Most Shopify brands can’t tell you who their best customer actually is. They’ll say “women 25-45 who like sustainable products” and call it a day. That’s not a customer avatar. That’s a demographic guess — and it’s the reason their Meta ads cost three times what they should and their email flows convert like a wet Tuesday.
What’s in This Article
Here’s what the top 10% of Aussie ecommerce brands do differently: they build a customer avatar that is so specific, so deeply researched, that every piece of copy, every ad angle, every product launch practically writes itself. When you know your best customer at that level, your cost per acquisition drops, your repeat purchase rate climbs, and your marketing stops feeling like guesswork.
This article is the Prospects playbook we run with every brand inside eCommerce Circle. It’s the exact process we use to turn a vague “target market” into a high-resolution picture of the one customer who drives 80% of your profit — and how to use that picture to make better decisions across every channel.
Why “Women 25-45” Is Costing You Money
Demographic targeting is a relic from the 2018 Facebook Ads era. Meta’s algorithm already knows more about your potential customers than you do. What it doesn’t know — and what you need to teach it — is the psychographic profile of the people who actually buy, rebuy, and refer.
A 2024 Nielsen study found that campaigns built around detailed psychographic personas outperformed pure demographic targeting by 37% on return on ad spend. That’s not a rounding error. That’s the difference between scaling profitably and burning cash.
Here’s the shift you need to make. Stop thinking about your customer as a segment (women, 25-45, metro Australia). Start thinking about them as a specific person with a specific problem, a specific desire, and a specific moment in their life when your product becomes the obvious answer.
When we work with brands inside the More Orders Operating System, Prospects is one of the four foundations — alongside Product, Store, and Strategy. You cannot out-spend, out-design, or out-discount a broken understanding of who you’re selling to. Get the avatar right, and everything downstream gets easier.
The Five Layers of a High-Resolution Avatar
Most avatar templates floating around the internet are surface-level — age, income, job title, hobbies. That’s fine if you’re writing a university marketing assignment. It’s useless if you’re trying to sell a $180 ceramic planter to a real human who’s got 14 tabs open and two kids yelling in the background.
The avatars we build at eCommerce Circle have five distinct layers. Each one answers a different question, and each one shows up in a different part of your marketing.
- Layer 1 — The Situation. What’s happening in their life right now that makes your product relevant?
- Layer 2 — The Pain. What frustration, fear, or friction are they experiencing? Be specific.
- Layer 3 — The Desire. What does the ideal outcome look like, in their words?
- Layer 4 — The Hesitation. What’s stopping them from buying?
- Layer 5 — The Trigger. What specific event pushes them from browsing to buying?
If you can fill in all five layers for your best customer — with real quotes, real situations, real language — you’ll write ad copy that stops the scroll, product pages that close the sale, and email flows that feel like they were written specifically for the person reading them. Because they were.
The 30-Day Avatar Research Sprint
You don’t build a proper avatar by sitting in a Google Doc and imagining. You build it by doing the research. Here’s the 30-day sprint we run with every Connect member during their onboarding.
Source 1: Your post-purchase survey. Install a single-question survey on your Shopify thank-you page. The question: “What was going on in your life that made you decide to buy today?”
Source 2: Your best customer interviews. Pull your top 20 repeat customers from Shopify. Email them personally and offer a $50 store credit for a 20-minute call.
Source 3: Review mining. Export every 5-star and 3-star review from your own site and your top three competitors.
Source 4: DM and email audit. Go into your Instagram DMs, Gorgias tickets, and email replies from the last 90 days.
Source 5: Competitor audience research. Use a tool like SparkToro to see what podcasts your audience listens to and what accounts they follow.
Do all five sources in the same 30 days. Pattern-match. The stories that come up again and again? That’s your avatar.
The Tool Stack: Klaviyo + Shopify Segments + a Spreadsheet
You don’t need fancy customer data platforms. The brands we work with get 95% of what they need from three tools they already pay for.
Klaviyo. Build a segment called “Best Customers” — customers who have placed two or more orders in the last 180 days with an average order value above your site average.
Shopify Segments. Create the same “Best Customers” segment inside Shopify so you can cross-reference what Klaviyo is telling you.
The Avatar Spreadsheet. One tab for each of the five layers. One tab for customer voice quotes. One tab for numerical data. One tab for the finished avatar summary.
Setting Up Your First Customer Avatar (30-Minute Walkthrough)
Block out 30 minutes and work through it in order:
- Pull your top 20 repeat customers from Shopify.
- Create your “Best Customers” segment in Klaviyo.
- Install a post-purchase survey.
- Open a new Google Sheet with tabs for each avatar layer.
- Email your top 20 customers offering $50 store credit for a call.
- Schedule a weekly 30-minute review.
Real Example: How Frank Green Built Its Avatar Around a Moment
Frank Green’s best customer is someone who has just made a small lifestyle upgrade and wants accessories that match the aesthetic of their upgraded life. The product isn’t a cup. It’s a status signal. That understanding shaped everything — packaging like Apple, seasonal colourways, design-magazine Instagram. Over $100 million in annual revenue by 2023.
Another Aussie example: Who Gives A Crap. Their avatar isn’t “eco-conscious households.” It’s the person who’ll pay a premium for a product that lets them feel like their everyday purchases are building a better world, and who will brag about it on their toilet roll holder.
The lesson? Your avatar isn’t a demographic. It’s a moment and a motivation.
Applying the Avatar Across Every Channel
- Meta and Google ads. Your hook is the Trigger moment. Your body copy is the Pain. Your offer is the Desire.
- Product page copy. The first subhead addresses the Pain. The feature list addresses the Hesitation.
- Email flows. Welcome series built around Situation. Abandoned cart built around Hesitation. Post-purchase built around Desire.
- New product development. Check it against the avatar before you launch.
- Customer service responses. Use Customer Voice language for higher NPS.
The compound effect is what makes this work. One avatar, built once, informs dozens of decisions every week.
The Avatar Refresh: When and How to Update
Review your avatar every quarter. Refresh properly every 12 months. Full rebuild when:
- CAC has climbed more than 25% in 90 days
- Repeat purchase rate has dropped more than 10% year-over-year
- You’ve launched a significantly different product category
The Data Sources Most Founders Never Open
The difference between a guessed avatar and a real one is where the inputs come from. You already own more customer data than most agencies could buy. The problem is that it lives in five different tabs you never open together.
- Post-purchase surveys. One question at the thank-you page (“What nearly stopped you from buying today?”) gets 30% to 50% response rates, far better than the 10% to 25% you see from email surveys. Tools like KnoCommerce or Grapevine plug straight into Shopify. We cover the exact five questions in our post-purchase survey playbook.
- Klaviyo segments. Build a segment of your top 20% by lifetime value and read 30 of their profiles. Where did they come from? What did they buy first? How long between first and second order?
- GA4 demographics and Hotjar recordings. GA4 tells you who they are, Hotjar shows you how they shop. Watch ten session recordings of buyers and ten of bouncers and you will learn more than a month of brainstorming.
- Review mining. Pull your last 100 reviews plus your top competitor’s. The exact phrases customers use to describe the problem become your ad hooks, word for word.
Putting Avatars to Work in Meta Ads and Email
An avatar that sits in a Google Doc is decoration. The payoff comes when each avatar gets its own angle in your marketing. If you have three avatars, you should have at least three distinct ad concepts running, each speaking to a different buying trigger.
Take a skincare brand. The “ingredient researcher” avatar gets an ad built around clinical percentages and before-and-afters. The “time-poor mum” avatar gets a 30-second routine demo. Same product, two completely different hooks. Brands that match creative to avatar consistently see cost-per-acquisition drop 20% to 40% versus running one generic ad to everyone.
In email, map each avatar to a Klaviyo segment and split your flows. Your welcome sequence should convert 5% to 10% of new subscribers, and the fastest way to get to the top of that range is sending the researcher long-form education while the convenience buyer gets social proof and a clean offer. Layer in RFM segmentation once you have purchase history, and use your NPS survey responses to find which avatar produces your promoters. That is the avatar you scale.
Prove the Avatar With Numbers, Not Vibes
An avatar you cannot test is just a nicely written opinion. Before you rebuild your ad accounts around it, attach it to figures you already have sitting in Shopify and Klaviyo. If the avatar is real, the numbers will separate cleanly. If they do not, you have described a demographic, not a customer.
Build a simple cohort table with three columns: the segment that matches your avatar, everyone else, and the gap between them. Then fill in four rows.
- 90-day repeat rate. Across most Australian DTC stores this sits between 20 and 30%. Your avatar segment should beat the store average by at least 5 points. If it does not, you have picked the wrong group.
- 12-month LTV. A useful avatar usually carries 1.5 to 3x the LTV of the general base. Anything closer than 1.2x is not worth reorganising your marketing around. Our LTV playbook walks through calculating this properly rather than dividing revenue by customers.
- Discount dependency. Check what share of the segment’s orders used a code. Your best avatar should be less discount-dependent than average, not more. A segment that only buys at 30% off is a promotion, not a person.
- Return rate. Apparel sits around 20 to 30% and homewares closer to 5%. A segment returning well above your category norm is telling you the product promise and the buyer are mismatched.
Run this in a cohort view rather than a lifetime total, because lifetime totals flatter your oldest customers and hide what is happening with the ones you acquired last quarter.
The Four Ways Avatars Go Wrong
Most avatar work fails for predictable reasons. Knowing them up front saves you a quarter of wasted targeting.
You built it from your best customers only
Interviewing five raving fans tells you why people who already love you love you. It tells you nothing about the 96% who visited and left. Balance every advocate interview with at least one lapsed customer and one person who abandoned checkout. The uncomfortable conversations are the ones with the information in them.
You described a person instead of a purchase trigger
“Sarah, 34, yoga teacher, drinks oat lattes” is a character sketch. “Buys within 48 hours of her old one breaking, compares three brands on shipping speed, needs it before the weekend” is a buying pattern you can write ads against. Triggers beat traits every time, because triggers tell you what to say and when to say it.
You made one avatar and stopped
Most stores over about $500k a year have two or three genuinely distinct buying modes: the self-purchaser, the gift-buyer, and the replenisher. They want different copy, different offers and often different landing pages. One avatar forces you into copy that is vaguely acceptable to all of them and compelling to none.
You never went back to it
An avatar built in 2024 is describing a customer who was shopping in a different economy, on a different iOS privacy regime, with different competitors. Diarise a review every six months and after any major product or pricing change. Feed it fresh data from your post-purchase survey so the refresh takes an hour rather than a fortnight.
Your Avatar Action Plan
- ☐ Export top 20 repeat customers from Shopify
- ☐ Create “Best Customers” segment in Klaviyo
- ☐ Install post-purchase survey
- ☐ Set up avatar spreadsheet with 6 tabs
- ☐ Email top 20 customers offering $50 credit for a call
- ☐ Book and run at least 8 customer interviews
- ☐ Export 50+ reviews and mine for language
- ☐ Audit last 90 days of DMs and support tickets
- ☐ Fill in all 5 avatar layers with real quotes
- ☐ Rewrite one ad, one product page, and one email flow
- ☐ Schedule a 90-day check-in
Where This Fits Inside the More Orders Operating System
Prospects is one of the four foundations of the More Orders Operating System. If your Product is solid but your Prospects work is vague, every dollar you spend on Promotion is working harder than it needs to.
Inside the eCommerce Circle, customer avatar work is one of the first things we run every new member through. If you’re scaling a Shopify brand in Australia and you’re stuck — the fastest way forward is usually more clarity on who you’re actually selling to. Let’s talk about what your Prospects work looks like today.
