Why Ecommerce Brands Lose Repeat Customers

From first order to repeat purchase: what ecommerce brands get wrong about retention

Most brands begin working on retention the moment a customer pays. The retention panel at RARE 2026 argued that by then, half the job is already done, or already ruined. The offer that brought someone in, the promise in the ad, and the first emails after signup all shape whether a second order ever happens.

The panel, "Retention is broken: why ecommerce brands still lose customers after the first order," brought together three practitioners who look at repeat purchases from different angles:

  • Andrei Georgescu, Co-founder and Managing Partner at White Image Loyalty, one of Romania's email marketing and automation pioneers, with more than 23 years in loyalty and lifecycle projects
  • Geneviève Masioni, Retention and Email Strategist and former Shopify developer, who helps sustainable brands increase repeat purchases without relying on discounts
  • Maciej Raczak, Head of Operations at Widoczni agency, who researches online purchasing decisions through neuromarketing

Alex Danchenko, Co-founder and COO at Yespo, and Dmytro Kudrenko, Founder and CEO of Stripo, moderated.

Alex opened with a familiar scenario: a brand meets its customer acquisition cost targets, but six months later, most of those customers haven't bought again. Here's what the panel said about why, and what to do instead.

Key takeaways

  1. Whether a customer orders twice is decided partly at acquisition: by the offer and the promise that brought them in.
  2. Deep discounts recruit discount hunters. Run promotions with a plan for that cohort, or give people a reason to buy beyond price. 
  3. Judge customers by how they buy and how profitable they are, not by how often they order.
  4. When behavior data can't explain why customers leave, ask them. The answer is often something you can easily fix.

The second order is decided before the first one

Asked whether a cohort that never reorders is acquisition's fault or retention's, Geneviève called a discount-driven drop-off a retention problem. But her reasoning, like Andrei's, started at acquisition: it's the first thing a customer hears from you, so it's the first retention decision too.

Geneviève's first move with one-time buyers is to look at what happened right after they joined. What did the welcome flow send, and does it match what the ad promised?

The offer matters as much as the message. A steep first-order discount sets customers' expectations about what your product should cost, making the full price feel like a markup. Andrei put it in terms of motivation: every customer buys for a reason, and if the only reason you gave them was a discount, that's the reason they'll wait for. Other reasons are available. People buy because a brand means something to them, or because buying from it says something about them.

Andrei Georgescu, Co-founder and Managing Partner at White Image Loyalty

 

"If you have customers who buy from you only when you have a discount, that's not necessarily a loyal customer. That's a discount hunter."

To tell a source problem from a post-purchase problem, Andrei compares channels. Say customers acquired cheaply through Facebook buy once and disappear. If customers from other channels have good retention rates, the problem is the acquisition source. If every channel shows the same drop-off, look at the product and at the service customers get after they buy.

Black Friday is a trade-off, so make it on purpose

Dmytro Kudrenko raised the obvious objection: Black Friday is one of the year's biggest acquisition windows, and it runs on discounts. Is every customer acquired then a lost cause?

Not if you plan for them. Andrei's position was the moderate one: discounts can have their moment, but they shouldn't be the foundation of customer lifetime value. Geneviève was more direct about the cost.

Geneviève Masioni, Retention and Email Strategist

 

"When you run a Black Friday campaign, you're making a trade-off. You're trading short-term revenue for lower-quality customers, because they enter the relationship with a massive discount."

She sees two ways to handle it:

  • Run it, then isolate the cohort. Keep customers acquired during the promotion apart from everyone else and put them through a dedicated nurture sequence that gets them used to your full prices.
  • Skip the public discount. Many of the brands she works with don't run Black Friday sales at all. They still create momentum by increasing the share of each order donated to charity and communicating this to customers.

Geneviève Masioni, describing the message her brands send instead of a discount

 

"During this Black Friday, more of your purchases will go to a cause we both care about."

Andrei's version of the same idea: when a business donates a share of every sale, customers get something a discount can't give them, the sense that their purchase helps someone.

Segment by how people buy, not by how often

Alex then asked the harder question. With purchase, engagement, and loyalty data in hand, how do you tell a customer who was a bad fit from one you simply served badly?

Andrei starts with profitability rather than order count. From there, he identifies what motivates each segment to return: 

  • High-value customers who are starting to drift respond to recognition.
  • New customers need a sense of progress and a clear next step.
  • Others respond to exclusivity or scarcity.

Geneviève went further and questioned the idea of a "bad" customer at all. In her clients' data, some people placed two orders with five items each, while others placed seven orders with just one item each. Some bought only during Q4 promotions. Each pattern represents a buying style worth serving.

Geneviève Masioni, Retention and Email Strategist

 

"Just because someone doesn't buy multiple times doesn't mean they're not the buyer we want."

So she segments by buying style (bulk buyers, frequent buyers with a lower average order value, promotion-only buyers) and sends each what they want to hear, instead of pushing everyone toward the same 30-day reorder cycle.

Asked by Dmytro how she manages so many segments, Geneviève said the discipline is in what she doesn't build: a segment exists only when it captures a real behavior, has enough people, and shows potential.

Only when a contact has gone through every flow (re-engagement, win-back, sunset) without reacting does she call them the wrong customer and remove them from the list. Removing them also protects sender reputation.

Build segments around buying behavior

In Yespo, dynamic segments can use purchase events and their parameters, such as order count, order value, brand, or category, over any period. Streak conditions find customers who buy in every week or month, and first-event and last-event conditions separate new buyers from lapsing ones. 

Show me how to segment by purchase behavior →

Ask first, then teach

Behavior data shows what customers did, but it rarely shows why. Andrei's example came from a brand selling mineral foundation, a product people should reorder when it runs out. Customers bought once and didn't come back. When the brand finally asked them why, the answer had nothing to do with price or competitors: many didn't know how to apply the product correctly and were unhappy with the result. The brand's response was a post-purchase email linking to a video demo of how to apply it.

The lesson is to build feedback into the customer journey, with several points where you ask customers about their needs and frustrations. Andrei also reframed what retention messages are for. Buying is the final action you want, but along the way you want other behaviors too: telling you who they are, how they use the product, and what they need. Each of those makes the next message more relevant.

Maciej added a reminder that data alone won't get you there:

Maciej Raczak, Head of Operations at Widoczni

 

"There is a saying in social psychology: different people behave differently in different situations because of different reasons."

His point was that teams focused on transactions lose sight of the emotions that come before a decision.

Turn feedback into a segment

Yespo widgets include NPS and rating components (0–10 scale, five stars, emojis, or like/dislike). A multi-step widget can show a different next page depending on the score, so you can send detractors to a follow-up question and promoters to a review request. Scores are saved to the contact profile or as event parameters, so you can build segments from them. 

I want to know more about NPS for my site →

Three retention myths the panel took apart

To close, Alex asked each panelist to finish one sentence: "The biggest lie brands tell themselves about retention is…"

Myth 1: More flows mean more retention. Andrei's answer was that brands treat retention as a communication problem and respond by building more automation.

Andrei Georgescu, Co-founder and Managing Partner at White Image Loyalty

 

"Retention is fundamentally about behavior, not about automation. Those are tools. They don't build loyalty by default."

His test: before sending any message, name the behavior it's meant to produce. Dmytro agreed, borrowing a principle from the Agile Manifesto: "individuals and interactions over processes and tools."

Myth 2: A good product retains customers on its own. Geneviève sees this often among sustainable brands that avoid discounts. They assume that without a discount they have nothing to say, so they lean on the product. But customer research tells you exactly what to say. You still need post-purchase flows, and they need to answer the objections and hesitations that stop people from buying again. Taken together, their points suggest that flows need a clear purpose, while even a good product may need post-purchase communication to encourage repeat purchases. 

Myth 3: We know our customers. Maciej called it the curse of knowledge.

Maciej Raczak, Head of Operations at Widoczni

 

"We are very often not looking at the product or service with the same pair of eyes that our customers do. Their decisions are not made along the path to purchase that we would choose as experts."

That gap, he said, creates bias in brand communication, and the bias costs retention. Dmytro summed it up: people don't behave logically.

Repeat purchase checklist

Run these on your customer base before you add another win-back flow:

  1. Compare repeat purchases by acquisition source.
  2. Read your ads and your welcome flow side by side. Anything the ad promised should show up in the first emails.
  3. Flag customers acquired through a deep discount. Give them their own nurture path toward full price instead of your standard flow.
  4. Decide on Black Friday deliberately. Either run it with a plan for the cohort it brings in or build momentum without a public discount.
  5. Score customers on profitability, not order count.
  6. Segment by buying style and motivation. Keep only the segments that capture real behavior and are big enough to be worth running.
  7. Ask one-time buyers why they didn't come back. If the answer is "I didn't know how to use it," add a how-to step to your post-purchase flow.
  8. Give every flow a behavior to produce. If you can't name it, question the flow.
  9. Remove inactive contacts only after they've completed your re-engagement, win-back, and sunset flows without responding. It keeps your list clean and your sender reputation intact.

Wrapping up

Asked for one piece of advice each, two panelists answered in a sentence or two. Andrei repeated his central point: retention starts before the first order, so look at everything you do before a customer buys. Alex added that the work is in connecting what you learn there to every message that follows. Geneviève kept hers practical:

Geneviève Masioni, Retention and Email Strategist

 

"Listen to your customers. And if you don't know what they're thinking, ask them. Keep iterating, keep collecting data, and never stop."

Recordings of all sessions are available on the RARE 2026 website. For the conference's AI panel, read our article.

Special Request Inline

Was This Post Helpful?

0.0 from 5 based on 0 reviews

Ivan Diulai

Copywriter

Top

Ivan Diulai

Copywriter

Comments 0