate as a starting point for investigation rather than an end goal, you unlock genuine insights. The most sophisticated optimization programs don't chase conversion rates. They chase understanding. They're more interested in behavioral patterns than surface metrics. Because once you understand why people convert (or don't), improving the numbers becomes a natural outcome. Don't optimize for conversion rate. Optimize for customer understanding. #cro #conversionoptimization #abtesting #experimentation

𝗗𝗲𝗺𝗼 𝗽𝗮𝗴𝗲 𝘃𝗶𝗲𝘄𝘀 𝗱𝗿𝗼𝗽𝗽𝗲𝗱 𝟮𝟮%. But demo submissions doubled. Sometimes the best optimization removes friction for your ideal customers while adding it for everyone else. A client's homepage redesign launched and the analytics looked brutal at first: ❌ Traffic up, but engagement down ❌ People leaving faster ❌ Demo page views dropped 22% But I dug deeper into the data. Instead of panicking over vanity metrics, we focused on what actually matters: conversions. 𝗧𝗵𝗲 "𝗚𝗲𝘁 𝗮 𝗗𝗲𝗺𝗼" 𝗳𝗼𝗿𝗺 𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻𝘀 𝘁𝗼𝗹𝗱 𝗮 𝗱𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁 𝘀𝘁𝗼𝗿𝘆: ✅ Submissions doubled ✅ Conversion rate jumped from 7.14% to 18.46% ✅ Quality leads were actually increasing The homepage wasn't broken, it was actually working better. Less time on site + higher conversions = more qualified traffic finding what they need faster. 𝗧𝗵𝗲 𝗻𝗲𝘄 𝗵𝗼𝗺𝗲𝗽𝗮𝗴𝗲 𝗱𝗶𝗱 𝘁𝗵𝗿𝗲𝗲 𝘁𝗵𝗶𝗻𝗴𝘀 𝗯𝗲𝘁𝘁𝗲𝗿: 👉 𝗖𝗹𝗲𝗮𝗿𝗲𝗿 𝗺𝗲𝘀𝘀𝗮𝗴𝗶𝗻𝗴 = Less confused visitors reaching the demo page 👉 𝗕𝗲𝘁𝘁𝗲𝗿 𝗾𝘂𝗮𝗹𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻 = Only serious prospects continued to demo 👉 𝗜𝗺𝗽𝗿𝗼𝘃𝗲𝗱 𝗖𝗧𝗔 𝗽𝗹𝗮𝗰𝗲𝗺𝗲𝗻𝘁 = Made it easier for qualified leads to convert Fewer but better visitors = higher conversion rates.

A recent client of mine was celebrating: "Ben, our conversion rate is up 10%!" I said: "That's great... but what about revenue?" Revenue was DOWN 3%. 😬 Here's what happened: - They had removed a bunch of low-intent traffic sources from their marketing mix. - Fewer tire-kickers meant a higher percentage of visitors were converting. - Efficiency improved (less marketing spend), but overall volume suffered. This is something I find myself explaining often: ➡️ Conversion rate is an efficiency metric, not a growth metric. That's fine if the business wants to tighten it's belt, but it's less fine if it wants to reach orbit. In any case, it's important for any business to remember--no single metric tells the whole story, and even if you're focusing on one metric, you should never take your eyes of the others (especially revenue). When have you found yourself zeroing in on the wrong metric?

One of the biggest data analysis mistakes I see people making is this: drawing conclusions from holistic data. Here's an example: Your website's conversion rate is 1.1%. You look for benchmark data on a good website conversion rate, and you see 2-5%, depending on industry. So you decide your website is bad because it isn't meeting industry benchmarks. And then you create an RFP for a new website. You spend hours vetting proposals and then drop $150,000 on a new website. The new website launches and your conversion rate is 1.1%. Here's the problem: that 1.1% conversion rate isn't the real story. That's all your traffic, from all channels to all pages on all devices. That number doesn't tell you anything about how good your site is. You have to dig deeper. If 75% of your traffic is top-of-funnel blog traffic, then of course your conversion rate is going to be low. Or maybe you are spending thousands of dollars on brand awareness campaigns that aren't designed to drive conversions. Or did you notice how your search ads to your core service page have a 0.2% conversion rate on mobile? So maybe the problem isn't your site at all. Maybe it's your targeting. Or your mobile experience. Or your landing page. The holistic number won't tell you any of this. You have to dive deeper. But the value of the holistic number is it can give you a signal that you should dive deeper. Because if your conversion rate had been 5.6%, you probably would have patted yourself on the back and said, "Wow, we have a great website and everything is working." Even when it's not. So no matter what the holistic number is, always look at the more granular level so you can find the real story.