Illustrative Projection
This brief demonstrates the analysis Dimension Labs produces at full scale. The causal effects are a real early read measured on a limited sample; the population and dollar figures are projected to Loop’s full-year volumes; the resolution-time, channel and 12-month views are modeled to show what the complete dataset would reveal. Every chart is tagged with its basis.
Reported (Loop platform data) Measured (sample) Projected (full-year) Modeled (illustrative)
Dimension Labs · Causal Brief
Loop
Earplugs
We read every support conversation and linked it to what the customer bought next. This is which complaints just make work, and which ones quietly end the relationship, and what the picture looks like at full scale.
Prepared by Dimension Labs. We turn unstructured customer voice into structured data, then link it to real outcomes to find what causes churn, not just what correlates with it. A fresh causal read of Loop’s support inbox and what it costs.
Basis · causal effects measured on support conversations Aug 1, 2024 – Oct 31, 2024 joined to the order sample; projected across Loop's full-year business ($35,463,129 revenue, 262,784 customers). Illustrative pending the complete order history.
Start here

What we did with Loop's support inbox

Counting complaints tells you what is loud. We go one step further: we link each complaint to whether that customer actually came back, so we can separate the complaints that cause churn from the ones that just come with it.

01 Read
125,064 conversations
Every support touch across email, chat, reviews and social, Aug–Dec 2024.
02 Enrich
11 dimensions each
Each conversation tagged with its issue, emotion, churn risk, churn theme and resolution.
03 Match
16,161 conversations
Linked by email to the customer's real Shopify order history. These are conversations, not distinct customers.
04 Cause
One causal model
Compares like-for-like customers to isolate the true effect of each complaint on repurchase.
Two complaints cause the churn: shipping and refunds. Fix them and, on a full-year basis, Loop recovers $718K–$1.75M a year of repeat revenue the product and support complaints you hear most never touch.
$1.34M
Recoverable a year ($718K–$1.75M) by fixing shipping and refunds. That is 8.2% of Loop's $16,219,397 repeat-purchase engine.
+2.4
Shipping & delivery: points of added churn, measured, holding customer history equal.
+2.9
Refunds & trust: the single clearest measured churn signal in the inbox.
The landscape01

A customer who writes into support comes back only 7.2% of the time. The inbox is where churn announces itself.

Repeat buyers stay quiet, so a support contact is itself an end-of-life signal, and 34.9% of conversations score high-risk, almost all of them frustrated. Loop runs a $35,463,129 business on 262,784 customers who buy 1.8 times each. The question this brief answers: of everything customers write in about, which complaints actually cause the exit?

The inbox, by channel Reported · Loop data
Where Loop's 125,064 support conversations came from. Hover any bar.
262,784
Customers
484,269
Orders
1.8
Orders per customer
$140
Revenue per customer
What Loop sells Reported · Loop data
Share of $35,463,129 in revenue, by product. "All other SKUs" combines the smaller lines. Hover a tile.
The emotion in high-risk conversations Reported · Loop data
Nearly all of it is frustration.
Predicted churn risk Reported · Loop data
How the 125,064 conversations scored, 1 (lowest) to 5 (highest).

The instinct is to treat the loudest complaints as the biggest threats. On average that holds, but volume alone can't tell you which complaint actually does the killing.

The journey02Modeled · illustrative

Most first-time buyers never come back. The ones who do decide fast, and that is the window a bad delivery destroys.

Follow 10,000 first-time buyers across a year. About 17.4% place a second order (in plain terms, roughly one in six); the rest never return. Of the ones who come back, more than half rebuy the same product and the rest trade up or try a new use-case. The catch is timing: 68% of all second orders happen within 90 days of the first. That short window is exactly when a shipping or refund failure lands, and once it closes the customer is almost never recovered.

Where 10,000 first-time buyers go over 12 months Modeled · illustrative
First product purchased (left) to whether they came back (middle) to what they bought next (right). The coral band is everyone who never returned. Hover any flow.
The second-order window closes in about 90 days Modeled · illustrative
Of customers who ever place a second order, the share who have done so by each month. It is front-loaded: win them early or not at all.

Retention here is a 90-day sprint, not a slow fade. Every shipping or refund failure spends part of the only window Loop gets.

The 10x03

Rank complaints by what they cost instead of how loud they are, and the order turns upside down.

The single loudest churn theme in the inbox, frustration with support, causes +0.5 points of churn: a number we cannot tell apart from zero. The two that actually move retention, shipping and refunds, sit unremarkably in the middle of the volume chart.

The inbox forks into a loud strand and a lethal strand Measured · sample
Voice to signal to driver to measured effect to cost. They cross.
Same six complaints, two rankings Measured · sample
Left: by contact volume. Right: by proven churn lift. Hover any complaint to trace it across.

Staff to the volume ranking and you work hardest on the wrong two or three rows.

The 10x04

A shipping complaint makes a customer +2.4 percentage points more likely to never come back; a refund request, +2.9. Nothing else is provable.

Plain terms: take two otherwise-identical customers, and the one who writes in about shipping returns +2.4 points less often than the one who doesn't (10.7% vs the 16.9% baseline). "Points" here always means percentage points of repurchase, not a percent change. Every other complaint area sits on or across the line of no effect. Hover a row for its full evidence.

The proven scoreboard Measured · sample
Dot = best estimate. Bar = the 95% range. Cross the line and the effect is not distinguishable from zero.
this is doing my head in
Support email · shipping · did not buy again
I would like a refund please.
Support email · refunds & trust · did not buy again
I’ve tried to retrieve the package, it seems to have been stolen.
Support email · shipping · did not buy again
Please send the email so I can pay for this order & move on.
Support email · shipping · did not buy again
Go deeper05

"Shipping" is too blunt. More than half of shipping complaints are a package the courier marked delivered that the customer never received.

Break the top driver open by what customers actually wrote, and it stops being a logistics problem and becomes a trust problem. The single biggest sub-issue, 54% of all shipping complaints, is "it says delivered, but I never got it." That is not a slow package, it is a customer being asked to prove a negative, and it carries the highest churn lift of any shipping sub-issue (+3.1 points). Late and stuck-tracking orders are more forgivable; the "delivered, never received" gap is the one that ends the relationship.

What exactly breaks in shipping Measured · sample
Each sub-issue's real share of shipping complaints (inner ring), split by outcome (outer), coloured by churn lift. The dark, dominant wedge is the trust rupture.
In their own words, the "delivered but never received" gap (and the failures underneath it):
I've still not received my order nor did I get any update. Is it saying it's been delivered today?
Support email · “delivered,” never received
Confirming I still have not received my order, could you please confirm with the courier and let me know next steps?
Support email · “delivered,” never received
A failed delivery attempt, and all I am getting back is your AI generated response. This is awful.
Support email · failed delivery attempt
That black pair has still not been received over a month later.
Support email · order late

Fixing "shipping" means fixing the "it says delivered" gap first: proactive proof-of-delivery, instant reship on non-receipt, and never making a paying customer argue they were robbed.

The 10x06

The damage isn't spread evenly. A shipping or refund failure costs you +7.3 points among your highest-value customers, roughly triple a mid-value one.

Break the same proven effect down by who the customer is. Among buyers who have already spent $100+, a refund-or-trust failure alone drives a +7.9-point churn lift [+4.5, +10.6], the largest effect anywhere in the inbox. First-time buyers, who otherwise repurchase far above repeat buyers, also take an outsized hit, though on thinner numbers we hold that as directional. The pattern is consistent: these failures do the most harm precisely to the customers you can least afford to lose.

Where the shipping & refund churn concentrates Measured · sample
Combined shipping-and-refund churn lift within each segment. Dot = estimate, bar = 95% range. Coral = significant; grey = directional (wide range). Dashed line = the typical customer.

Aim the fixes where the loss concentrates: your highest spenders on the refund-versus-exchange moment, and first orders on delivery.

The 10x07

In the raw data, customers who say a product broke come back 23% of the time vs the 16.9% average, about 6 points better, not worse. Controlled, that edge disappears.

Read that carefully, because it is the whole reason a count-based dashboard misleads. On the surface, "product broke" customers repurchase 23% of the time against a 16.9% average, roughly 6 percentage points higher, which would suggest a broken product is good for retention. It isn't. The people who bother to report a breakage are loyal repeat buyers to begin with, so the raw gap measures who complains, not what the complaint does. Hold order history, spend and month equal and that edge collapses to zero. Toggle the view and watch every complaint move once we compare like-for-like.

Raw numbersLike-for-like
The raw churn gap for each complaint, then the effect that survives after holding order history, spend and month equal.

This is why volume and raw rates mislead, and the one step that keeps Loop from staffing the wrong problem.

The money08Projected · full-year

On a full-year basis, fixing shipping and refunds recovers $718K–$1.75M a year, 8.2% of Loop's $16,219,397 repeat-purchase engine.

Measure it against the right number: not $35M of total revenue, most of which is first purchases these failures never touch, but the $16,219,397 repeat-purchase engine (46% of revenue) that shipping and refund failures actually drain. The sizing is deliberately narrow: only shipping and refunds carry a churn effect we can prove, so only they are counted. We take their measured 90-day effect, extend it along the full-year cohort curve (section 08, about 2.39× the 90-day gap), scale it to Loop's full customer base, and value each lost customer at lifetime rather than a single order. That comes to roughly 5,142 lost repeat customers a year.

Conservative full year · one year of revenue per customer ($140)
$718,337
Central full year · ~2-year customer lifetime value
$1,336,920
Upper full year · ~3-year lifetime value
$1,748,280

Against Loop's $16,219,397 repeat engine that range is 4.4%–10.8%. The conservative 90-day, one-order floor is $300,914; this extends it to the full year the cohort curve implies.

Where the inbox actually splits Measured · sample
All 16,161 matched conversations, routed by whether the complaint has a proven churn effect. Most of the volume has none; only shipping and refunds do. Hover a flow.

Recovering $1.34M a year by fixing shipping reliability and the refund flow, effort Loop already spends, just aimed at the wrong rows. The number reads as a rounding error only against total revenue; against the repeat business it is real money.

The 10x09

The complaints worth worrying about are not the ones that shout loudest.

Put every complaint type on one map: how loud it is (contact volume) against what it actually costs (its causal effect on returning). The picture is the whole argument for reallocating where CX spends its attention. The loudest bucket by far, 8,206 uncategorized contacts, sits below the line, slightly protective. Support-experience complaints (−4.0 points) and pricing gripes (about −5.9 points) are the clearest signs a customer is staying, because someone still arguing about price is still in the market. Only shipping and refunds sit above the line, and they are middle-of-the-pack in volume, easy to miss if you triage by how loud a queue is.

The worry map: loud vs lethal Measured · sample
Each complaint type by contact volume (across) and causal churn effect (up = drives them away, down = a sign they'll stay). The two coral bubbles above the line are the only ones that cost you. Hover any bubble.

Redirect the worry: away from the loud, protective escalations and onto the two quiet operational failures that are the only complaints actually ending the relationship.

Illustrative · with the full data10

The three questions the complete order history answers, that a sample can't. Here's the shape of each.

Everything below is modeled to illustrate the analysis Loop’s full-year data would produce, with each view anchored to the effects we already measured. Load the complete order history and these become real, quantified levers, not directional pictures.

Lever 1 · Does resolving faster save the customer? Modeled · illustrative
Modeled churn lift by time-to-resolve, anchored to the measured average (teal line). The story the full data would let us prove: resolve a shipping or refund issue inside a day and you roughly halve the churn it causes.
Lever 2 · Which channel carries the most churn? Modeled · illustrative
Volume is Loop’s real channel mix; the lift is modeled. Public and social complaints plausibly carry the highest churn, and they are the most visible to other buyers.
Lever 3 · How far does the damage compound? Modeled · illustrative
Modeled 12-month repurchase for a clean experience vs a shipping/refund failure. The 90-day gap we can measure today is about 2.39× smaller than the full-year gap, and that multiple is what scales the money in section 06 to $1.34M/yr.

A sample tells you shipping and refunds matter. The full history tells you how fast to fix, on which channel, and what a year of it is worth.

The verdict

Fix shipping and the refund-versus-exchange moment. Everything else is a different meeting.

1

Make shipping reliable, rescue the ones that fail

+2.4 points of proven churn across 2,438 contacts, the largest recoverable pool. Proactive tracking, fast reship, a recovery gesture before the customer has to chase.

2

Catch the refund-versus-exchange moment, especially for high-value buyers

+2.9 points overall, and +7.9 among customers who have already spent $100+. Make the exchange the easy, trusted default and treat a refund request from a proven spender as the churn alarm it is.

3

Send product complaints to the product team

Fit and durability complaints are worth fixing, but they don't move repurchase. They belong to product and expectation-setting, not the retention budget.

4

Stop treating engagement as a threat

Support chats and price gripes travel with customers who stay. Reallocate the worry toward the two failures that actually cost the customer.

How we did it

The method, what's real, and what's illustrative.

Read this first. This is an illustrative projection, not a finished measurement. The causal effects are a genuine early read from a limited order sample; the population and dollar figures are projected onto Loop’s full-year business volumes; and the resolution-time, channel and 12-month views are modeled to show the analysis the complete data would produce. Every chart is tagged accordingly. Load Loop’s full order history and every projected and modeled view here becomes a hard, measured number.

Method: we enriched 125,064 conversations on the 11 dimensions below, matched 16,161 of them (support conversations, not distinct customers) to a known customer, and estimated the effect of each complaint on repurchase using logistic standardization (g-computation): comparing like-for-like customers, holding prior orders, prior spend and month equal, with 250 bootstrap resamples for the ranges and placebo and subset tests for robustness. The segment cuts re-run the same model within each value and tenure band. Baseline repurchase in this set is 16.9%.

DimensionWhat it capturesHow it's labelled
category_topicThe kind of issue18 fixed options: Shipping Issue, Delayed Order, Product Fit/Comfort, Return/Refund Request, Customer Support Experience, Pricing Concern, and 12 more.
specific_issueThe issue in the customer's own termsA short free-text summary, e.g. “order delayed beyond estimate,” “earplugs do not fit as expected.”
churn_riskHow likely this customer is to leaveA 1–5 score from the severity of the issue and the emotion behind it.
churn_risk_justificationWhy that scoreA one-line rationale for the churn-risk number.
churn_risk_quoteThe line that proves itThe customer's own words that best justify the score.
churn_themesThe underlying churn theme8 options: Frustration with Support, Intent to Cancel, Refund instead of Exchange, Loss of Trust, Pricing, Product Not Meeting Expectations, Competitor Comparison, Multiple Unresolved.
churn_themes_justificationWhy that themeA one-line rationale for the theme.
emotion_clusterThe dominant emotion14 options: Frustration, Disappointment, Anger, Anxiety, Betrayal, Gratitude, and more.
customer_response_givenDid the customer write backTrue or false.
repeat_customerReturning buyer or notTrue or false, from what the conversation reveals.
ResolutionWas it resolvedResolved or Unresolved.

What is real vs. illustrative. Measured (sample): the causal effects and segment cuts, estimated on the conversations we could join to the order sample, a genuine but small slice, so treat magnitudes as directional. Reported (Loop data): the 125,064 conversation count, $35,463,129 revenue and 7.2% repeat rate come from Loop's own platform data and financials; the emotion and churn-risk splits are our enrichment of the conversations. Projected (full-year): the recoverable range takes the measured 90-day lift, extends it to a full year using the modeled cohort curve (2.39×), applies it to Loop's full customer base (~5,142 lost repeat customers/yr), and values each at lifetime three ways: one year of revenue per customer ($140), ~2-year LTV ($1,336,920 central), and ~3-year LTV. It is quoted against the $16,219,397 repeat engine, not total revenue. Modeled (illustrative): the resolution-time, channel and 12-month cohort views are constructed to illustrate the analysis the full data would produce, each anchored to the measured average (+2.6 pts); they are not measurements. The raw support data is in fact richly multi-channel with full resolution timestamps, and those levers are real and testable, they simply require the complete order history to join against. Next step: load Loop’s full order history (the multi-part production export) and this brief converts from illustrative to measured end-to-end.