The Declining Portfolio: How to Spot Customers Before They Disappear — and Decide Which Ones Are Worth Keeping
Chris Daly, Founder of I Want That! — 25 years in retail, worked with 40+ ecommerce brands.
Who This Is For
Shopify operators in the $300K–$5M range who have customers showing up as active in their 12-month window but spending materially less than the prior year — and no clear system for deciding which ones to fight for and which ones to release. If your declining segment is getting the same promotional cadence as your growth segment, you are spending recovery dollars on customers who have already decided to leave.
What You Need Before You Start
- 24 months of transaction data with YOY net sales comparison by customer
- Attrition Migration Tracking output identifying Existing Decreasing customers — active in both P1 and P2, with P1 net sales below P2
- Lifetime contribution margin by customer if your system tracks it — gross sales minus discounts, returns, and a loaded cost of goods estimate
- Category purchase history to identify whether remaining engagement is in core or ancillary product lines
- Quintile scores based on current-period (P1) net sales
Estimated time: 3–4 hours to build the segmentation; 1 hour per quarter to refresh Difficulty: Intermediate to Advanced
The Honest Truth About Declining Customers
Most declining customers have already left. They just haven't stopped buying yet.
A customer who transacted four times last year and once this year is classified as Existing Decreasing. They are technically active. But their purchase behavior tells a different story — the brand relationship is contracting, the repurchase cycle has stretched, and the category engagement has narrowed. Often, that single remaining transaction is on a low-margin ancillary product. A pair of socks from a shoe brand. A cleaning kit from a camera store. The last thing they buy before they stop buying.
The instinct is to fight for them. The discipline is to ask first: fight for them with what resources, at what margin cost, and against what probability of recovery? Because the answer is different for a Q1 declining customer who spent $800 last year and $400 this year than it is for a Q5 customer who spent $50 lifetime and has a negative contribution margin.
The declining portfolio is the most margin-dangerous segment to mismanage. Blanket discounts, untargeted counter-offers, and promotional cadences designed for growth customers applied to declining customers is one of the most reliable ways to accelerate their exit while destroying your margin on the way out.
Step 1 — Score the Declining Segment and Calculate Contribution Margin
Pull your Existing Decreasing customers and sort by current-period (P1) net sales descending. Apply quintile scoring. Now append two additional data points for each customer: lifetime contribution margin and primary category of remaining purchases.
1.1 The Contribution Margin Screen
Lifetime contribution margin is gross sales minus discounts minus returns minus a loaded COGS estimate. If your system does not calculate this automatically, a spreadsheet approximation using average margin by category is sufficient for segmentation purposes.
Flag every customer with a negative lifetime contribution margin. This is not a small number on most files. Customers who have returned frequently, used coupons on every transaction, and whose remaining engagement is on low-margin products can easily sit at negative contribution despite being technically active. These customers are a cost center, not a revenue line.
The contribution margin screen changes the quintile picture materially. A Q3 declining customer with positive contribution margin is a different business problem than a Q3 declining customer at negative margin — even if their current-period spend is identical.
1.2 The Category Signal
For each declining customer, identify whether their remaining purchases are in your core high-margin categories or in ancillary low-margin products. A customer who used to buy shoes and accessories and is now only buying socks is not a shoe customer anymore. Their remaining transaction is completion behavior — they are buying the last thing they need from you before the relationship closes.
This matters enormously for CGO strategy. A counter-offer on a core high-margin product from a declining customer has recovery potential. A counter-offer on socks does not. You are discounting a margin-thin ancillary product for a customer who has already decided to leave — that is two losses on one transaction.
Step 2 — Design the Approach by Quintile
Q1 — Meaningful Recovery Investment
Q1 declining customers were your top spenders and are still spending more than most of your file in absolute terms. A Q1 customer who spent $800 last year and $400 this year is declining — but their floor is still above your file average. They warrant a real recovery investment.
The first question is why they are declining. The answer is almost always one of three things: a competitive alternative has captured part of their wallet, a service or product failure went unaddressed, or natural life-stage drift has reduced their need for your category. Each cause requires a different response.
Direct outreach: for Q1 declining customers, a personal email from the founder — not a triggered sequence, not a batch send — that acknowledges the relationship and asks a direct question: is there something we missed? This is not a discount offer. It is a service gesture. The response rate is low but the information value is high, and the customers who respond are often recoverable.
Recovery offer: if direct outreach does not produce a response within 14 days, a bespoke recovery offer sized to their prior AOV is appropriate. A threshold credit — spend $300, receive $40 — is preferable to a percentage-off because it anchors the transaction at a spend level closer to their historical average rather than enabling a small discounted transaction.
CGO for Q1 declining: accept and counter CGOs from Q1 declining customers on core high-margin products. Their offer will reflect declining brand engagement — expect offers 15–20% below retail rather than the 10–15% typical of growth customers. A counter that meets them at 10–12% off with free shipping is defensible on a customer who was spending $800 annually. Do not counter CGOs from Q1 declining customers on ancillary products. Accept full price or let the transaction close without engagement.
Q2 — Structured Intervention
Q2 declining customers have meaningful prior value and a recovery window that is not yet closed. They are worth a structured investment — more than passive tactics, less than the personal outreach appropriate for Q1.
Category re-engagement: pull their prior category purchase history and identify the core category where their engagement has dropped. A curated product recommendation in that category — not a sitewide sale, not a generic email — is the first touch. You are reminding them why they came in the first place.
Threshold offer: if the category recommendation does not produce a purchase within 21 days, a threshold offer anchored above their current-period AOV is the follow-up. You are not trying to close a single transaction — you are trying to reset the purchase cadence at a higher level than where it currently sits.
CGO for Q2 declining: same discipline as Q1 — counter on core products, decline to engage on ancillary low-margin items. The counter for Q2 can go to 12–15% off on core categories. On ancillary products, a polite full-price response or no counter is the correct answer. You are not in the business of discounting socks for customers who are leaving.
Q3 — Passive Intervention with a Hard Stop
Q3 declining customers are at the margin of recovery viability. Their prior spend was average, their current spend is below average, and the probability that a meaningful offer investment produces a profitable recovery is materially lower than for Q1–Q2.
The approach here is passive and time-bounded. One category re-engagement email. One follow-up with a modest offer at 30 days. If neither produces a purchase, suppress them from further promotional cadence and route them to the defected tracking pool at the next quarterly reporting cycle.
CGO for Q3 declining: this is where the math gets hard. A Q3 declining customer submitting a CGO on a core product is worth a light counter — 10% off, no additional incentives. A Q3 declining customer submitting a CGO on an ancillary product with a negative contribution margin history is not worth countering at all. Accept full price if they'll take it. If they won't, let the transaction go. The margin cost of countering a low-value offer on a low-margin product from a customer who is leaving is not recoverable.
The hard stop rule: if a Q3 declining customer has a negative lifetime contribution margin, do not invest in recovery. Not one email sequence. Not one offer. Route them to passive suppression. Paying to retain a customer who costs you money on every transaction is not a retention strategy — it is a recurring loss.
Q4 — Triage and Release
Q4 declining customers spent modestly in their prior period and less in the current one. The recovery math is almost never positive. A single low-cost email touch — a new arrivals notification in their prior category, no offer attached — is the ceiling of appropriate investment.
If they respond and purchase, process normally and route to the standard post-purchase flow. If they do not respond, suppress and move on. Do not build a Q4 declining recovery sequence. Do not counter their CGOs on anything other than core full-margin products, and only then with an acceptance rather than a negotiated counter.
The practical discipline for Q4: ask whether the cost of the intervention — email platform cost, offer cost, operational overhead — exceeds the expected margin recovery from the segment. For most brands at most platform costs, it does.
Q5 — The Negative Margin Decision
Q5 declining customers with negative lifetime contribution margin are the clearest case in portfolio management. They are not a retention problem. They are a cost center that is still technically active.
Do not counter their CGOs. Do not build recovery sequences. Do not include them in promotional cadence that carries offer costs. If they arrive inbound and purchase at full price, process the transaction. If they submit a CGO, particularly on ancillary low-margin products, decline to counter.
The hardest part of managing Q5 declining customers is accepting that releasing them is the right business decision. Every dollar spent on Q5 declining customer recovery is a dollar not spent on Q1–Q2 growth customer development, new customer acquisition, or Q1 reactivation. The opportunity cost is real and it compounds.
One exception: if Q5 declining customers are purchasing from your clearance inventory at full clearance price with no coupon and no return history, they are serving a markdown velocity function. Let them. Process the transaction without engagement investment and move on.
Step 3 — The CGO Decision Framework for Declining Customers
Customer-generated offers from declining customers require a two-variable screen before you decide to counter.
Variable 1: Product category margin. Is the CGO on a core high-margin product or an ancillary low-margin product? Core products warrant a counter. Ancillary products — the socks, the cleaning kits, the add-ons — do not. You are not in the business of negotiating your way to a lower margin on a product that was already thin.
Variable 2: Customer contribution margin. Is this customer at positive or negative lifetime contribution margin? Positive margin customers in Q1–Q3 warrant a counter on core products. Negative margin customers warrant no counter regardless of quintile or product category. Countering a negative-margin customer's offer on a low-margin product is the worst possible outcome — you are paying twice to lose money.
The decision matrix is simple:
- Core product + positive contribution margin customer → counter, scaled to quintile
- Core product + negative contribution margin customer → accept full price or decline
- Ancillary product + any customer → accept full price or decline, do not counter
- Ancillary product + negative contribution margin customer → decline engagement entirely
Common Mistakes
Mistake 1: Applying growth customer tactics to declining customers. Why it happens: the segmentation is not built and everyone gets the same campaign. Fix: build the Attrition Migration Tracking classification before you build any campaign. Declining customers receiving early access emails and category expansion offers designed for growth customers accelerates their exit — they are being marketed to as if nothing has changed when everything has.
Mistake 2: Countering CGOs on ancillary low-margin products. Why it happens: the instinct is to close every transaction. Fix: build the two-variable CGO screen — product margin and customer contribution margin — before any counter goes out. A declined counter on a sock CGO from a declining customer is not a lost sale. It is a margin-protected decision.
Mistake 3: Not identifying negative contribution margin customers before running recovery campaigns. Why it happens: contribution margin at the customer level requires more data work than most operators do. Fix: a spreadsheet approximation using average category margin applied to transaction history is sufficient for segmentation. You do not need a perfect number — you need a directional screen that separates positive and negative margin customers before you spend offer dollars.
Mistake 4: Treating all declining customers as recoverable. Why it happens: operators conflate active status with relationship health. Fix: a customer active in the 12-month window with a lengthening repurchase cycle, narrowing category engagement, and declining AOV has already made a behavioral decision. The data will catch up to them at the next quarterly reporting cycle. Act on the behavioral signal now, not on the active status flag.
FAQ
What is a declining customer in ecommerce portfolio management? A declining customer is classified as Existing Decreasing in Attrition Migration Tracking — active in both the current and prior 12-month periods, with current-period net sales below prior-period net sales on a year-over-year basis. They are technically active but contracting in value. The decline can reflect competitive displacement, product-market drift, service failure, or natural life-stage change. The key diagnostic question is whether the decline is recoverable — which depends on prior-period value, current contribution margin, and the category of remaining engagement.
How do I calculate contribution margin at the customer level? Customer-level contribution margin is gross sales minus discounts applied minus return credits minus a loaded cost of goods estimate for the products purchased. If your system does not calculate this automatically, apply your average margin by category to each customer's transaction history. Flag customers where cumulative discounts and returns have eroded margin below zero. This screen does not need to be precise — it needs to be directional enough to separate customers worth investing in from customers who cost you money to retain.
Should I counter a CGO from a declining customer? It depends on two variables: the product category margin and the customer's lifetime contribution margin. CGOs on core high-margin products from positive-margin customers in Q1–Q3 warrant a counter scaled to their quintile. CGOs on ancillary low-margin products — regardless of customer quintile — do not warrant a counter. CGOs from negative contribution margin customers do not warrant a counter regardless of product category. The discipline is running the two-variable screen before any counter goes out, not after.
How do I know if a declining customer is recoverable? Recovery probability correlates most strongly with three variables: the depth of the YOY decline (a 20% decline is more recoverable than a 70% decline), the category of remaining engagement (core product engagement is more recoverable than ancillary-only engagement), and the length of the prior brand relationship (a customer declining in year 4 of a 6-year relationship is more recoverable than a customer declining in year 1). Customers with negative lifetime contribution margin are rarely worth the recovery investment regardless of other variables.
What happens to declining customers at the next reporting cycle? Declining customers who do not respond to intervention and do not transact in the current period will reclassify as Defected at the next quarterly Attrition Migration Tracking run. This is the natural outcome for customers who have behaviorally exited. The goal of the declining portfolio playbook is not to prevent all defection — it is to recover the Q1–Q2 customers worth fighting for and release the Q4–Q5 customers whose recovery cost exceeds their recovery value before the defection reclassification happens.
Is it ever right to do nothing for a declining customer? Yes — for Q5 customers with negative contribution margin, doing nothing is the correct answer. No email sequence, no counter-offer, no promotional investment. If they arrive inbound and purchase at full price, process the transaction. If they submit a CGO on an ancillary product, decline to engage. The opportunity cost of Q5 declining customer recovery investment is real — every dollar spent there is a dollar not spent on growth customer development or new customer acquisition.
Knowing which declining customers are worth fighting for — and which ones are costing you money to retain — starts with seeing your portfolio clearly. Install Vector to build your Attrition Migration Tracking segmentation and run your first contribution margin screen →
Codex handoff notes:
- External citation slot 1: customer defection early warning signals — Bain, HBR, or similar source on behavioral indicators of customer exit before data reflects it
- External citation slot 2: contribution margin at customer level — any published source on negative LTV customer segments in DTC ecommerce
- Sibling links:
/blog/portfolio-growth-customers,/blog/portfolio-stable-customers,/blog/portfolio-reactivated-customers,/blog/portfolio-defected-customers— link bidirectionally when siblings ship - Internal link:
/blog/customer-portfolio-management(DE overview) and/blog/six-customer-portfolios-framework— reference from opening section - Image path:
/images/blog/portfolio-declining-customers-hero.png - Hub confirmation:
/customer-portfolios - Asset cut: the CGO decision matrix (two-variable screen: product margin × customer contribution margin) is the single best LinkedIn asset — a 2×2 grid, four cells, four decisions. One image, immediately actionable.
- Note: the negative contribution margin customer argument is the most counterintuitive claim in this post — flag for Chris to add a real client anecdote in the Backstory position if this gets expanded to a RE later