Speero

By Ben Labay: "Advanced personalization involves 'growth engineering' as a new 'role' to connect the dots and architect the data, from tools/data like: ? CDPs ? Data warehouses ? Testing tools that enable adaptive approaches, e.g., Contextual Multi-Arm Bandits (or similar) ? And advanced 'intent' or 'propensity' data and models The last point is where Mr. David Mannheim comes in. He just pushed out a cool Ecom report on intent. (https://lnkd.in/gFcB-s7f) There, we can find concepts, vocabulary, and taxonomy that influence the last point of propensity data. Things like: - 63% feel manipulated by ecommerce tactics (only 11% don’t) - 46% feel overwhelmed on ecommerce sites - 83% use discount codes when they would have bought at full price - 1 in 5 will stop shopping if they get an early pop-up The TRICK is to get this data accessible to the testing and engagement platform setup. Feature attribute data: > CDP defined User-level attributes: account tier, number of past upgrades purchased, engagement metrics (time on site, feature usage). > Session-level attributes: current time of day, day of the week, user’s device type, current navigation path or product page. > External attributes (optional): Geo-location, known seasonal promotions, pre-determined propensity model data All this sounds cool, but WHY/WHERE to apply this stuff? > Adaptive Learning: A dynamic personalization approach continuously updates the probability distribution of reward for (offer/product/promo) as new data is collected. Unlike a static A/B test, it doesn’t wait for a full experiment cycle to end before updating which offer to show next. (we don't care what wins, just push to what is working best now) > Context Utilization: This setup leverages user and environmental context (e.g., user account age, user’s current usage tier, time of week, location). This allows for personalized experiences rather than one-size-fits-all solutions. Add in explicit propensity and 'intent' data (h/t to David here) and you really get cooking. > Handling Concept Drift: If certain upgrades become more or less attractive over time, the testing/personalization algorithm automatically adjusts. This adaptability ensures that the system remains optimal in the face of changing market or user conditions. Yes this is where AI experimentation tools come into play, but the foundation of tooling and explicit data ontology needs to be there first. A personalization (also AB testing) recipe is only as good as its ingredients. Bottom line? The right data, connected smartly, powers personalization that works— Want to dig deeper into David’s intent report, architect your own growth engineering setup, or just swap ideas on making this real for your team? I’m all ears—DM me or drop a comment below. Let’s cook up something impactful together!" Original post by Ben Labay In this week's Speero's best take of the week: when we celebrate the best takes and opinions from our fellow experimentation experts.

Speero tested offering an easy way to apply coupons to first-time visitors and increased transactions (+7.73%), revenue per visitor (+6.65%), and add to cart (+3.88%) in the process. The story started with us noticing an opportunity to help first-time customers convert more by adding an option to apply a coupon with one click on the cart page. Hypothesis: If we add a checkbox/button to apply a coupon with one click, this will increase motivation and thus revenue per visitor. The result? A winner: +7.73% in Transactions (100% CTBC) +6.65% in Revenue per Visitor (99% CTBC) -0.02% in Orders with Subcriptions (49% CTBC) +3.88% in "Add to Cart" (PDP) (99% CTBC) Observed Monetary Impact: $136,995 /month (AOV * Extra Conversions) There's a significant increase in Revenue per Visitor and Transactions. The pattern is consistent across all devices. Implement the variant and iterate with more custom targeting (e.g. exclude returning visitors).

Here’s how Speero helped Tipalti increase the lead-to-opportunity ratio from 5% to 12-24% in its highest volume lead gen channels. During that time, Tipalti had several high lead generation channels. However, this advantage became an issue since many of those leads weren’t turning into opportunities. Speero helped Tipalti turn this around by identifying growth opportunities across all those channels. For instance, their paid channels drove lots of leads, but the lead-to-opportunity ratio was only 5%. Today, they’re averaging (depending on the sector and time zone) between 12-24%. Most of that came from learning and partnership with Speero.

Product and marketing teams fight a flood of (bad) data and channels that come and go each year. TikTok became popular in 2020 but was almost banned in the US. SEO looks completely different since AI search came into play. The world has changed dramatically for marketing, product, and experimentation practitioners. Decision-making on the ‘front line’ must now keep pace with a flood of (often unreliable) data and evolving consumer channels and behaviors. Practitioners prioritize: - Authority to Test: Teams need the autonomy to run experiments without relying on cumbersome approval chains. Solution: Don’t give responsibility without authority, ”trust but verify”. - Strategic Integration: Experiments must align with overarching business goals rather than being siloed as mere tactics. Solution: clarify what core business opportunities and customer problems need to be solved with a Goal Tree Map workshop. - Systems of Trust: Reliable data and consistent methodologies are non-negotiable to build confidence in experimental results. Solution: Ensure access to quality data systems. Start operating with the mantra, "Move fast, but with certainty." Focus on solving strategic problems, not just optimizing minor metrics.

Winning with experimentation isn’t just about running tests—it’s about the people who drive growth and the platforms that power them. Experimentation Live 2025 (by Speero and Eppo) is dealing with just that. Every talk, workshop, and speaker will align with one of these focus areas: - Growth: Content tailored for marketers, product managers, and metric owners driving business impact through experimentation. - Platform: content designed for ops, enablement, and data/tooling experts who build the infrastructure that makes experimentation possible. We’re keeping the same format and focus – Mini-group table talks (5-8 people) where you network and discuss problems you share with your peers. Keynotes are there to start conversations, not steal the limelight. - We’ll ask you about your issues, goals, and priorities. - We’ll find people who have them as well. - You all get together and discuss them. - Find solutions and mingle. Experimentation Live 2025 (by Speero and Eppo). Happening on 21-22 October 2025, in Austin, Texas. October 20 is reserved for workshops (more on those soon). Get an early bird 25% off coupon, enter: EARLYBIRD25 Consultants who bring a client get a bigger discount can contact Speero’s CEO ([email protected]) for more info. The coupon is available through July 31st or the first 50 coupons used, whichever comes first. Experimentation Live 2025, by Speero and Eppo.

Here’s how Speero tested anchoring the bigger price and increased revenue per visitor by 5.2% and AOV by 1.5%: People rely way too much on the first number they see–AKA anchoring bias. If the first price looks high, everything after seems cheap by comparison. Speero decided to run a test and use this. The client’s product page defaulted to the Deluxe option. We wanted to anchor to the higher price. Hypothesis: If we order the size options from highest to lowest, it will anchor users to the grand option, which will raise the value perception of the default deluxe choice and lead to more users choosing a larger bouquet and higher overall AOV. The Result? Big Win. +5.2% Revenue per Visitor +1.5% AOV 99% stat sig. Implement the variation and iterate with: 1) Highlighting “best value” product. 2) Adding a small review of someone loving the “Deluxe” package. 3) Explaining how much more value they get with the most expensive option.

Original post by Ryan Lucht: "The validity and productivity of A/B testing is, at this point, a decades-long consensus held by everyone from Nobel Prize-winning economists, to some of the best-performing CEOs of the century thus far, to smart practitioners at every seniority level in fields as broad as Marketing, Product, Engineering, and AI/ML. This is a pretty strong consensus - if you think a handful of W Edwards Deming quotes has suddenly given you a unique insight as to why it's all BS and you're making it your mission to disprove A/B testing... well, extraordinary claims require extraordinary evidence. Good luck!" Original post by Ryan Lucht. In this week's Speero's best take of the week: when we celebrate the best takes and opinions from our fellow experimentation experts.

Speero tested matching the fragrance selectors for shampoos on the website with the same colors the product bottles use. This increased, transactions (+2.26%), revenue per visitor (+2.21%), reached checkout page (+1.38%), and click on "add to cart"(+2.07%). This time, we decided to test something creative and not so easily connected to typical UI testing. We noticed an opportunity for increased clarity by differentiating the fragrance selectors for shampoos on the website with the same colors the product bottles use. So fragrance selectors will be in the same colors as shampoos with that fragrance. Hypothesis: If we match the scent selector with the color of each of the bottles, the scent offering will be clearer, and so more transactions will be completed. The result? Winner - Implement Variation resulted in: +2.26% in Transactions (96% CTBC) +2.21% in Revenue per Visitor (95% CTBC) +1.38% in Reached Checkout Page (92% CTBC) +2.07% in Click on "Add to Cart" (PDP) (98% CTBC) Observed Monetary Impact: $45,273 /month (AOV * Extra Conversions) Our recommendation was based on realizing that users struggle to interact with the scent options: 1) Apply this learning to other areas of the website. 2) Implement this variant and continue testing to optimize the interactions with the scents.

So many Shopify merchants spend hours tweaking their store, running ads, and testing ideas, only to see minimal results. Why? Because they’re guessing and copying competition, instead of having a clear strategy for: - Running trustworthy A/B tests - Optimizing their store for conversions - Finding budget-friendly tools that fit their needs And honestly? It wasn’t their fault. There’s so much noise out there. It’s hard to know what actually works. That’s why Speero is teaming up with Shopify to host an Ask-Me-Anything (AMA) Forum—a live event where you can get real answers to your biggest questions. This isn’t about theory or fluff. It’s about actionable advice you can use right away: - Clear steps to run meaningful experiments in Shopify. - A list of the best tools for A/B testing and optimization. - Insights into what’s worth testing in 2025 (and what’s not). - How to use AI to grow your store Event details: - February 27, 12-2pm EST - On Shopify community platform Here’s the best part: You don’t have to prepare anything. Just show up with your questions. Or drop them in the comments below. The link to register is down in the comments. As for the format: 1. Join the Shopify community if you don't have a Shopify store. You can join for free with your email. 2. Ask questions during AMA 3. This isn't a video webinar but a textual, board AMA.

If you’re running a Shopify store, you’ve probably asked yourself: “Why aren’t my ads converting?” “What’s the secret to turning visitors into buyers?” “Can I A/B test my Shopify store without outside help, quickly and cheaply?” Here’s the truth: You don’t have to figure it out alone. Speero is teaming up with Shopify to host an exclusive Ask-Me-Anything (AMA) forum, designed for merchants like you who are ready to stop guessing and start growing. This isn’t just another webinar or lecture. It’s your chance to: - Ask anything: From A/B testing to AI trends, no question is off-limits. - Get expert answers: Speero’s team of marketing and experimentation pros will break down exactly what works (and what doesn’t). - Learn from peers: Connect with other Shopify merchants who are facing the same challenges—and share what’s worked for you. We’ll cover questions like: - How do I run trustworthy A/B tests in Shopify? - What’s the best way to optimize my store for conversions? - What tools should I be using to grow my store in 2025? - How do I use AI to stay ahead in ecommerce? - What’s actually worth testing right now? This is your chance to get actionable, no-BS advice that you can implement immediately. Event details: - February 27, 12-2pm EST - On Shopify’s community platform Drop a comment below with the #1 question you’d like us to answer—or just show up and ask us live. Let’s turn your Shopify store into a growth machine.