Speero

The client wanted to expand into the German market. So we copy tested (via Wynter) the client's German audience to explore their emotions and perceptions of the client’s website claims. Objective: Find out which elements of confidence building the German audience resonates with and does not resonate with. Plan: 1 test run via Wynter. 30 panelists, 3 feedback areas. What will we do with the data: Adjust (Brand) PTR page based on feedback from audience. Default Questions for Overall Insights: - What is your first reaction after reading this section?  - After reading everything, what’s still unclear? - How compelling is this offer? (Likert scale) - Please rate the overall clarity.  (Likert scale) Questions for Highlight Insights: - What is your first reaction after reading this section? - What about this is unclear or off-putting? - Rate the clarity of the highlighted text.  (Likert scale) - What information on this section could be clearer? Level of audience: German - Mid-level Founders/Owners/CEOs (company size is 2-50) plus Operations and Finance. Size of audience: 30 per test. Results: - Reviewers mentioned that the header was vague in messaging and wanted clarity of benefits and outcomes. - They also mentioned that the product description section was well articulated but had an overwhelming amount of copy. - Respondents mentioned better organization of the copy and using visual elements could make the information better. - The credibility sections had a positive impression. However, presence of too many awards raised doubts on the authenticity of the trust icons. - Overall, the page had a positive impression on the reviewer. They mentioned a need for visual demo of the product (either through a video or real pictures). - The call to action was termed as pushy and presented too early. Recommendation:  - Test a clearer header with specific benefits and outcomes. - Reduce and reorganize text in the product description with more visual elements. - Add a video or real images to demonstrate the product visually. - Experiment with the timing and placement of the call to action. - Limit the number of trust icons to avoid credibility concerns.

How can you effectively implement multivariate testing to optimize complex landing page elements simultaneously? Can you leverage ML to predict and optimize the landing page conversion rate? The costs of shipping and returns are going up: - How can you design landing pages to set realistic expectations and potentially reduce return rates? - What strategies can you use to present shipping costs in a way that doesn’t lower conversions? Join Ben Labay, Rand Owens, and Michael Aagaard for a casual, panel-style Q&A about UX designs for maximum conversions and AI in eCommerce. Happening on Wednesday, Oct 23, 2024, 10 am - 11 am (CDT). Link in the comments.

EXL + Unite was a blast! 3 days of workshopping, listening to great keynotes, discussing experimentation, and partying, of course. We had some high-quality presentations but most of the conference was focused on topic-specific table discussions with peers. These were deeply insightful, chock-full of ideas and people ended up with at least 5 things to start building this week! People ate 20 tacos on average and drank some of the best mocktails, frozen drinks, and alcohol of course! People met, shared stories and insights. Some even became friends! PS: Based on the data from selfies, Shiva Manjunath became friends with everyone.

EXL + Unite has ended, and what a phenomenal experience it’s been! Here's what people say about it: - "I can not believe how much I learned since I got here, how many friends I made and how much great food I ate 😋 Had the expectations high for this event and still got me surprised with everything. Definitely coming back!" Ana Catarina C. - "The roundtable discussions and casual chats during meals and breaks are the best parts of this conference. The talks and super cool venue aren’t bad either!" Kathryn Mueller - "The last few days were a whirlwind of speakers, roundtables and - most important - getting to know more of the experimentation community from all over the world...The roundtable discussions were great, special thanks to all the moderators who made them happen." Greg Wendel - "It was an incredible three days filled with insightful workshops, moderated table sessions, engaging talks, and plenty of fun. 🎉 A special highlight for me was the "Advanced Topics in Practical A/B Testing" workshop. Huge thanks to Ron Kohavi for patiently answering all my questions!" Lidija Turkovic Missed out? Don’t worry—you can already start planning for EXL 2025! Expect more insights, more innovation, and even greater opportunities to transform your approach to experimentation and optimization. 📅 A huge thank you to all the brilliant people who contributed to making this event a success! Until next year!

What an incredible three days at EXL + UNITE by Speero and Kameleoon 2024 in Austin, TX! 🎉 To everyone who came, thank you for bringing your energy and passion—this is what makes the core of any event. While EXL 2024 may be over, the journey continues. We’re already planning for EXL 2025, and it promises to be even bigger. 🌟 Don’t miss your chance to be part of the next wave of growth, optimization, and experimentation! Big thanks to our speakers, partners, and all the attendees for making this such a memorable event. Looking forward to seeing you all next year.

Yesterday was a blast at EXL + Unite! Everyone was hanging out and talking (hopefully some were listening as well). We had great keynotes from: - Melissa Weiss (When experimentation isn't love at first test) - Carole Robin “Connect” Workshop: Grow Impact by Building Stronger, More Robust Relationships - Lukas Vermeer: Bridging the gap: Integrating the Product Operating Model and the Experimentation Growth Model - Stewart Ehoff: The Golden Thread - Connecting research, experimentation & business ROI And of course, our branded small table discussions where most of the networking happens (minus the parties).

EXL Unite is in full swing since yesterday! Lots of networking, lots of learning! We started with two workshops: Advanced topics in practical A/B testing by Ron Kohavi. Applications of AI in experimentation by Craig Sullivan Shawn David, Iqbal Ali and Andrew McNaught. Today there'll be lots of great keynotes and even more networking! See you there!

When I was running Native Deodorant, the first thing our customers saw on the PDP was the product ingredients. We thought, people want to know this. They want to read about the ingredients, return policy, and effectiveness. It turned out, people wanted to see the reviews. No one cared about the ingredients. I like customer psychology because it often proves me wrong. So that was really fun for me and continues to be really fun for me. I think that was the most enjoyable part of it. Also just constantly being surprised. I remember early on there was some guy who was running a Facebook ad platform.I forgot what it's called, but it basically helped you buy Facebook ads. And he's like, which one of these two creatives do you think work better? He said: it's this crappy one, not the beautiful one that someone spent a lot of money on. He was right.

How can multivariate testing be effectively implemented to optimize complex landing page elements simultaneously?What advanced personalization techniques you can use to dynamically adjust landing page content based on user behavior and data? How can machine learning algorithms be leveraged to predict and improve landing page conversion rates? What advanced analytics methods can be used to gain deeper insights into user behavior and decision-making processes on landing pages? Given the rising costs of shipping and returns in e-commerce, how can landing pages be designed to set realistic expectations and potentially reduce return rates? What strategies can be employed on landing pages to present shipping costs in a way that doesn't deter conversions but also maintains profitability? We're running a free webinar ‘Perfecting Ecommerce Landing Pages with Michael Aagaard’. Come and join us for a casual, panel-style Q&A about UX designs for maximum conversions and AI in eCommerce. Happening on Wednesday Oct 23, 2024 10am - 11am (CDT). Link in the comments.

Should and can you run multiple experiments at the same time? Newcomers think you shouldn’t. What if the users see variant B on Homepage, but variant C on the product page? What’s the impact of this? #1: Interaction effects are possible, but very rare. There’s a lot of research[1] backing this argument (most points to zero interaction effects). There, we cleared #1. As for #2… There are tools to help you determine how big and nasty are your interaction effects. If you’re worried, you can always calculate it. Even more, you want to measure interaction effects because those tests you’re running on your homepage and product page… You wanna know if they fit together. In the end, most of the experiments you’re running are like looking into a glass orb. So you wanna know if the scrying glass is smudged. Now let’s get into the nitty-gritty details: The first option is to run tests fully next to each other. So they should have the exact same data range. But typically, you won’t have this situation. To be frank, it’s quite hard to actually do this. So, most of the time, you’ll have a partial test overlap.For instance, one test on the homepage, one test on the category page, or maybe two tests on the homepage. This means you’ll have an overlap depending on your traffic. But, as with everything, you have exceptions. So if you have lots of traffic and a tool to create mutually exclusive groups or non-overlay tests, you can run mutually exclusive tests. If this isn’t your way, you can always create an extra variant in your test that mixes multiple tests together. But this gets complicated quickly. Because most of the time you got a hypo you wanna prove. Another option for you is multi-page experiments. But these experiments get QUITE complicated, so we don't recommend doing this. But if you want, you can. Of course, you’ll have to know the pages where you’re running tests, what’s the conversion goal, does it match up, etc. And your last option is waiting for other tests to finish and isolating tests. But this lowers your test velocity, meaning fewer wins, learnings, and progress. Thanks a lot to the Blueprint owner, Lucia van den Brink ✨️ Link for Blueprint in the comments.