Utilizing Data For Ecommerce Decision Making

Shopify SEO tip: How Google crawls collections (products) to map rankings and clicks: Google is ranking based on how you structure collections. I have read patents (and tested) how Google often treats the first 1-5 products as the “signature items” of that collection. This makes us question which products slot into collections (also at the top). Brands usually have "best sellers", which makes sense. But here’s what the data and SERPs suggest for Shopping rankings: - Shopping carousels can pull SKUs from the top rows - Collection relevance influenced by anchor products - “Popular products” snippets above-the-fold - Crawl priority favours first-listed SKUs This means product order isn’t just UX or merchandising. It’s quietly influencing SEO visibility. Brands need to match SERP and customer intent, you can do SERP searches to understand this. What the data suggests to rank better: 1. Anchor products: (SEO focus) - Evergreen, keyword-rich SKUs locked in top 3 - Category-defining -> “signal” products 2. Attribute-rich SKUs: (structured data) - Reviews, GTINs, availability, priceValidUntil - Feeds into organic Shopping features 3. Seasonal rotation: (You should swap in seasonal relevance) - Surface promos without displacing anchors - Blend relevance + recency 4. SERP mirroring (reverse-engineer Google) - Identify which SKUs Google surfaces already - Align your top slots to match search intent ----- The good news? I have used this strategy to drive 100,000s of product clicks and profit increase. This doesn’t require you to re-invent the wheel, just smarter ordering rules in Shopify (or whichever platform you use). Your Shopify collection pages aren’t just “lists”, they’re entity signals and SKU order matters. When I work with a brand, I'm mapping out intent to rank collections and drive profits. Do you treat product order as an SEO lever, or is it still just merchandising for you? P.S The screenshot is an example of SKU placement at category level.

🚨The greatest drop-off is from Product Details Page To Cart Page, so we must improve our Product Details Page! Not so fast ✋ In today's age of data obsession, almost every company has an analytics infrastructure that pumps out a tonne of numbers. But rarely do teams invest time, discipline & curiosity to interpret numbers meaningfully. I will illustrate with an example. Let's take a simple e-commerce funnel. Home Page ~ 100 users List Page ~ 90 users Product Display Page ~ 70 users Cart Page ~ 20 users Address Page ~ 15 users Payments Page ~12 users Order Confirmation Page ~ 9 users A team that just "looks" at data will immediately conclude that the drop-off is most steep between Product Details Page & Cart Page. As a consequence they will start putting in a lot of fire power into solving user problems on Product Display Page. But if the team were data "curious", would frame hypothesis such as "do certain types of users reach cart page more effectively than others?" and go on to look at users by purchase buckets, geography, category etc and look at the entire funnel end to end to observe patterns. In the above scenario, it's likely that the 20 cart users were power users whilst new & early purchasers don't make it to this stage. The reason could be poor recommendations on the list page or customers are only visiting the product display page to see a larger close up of the product. So how should one go about looking at data ? Do ✅ Start with an open & curious mind ✅ Start with hypothesis ✅ Identify metrics & counter metrics that will help prove/disprove hypothesis ✅ Identify the various dimensions that could influence behaviours - user type, geography, category, device type, gender, price point, day, time etc. The dimensions will be specific to your line of business. ✅ Check for data quality and consistency ✅ Look at upstream and downstream behaviour to see how the behaviour is influenced upstream and what happens to the behaviour downstream. ✅ Check for historical evidence of causality Dont ❌ Look at data to satisfy your bias ❌ Rush to conclude your interpretation ❌ Look at data in isolation - - - TLDR - Be curious. Not confirmed. #metrics #analytics #productmanagement #productmanager #productcraft #deepdiveswithdsk

“Let’s collect ALL the data first, analyze later?” Classic mistake, it’s called “boiling the ocean”. Instead, use a “hypothesis-driven approach”. …I also hate theory, so let’s jump into the example: Example: Checkout conversion plunged from 2.8% → 1.9% last week. #1 Structure → Use a logic tree to break the problem into major components (𝐭𝐡𝐞 𝐦𝐨𝐬𝐭 𝐞𝐱𝐩𝐞𝐧𝐬𝐢𝐯𝐞 𝐬𝐭𝐞𝐩). → Then select a few hypotheses based on initial data, intuition, and past experience. #2 Hypotheses H1: Price shock (list price up). H2: Payment outage. #3 Tests & Results H1: Check if the drop was concentrated on SKUs with price increases? → Result: No difference found. Hypothesis killed. H2: Check if payment failure codes spiked? → Result: Validated, +240% errors on Android, clustered at midnight. Outcome: Payments were redirected to the backup gateway, and conversion recovered to 2.7% within 24h. 𝐍𝐨𝐭𝐢𝐜𝐞 𝐰𝐡𝐚𝐭 𝐰𝐞 𝐝𝐢𝐝𝐧’𝐭 𝐝𝐨? 𝐏𝐮𝐥𝐥 𝐞𝐯𝐞𝐫𝐲 𝐝𝐚𝐭𝐚𝐬𝐞𝐭. That’s the power of hypothesis-driven problem solving. 👉 More at: aseptamar.com https://lnkd.in/gMhS9-jK

In #ecommerce growth, the first big challenge isn’t always what you think—it’s coming up with the right hypotheses to test. But here’s the good news: it’s not as hard as it seems. It all starts with your business analytics. Dig into the numbers, uncover patterns, and spot areas for improvement. The key is to focus on real indicators that impact revenue – conversion rates, cart abandonment, AOV, and retention. Once you have the data, it’s time to generate hypotheses. Think of it as structured brainstorming: ➝ What small tweak could improve checkout completion? ➝ How might a new product bundle increase AOV? ➝ Would a different ad angle bring in higher-quality traffic? At this stage, documentation is everything. Capture each idea, including: * The assumption behind it * The solution you want to test * The method for testing * The expected outcome * The current and target metrics But hypotheses don’t happen in isolation. Regular team discussions fuel better ideas. The best sessions: - Have a clear agenda - Ask specific questions about impact, effort, and ROI - Build on past experiments to refine strategies Yes, the process might feel slow at first. There might even be some resistance. But after a few cycles, it becomes second nature – and the rewards are huge. Not just higher profits, but a clear roadmap for growth, backed by data, experience, and continuous learning. ––– 🤘 Follow me, Gadashevich, for more insights on growing your e-commerce business #shopify

Most business leaders shout about their data-driven approach to decision making. Yet in 90% of critical decisions, data is sidelined in favour of the seniority of decision makers. The reasons for this are rooted in human nature: - Fear of failure - Distrust in their team - Assuming data analysis will take too long The problem is: When developing your business with #Amazon, making decisions based on your "gut" backfires almost always immediately. That's because Vendor Managers inform all their decisions based on the performance of your account. And Amazon's business model is hardly comparable to any other retailer. Which means your past experience most certainly won't apply to its business model. So how can you increase your chances of making better decisions with Amazon? 𝗕𝘆 𝗯𝗮𝘀𝗶𝗻𝗴 𝟵𝟵% 𝗼𝗳 𝘆𝗼𝘂𝗿 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻𝘀 𝗼𝗻 𝗱𝗮𝘁𝗮. Want to know how much you can increase your cost prices with Amazon? 👉 Analyse how your Average Selling price compares to your Average Cost Price. Don't know if you should continue selling unprofitable items on Amazon? 👉 Analyse how your low-margin range contributes to first-time shoppers buying your brand. Here's the thing: Making decisions without data should form the exception, not the norm of running your Amazon business. --- How are you making decisions when it comes to Amazon? Let me know in the comments! #amazonvendor #amazonstrategy

Your Shopify collection structure might be the #1 factor that determines if ChatGPT recommends your products or your competitor's. (50% of our agency's success comes from how we structure collections for our eCom clients) AI doesn't think in broad categories. ChatGPT users don't search for "men's clothing." They provide DETAILED information to get results uniquely tailored for them. As a result, ChatGPT is searching the web for specifics like "lightweight post-surf hoodies." Most brands make the critical mistake of creating just a few broad collections: - Men's Clothing - Women's Apparel - Accessories These won't get you visible in AI shopping results. Instead, create hyper-specific collections that match how people actually search: - Lightweight Surf Hoodies - Cold Water Surf Gear - Quick-Dry Beach Apparel This strategy works because: 1. These specific collections convert at a higher rate (people find exactly what they want) 2. They're much easier to rank for than broad terms (less competition) 3. They match how people naturally talk to AI assistants You can include the same products in multiple collections. One product can exist in 5+ collections, each targeting different search intents. We've tested this across 75+ brands. The ones with detailed collection structures consistently outperform in both traditional SEO and now in AI recommendations. With ChatGPT likely integrating Shopify shopping soon, your collection structure will directly impact whether AI assistants recommend your products. Are your collections specific enough for AI shopping?

They told you to build categories. I'm telling you to build ranking machines. Here's the difference. Most ecommerce brands lose rankings on their most important pages. Why? Their collections are built for sorting, not searching. You don't need a household name to own those crucial product category rankings. You just need to shift your approach and build SPECIFIC pages designed to rank. You can’t have pages like: “Necklaces” “Rings” “Bracelets” …and expect to rank. You’re not Tiffany. Here’s how you win: 🤫 1️⃣ USP-led optimisation: → Not just: "Necklaces" (aimless browsing) → But: "[Your Unique Material/Style] Necklaces for [Your Ideal Customer]" (ranking for specific needs) - Add your USP to the H1 and meta title - Structure your URL to include key terms - Weave your differentiator throughout the description - Make the meta description work for you (not against you) 2️⃣ Audience targeting: → Not just: Listing everything vaguely → But: Creating collections that directly address specific customer needs and search queries - Speak your customers’ language - use their exact search terms - Create sub-categories for different audience segments - Create multiple collection pages for different customer needs - Build collections for different needs (and add links between them) 3️⃣ Search intent matching: → Not just: Showing products → But: Answering the "why" behind the search, removing all friction to purchase - Use schema markup to become more visible - Highlight product features your customers care about - Match both browsing and buying behaviour (e.g. filters vs. comparisons) - Remove barriers between searching and purchasing - no distractions, no dead ends This works. 🔹 One client added hundreds of new visitors with these changes on just 8 collection pages 🔹 Another jumped from page 2 to position 2-3 just by adding “waterproof” to their jewellery collections P.S. What specific USP could you add to your collection pages right now? 🧠 Follow me – Freddie Chatt – for ecommerce SEO that makes your existing pages work harder.

I have spent 1000+ hours in meetings with Jeff Bezos. His process of inquiry and review challenged and improved my thinking more than anyone I have worked with. Here is how I prepared for meetings with him: 1) The famous narrative memo Many people know that Amazon meetings are conducted using a structured, six-page narrative memo. However, few people understand what is required to write one. It isn’t something you can throw together the night before. The first draft is rarely strong enough, and preparation must begin at least a week in advance to allow for revisions. Once in the meeting, Jeff will spend 15-20 minutes carefully reading the document, making notes, and highlighting items to question. If your memo is thorough, he will cross out most questions as they are answered later in the document. After the read-through, what follows is 40 minutes of in-depth questions and discussion. No detail is too small to escape his attention, and no problem too complex to dissect. Prepare yourself by anticipating his questions/concerns and addressing them in the document. 2) Be prepared to say, “I don’t know.” “I don’t know” is a sufficient answer, as long as it is followed by “Here is when and how I will get the answer.” It is better to acknowledge that you don’t have the answer and follow up later than to wing it or be vague. 3) Metrics must be accurate and thoughtful Don’t present numbers based on bad data or without a deep understanding of the data sources and how metrics are calculated. Jeff takes the concept of inquiry to its utmost — he assumes that every number and statement is false until his inquiry process yields sufficient evidence. Your meeting will derail if you don’t know the details. 4) Diagnose Root Causes The objective of meetings is a decision. Good decisions start with an understanding of the current problems and opportunities. Many leaders confuse simply providing data with diagnosing the root cause of issues. Data and metrics are tools used to get to the root causes. You can’t fix a problem or win an opportunity without an accurate diagnosis. Jeff expects you to distill data down to root causes. You won't get to a decision with Jeff until you are aligned on the problem you are trying to solve. 5) Give objective options. Jeff expects a structured evaluation of multiple alternative solutions, with an objective assessment of the pros and cons of each. It is not sufficient to say, “This is our plan.” Effective narratives describe multiple options, highlighting the one the team recommends and why. 6) Assumptions will be challenged. “What makes you so sure about that?” is a frequent question. Every claim must be backed by verifiable data or logical reasoning. 7) Highlight problems, not just successes. Intellectual honesty is valued over good news. Problems should be addressed proactively with clearly recommended solutions, not ignored or obscured. Continued in the first comment.

I’ve noticed a concerning trend that I’ll call “over optimization” As ecommerce marketplaces have evolved, we’ve been given more and more data sets And they’ve been made available faster and faster You can pull up an Amazon app on your phone to see your sales and ad performance up to date to the minute almost Same for Shopify This feeds into the instant-gratification mentality that often urges us to make changes when things seem to be off track but aren’t necessarily It’s the emotional part of our brain that lets anxiety and fear drive our action That’s bad in this instance in case that isn’t clear The solution is to install a process to guard against these urges You want to measure data sets with enough statisical relevancy and then develop a hypothesis to test You then need to implement the changes to test your hypothesis and give it time so you can collect statistically relevant data You also need to step back and evaluate at a macro level every so often such as a mid-year review or QBR You have to be able to trust in your process even if it isn’t easy to do That’s what we do as an agency If brand partners push back and want to make impulsive changes, we push back hard Unfortunately the decision is not always up to us, but we at least stand by our principles It’s hard to stay disciplined, but almost all successful companies and people are very disciplined #ecommerce #process #amazonsellers #amazonvendor