Benchmark 2026 · Chile

Abandoned cart recovery via WhatsApp

11 e-commerce accounts · Shopify and WooCommerce · Chile · 30-day window ending September 1, 2026 · 2,262 carts contacted, 617 recovered.

Download benchmark PDF

How ROAS is calculated

ROAS = conversion × ticket ÷ 94

The cost per template is identical across all 11 accounts: 94 CLP, no exceptions. Nobody has a cost advantage, so the whole ROAS range of the accounts with volume, from 80x to 728x, comes from just two levers: how many carts get recovered and how much each one is worth.

Before reading these figures. This is not a statistical study. It is eleven accounts observed over a single thirty-day window: a universe too small to talk about significance, closer to organized casework than to a representative sample of the market. What it does deliver is what actually happened in these accounts, measured with the same criteria for all of them. We share it because we believe it helps you know where you stand against the rest of the portfolio and decide with that in view. Read it as a reference for conversation, not as an industry standard.

Where the data comes from. The “Recoveries by Eclectic” card on each account's dashboard, Recoveries tab: only sales directly attributed to the recovery message, with no assisted recoveries. It is the most conservative of the three readings the dashboard offers. Accounts are anonymized; each client recognizes their own.

What goes into the ranges. Percentiles are computed over the 9 accounts with at least 30 carts contacted in the window. The other two (2 and 10 sends) appear in the tables marked with †, but not in the ranges: at those volumes a single order moves the rate by more than ten points.

01 / Summary

The portfolio in four numbers

Conversion

27.3%

617 of 2,262 carts contacted

Recovered sales

40.4M

CLP in 30 days, 11 accounts

Value per cart

17,956

CLP recovered per cart contacted (median)

Portfolio ROAS

191x

40,429,706 CLP over 211,500 CLP spent

02 / Conversion

Conversion by account

Carts recovered over carts contacted, 30-day window.

median 27.0%

C05
50.0%
1 of 2 carts — small sample
C04
34.5%
58 of 168 carts
C09
34.1%
14 of 41 carts
C02
33.7%
93 of 276 carts
C07
27.3%
12 of 44 carts
C03
27.0%
292 of 1,083 carts
C01
23.7%
71 of 300 carts
C08
23.7%
37 of 156 carts
C06
21.5%
28 of 130 carts
C10
20.0%
2 of 10 carts — small sample
C11
17.3%
9 of 52 carts

The band is narrow: 17.3% to 34.5% across accounts with volume, with the median at 27.0%. Hatched bars have fewer than 30 sends and are not comparable.

03 / ROAS

ROAS does not measure recovery

Since the cost per message is the same for everyone, an account's ROAS is determined by its conversion and its ticket. The ticket, however, is not set by the recovery: it is set by the catalog. Comparing clients by ROAS is comparing them by the price of their products.

In the chart each dot is an account. The dashed lines join combinations that yield the same result: moving along one of them changes nothing.

Accounts with 30 or more carts contacted. Both axes on a log scale, which is why the equal-result curves are straight lines. Moving up the vertical axis means changing catalog; moving right is the recovery work.

The case that sums it up. C11 has the lowest conversion in the study (17.3%) and the second best ROAS (728x), only because its ticket is 372,742 CLP. C08 has the worst ROAS (80x) with a normal conversion (23.7%, in the lower quartile but within the band) and a ticket of 31,391 CLP. If C11 brought its conversion up to the portfolio median, it would recover 1,871,284 CLP more in thirty days without sending a single extra message.

What to look at instead

Conversion to judge the recovery work, because it is the only thing that depends on the message, the timing of the send and the conversation that follows.

Value per cart contacted (recovered sales divided by carts contacted) as the single number to size the opportunity: it combines both levers without letting the ticket dominate the picture, and multiplied by the monthly abandonment volume it gives directly what is at stake.

ROAS answers a single question, and it answers it decisively: whether the channel pays for itself. The worst account in the study returns 80 times what was invested.

04 / Reference

Reference ranges

Metric	How it is calculated	Min	P25	Median	P75	Max
Conversion	carts recovered / carts contacted	17.3%	23.7%	27.0%	33.7%	34.5%
Recovered ticket	recovered sales / carts recovered	31,391 CLP	52,585 CLP	58,172 CLP	68,848 CLP	372,742 CLP
Value per cart contacted	recovered sales / carts contacted	7,445 CLP	12,982 CLP	17,956 CLP	23,040 CLP	64,513 CLP
ROAS	recovered sales / template spend	80x	139x	191x	245x	728x

Account-by-account percentiles over the 9 accounts with 30 or more carts contacted, unweighted by volume: the middle half of accounts falls between P25 and P75.

05 / Opportunity

Value recovered per cart contacted

C11
64,513
17.3% × 372,742 CLP
C02
23,199
33.7% × 68,848 CLP
C04
23,040
34.5% × 66,736 CLP
C06
20,419
21.5% × 94,805 CLP
C09
17,956
34.1% × 52,585 CLP
C03
15,684
27.0% × 58,172 CLP
C01
12,982
23.7% × 54,853 CLP
C07
11,393
27.3% × 41,776 CLP
C08
7,445
23.7% × 31,391 CLP

Recovered sales divided by carts contacted, with the breakdown into its two factors to the right of each bar. Portfolio median: 17,956 CLP for every abandoned cart that gets contacted.

06 / Detail

The 11 accounts

Account	Platform	Carts contacted	Recovered	Conversion	Ticket CLP	Value per cart CLP	Recovered sales CLP	ROAS
C05 †	Shopify	2	1	50.0%	735,500	367,750	735,500	3,912x
C11	Shopify	52	9	17.3%	372,742	64,513	3,354,682	728x
C02	Shopify	276	93	33.7%	68,848	23,199	6,402,905	250x
C04	WooCommerce	168	58	34.5%	66,736	23,040	3,870,675	245x
C06	Shopify	130	28	21.5%	94,805	20,419	2,654,529	217x
C09	Shopify	41	14	34.1%	52,585	17,956	736,188	191x
C03	Shopify	1,083	292	27.0%	58,172	15,684	16,986,171	167x
C10 †	Shopify	10	2	20.0%	65,843	13,169	131,686	140x
C01	Shopify	300	71	23.7%	54,853	12,982	3,894,587	139x
C07	Shopify	44	12	27.3%	41,776	11,393	501,311	121x
C08	Shopify	156	37	23.7%	31,391	7,445	1,161,472	80x

† Fewer than 30 carts contacted: shown for completeness, but excluded from every range and percentile in this report.

07 / Scale

Scale does not wear down conversion

In mass campaigns to a cold base, performance drops as you keep hitting the same list. That does not happen here: the highest-volume account in the study contacts 1,083 carts in thirty days and converts 27.0%, exactly the portfolio median.

Volume tier	Accounts	Carts contacted	Conversion	Ticket CLP
Under 30 carts	2	12	25.0%	289,062
30 to 99	3	137	25.5%	131,205
100 to 299	4	730	29.6%	65,230
300 or more	2	1,383	26.2%	57,523

Weighted conversion within each tier. The ticket rises in the small tiers because that is where the expensive-product accounts concentrate, not because of volume.

08 / Delivery

Delivery is not the bottleneck

Abandoned carts this study

94.2%

100.0%

Mass campaigns to a cold base Aug-2026 benchmark

74.8%

83.3%

60%70%80%90%100%

Templates delivered over carts contacted, range across accounts. The abandoned-cart message reaches someone who just left their number on the site: that is why delivery starts at 94.2% and its median is 100%, while a mass campaign over a cold CRM base moves between 74.8% and 83.3%. They are two different products and should not be compared with each other.

09 / Conclusions

Five conclusions

1
Conversion moves in a narrow band, from 17.3% to 34.5%, with a 27.0% median. It is the metric that measures the recovery work and the only one that allows comparison between accounts.
2
The cost is identical for everyone: 94.00 CLP per template. No account has a cost advantage or disadvantage, and ROAS is entirely determined by conversion and ticket.
3
ROAS ranks by ticket, not by quality. It confirms the channel pays for itself (between 80x and 728x in this study), not who recovers best.
4
Scale does not penalize. Between 30 and over a thousand carts contacted per month, conversion stays around the median.
5
Delivery is solved (median 100%): the room for improvement is not there.

10 / Scope

What this study does not show

A single 30-day window. It cannot distinguish a real improvement from a monthly fluctuation. Two consecutive windows per account can.
Direct attribution only. Purchases where the message helped but the close came through another path are left out. The real recovery figure is higher than this report's, never lower.
No control for industry or product type. Part of the conversion difference between accounts may come from what is sold rather than how it is recovered.
It does not look inside the flow. How many touches each cart gets, how long the first one takes, what the template says and what happens in the conversation afterwards: that is where the 17 points separating the best account from the last one are explained, and it is the first thing to measure in the next version.

How to read your own position. Find your account in the 11-account table and compare your conversion against the 27.0% median first: that difference, multiplied by your carts contacted and your ticket, is what is on the table this month. Only then look at ROAS, and only to confirm the channel pays for itself.

Cart recovery with Eclectic

Want to know where your store would land in this table?

We connect your Shopify, WooCommerce or Jumpseller store and in thirty days you have your own recoveries card to compare against the portfolio.

Download benchmark PDF

Study prepared by Eclectic · anonymized accounts, aggregated data · September 2026 · try-eclectic.com