CASE STUDY · ARGYLE WINERY × PREMIER CRU
How Argyle and Premier Cru drove 58× email revenue per recipient in one week
What if we gave each customer an offer that aligned with their behavior?
3%
of recipients · targeted
+77%
web revenue vs the full list
58×
web revenue per recipient
EXECUTIVE SUMMARY
Aligning the offer with behavior
Tasting room staff are great at delivering a personalized experience: they build rapport, listen, surprise and delight, and sell more wine because of it. The best are building a picture of the people in front of them and figuring out what offer to make, based on how the customer is reacting. In other words, the best offers are made just in time, based on a person’s behavior.
Argyle’s team does this so well in the tasting room and wanted to bring the same experience to their email marketing.
Like many wineries, Argyle ran their email in three standard modes:
Full list — the same offer made to everyone reachable on their list
Product-specific — offering certain wines to people who have bought those SKUs before
Club — special offers for club members, because they’re special
Argyle runs their e-commerce, club, and tasting room through Commerce7 and uses Klaviyo for email marketing. Both are needed to analyze customer behavior, but exporting CSV files and building pivot tables is tedious, and the result is out of date the moment it’s finished.
Their partners at Premier Cru know how to build these analyses by hand, to figure out which customer segments should receive a different offer, but it takes days to assemble and is only current the day it’s finished. Marketing is iterative, and iteration stalls out when it takes days to even find an insight worth testing.
With New Vintage connected to Claude, the team ran a full buyer analysis in one pass, just by asking. They identified a promising customer segment that hadn’t received offers aligned with their behavior, and tested two campaigns that returned 39× and 58× more revenue per recipient.
01 · The problem
Everyone gets the same emails.
Consumers value personalized experiences. We can do this for the 100 people we know, but doing this for 10,000 people requires data analysis and systems.
For most wineries, data is fragmented across systems, and analyzing it takes time and expertise. Exports, cleanups, and spreadsheet gymnastics are prerequisites to identifying customer trends, let alone new email campaigns to try.
That means days of analysis that DTC teams don’t have, just to build a customer list they cannot easily reproduce.
Full list
The same offer for everyone reachable.
Product-specific
Certain wines for buyers of those SKUs.
Club
Special offers for club members.
Behavior
An offer aligned with how each customer buys.
02 · The solution
Ask the whole customer story at once.
The Premier Cru team installed New Vintage in Commerce7, connected it to Claude, and built a full buyer analysis in one session.
This is where New Vintage provides immediate value. In a few clicks, Premier Cru connected Commerce7, Klaviyo, and RedChirp data so that it lives in one place, eliminating the need for manual exports while enabling users to directly query the data via natural language.
Orders & club history
Customer conversations
query results
03 · The finding
High-value e-commerce customers weren’t converting on generic email offers.
Klaviyo reports revenue conversions based on last-touch attribution. This alone doesn’t tell you which website orders came from a given campaign.
Premier Cru profiled website orders via New Vintage, placing buyers on two axes: price per bottle and bottles per order.
Within a few prompts, Premier Cru took three passes over the data, each one sharpening the picture:
Pass 1 — Build the personas. Five buyer personas were built from Commerce7 order history, cutting on club status, customer lifecycle stage, and buying mode.
Pass 2 — Segment on price and volume. The same personas were analyzed on two axes: price per bottle and bottles per order. This is where the new segment surfaced: buyers with no club relationship, in the top price tier, ordering anywhere from single bottles to full cases.
Pass 3 — Overlay engagement. Klaviyo engagement data was layered on top to see how each persona was engaging with email campaigns. This high-value segment was receiving fewer campaigns and engaging less than the most engaged cohort.
Buying the highest-priced wines. Missing from the most-engaged cohort.
Fewer campaigns. Lower engagement. A high-value audience hiding in the data.
Most engaged cohort
67.5%
open rate
4.1%
click rate
19
campaigns
50.9%
open rate
2.2%
click rate
15
campaigns
It was during the second pass that uncovered a new segment that previously wasn’t analyzed. The buyer with no club relationship, within the top price tier, and purchased anywhere from single bottles to full cases. This discovery is what initiated the third pass and ultimately led to Premier Cru identifying an audience. An audience that purchased the highest-priced wines yet was silently being filtered out of the cohort that historically showed the highest engagement.
HYPOTHESIS
If this audience received an offer aligned with their behavior, then their conversion rates will increase.
04 · The proof
Two email campaigns to test the new audience.
With this new insight a couple of experiments were conducted to validate the findings.
Experiment one
The right wine to the right buyer
James Suckling scored Argyle’s wines from 92 to 97 points across every price tier. Referencing these new scores Premier Cru built a campaign plan matching scores to segments on one principle: reward the case, not the discount. Each email used a shipping incentive for a case rather than a price cut.
Two versions of the email were sent to mutually exclusive audiences. The segmented version led with the 95+ point single-vineyard Pinots and invited recipients to build a case. The full-list version led with a 94-point Pinot Noir as a weeknight bottle. The targeted segment had better performance all-around.
Sent the same morning to mutually exclusive audiences · revenue is paid, non-club, Commerce7 web-channel orders only
Experiment two
Same email, different audience
The first experiment moved two things at once: the audience and the wine. To isolate the audience, Premier Cru took the email the full list had already received and resent it, unchanged, to the targeted segment five days later — same copy, same creative, same subject line, same offer. The only thing that differed was who received it.
The same email sent to the right audience drove 58x more web revenue per recipient, in two fewer days.
Full list figures are the original send from campaign one; the segment received the identical email five days later, measured over a shorter window · revenue is paid, non-club, Commerce7 web-channel orders only
05 · Learnings
Three lessons from this experiment.
01
Behavior adds a dimension for segmentation.
The entire discovery in this case study came from segmenting buyers by price tier and bottles per order instead of just by club membership or SKU-level purchase history. Club status told Argyle who was already inside the tent. It didn’t tell them who was buying like a top customer while standing outside it. That’s the more general lesson: the highest-value behavioral segments in a winery’s data often cut across the categories that email programs are typically built around: club, full list, and product-specific.
02
A live connection lets you recreate lists easily.
Klaviyo can segment on what it observes: opens, clicks, prior sends, and sometimes website data. It can’t segment on price per bottle or bottles per order, because those measures are derived from Commerce7 order data. So the audience Premier Cru had just identified—buyers with no club relationship in the top price tier—existed as a description rather than as an audience anyone could mail. With New Vintage, they can refresh a list by asking Claude to do it again, or even schedule a task to rerun it periodically and add it to Klaviyo.
03
Reconcile conversions before you celebrate.
The email platform credited this campaign with orders that never happened on the website. Measuring against actual Commerce7 web orders provides a clearer signal on campaign ROI. Using New Vintage to do this helps the team separate signal from noise for campaigns meant to drive website orders.
06 · The takeaway
Personalizing email campaigns is finally within reach.
Personalization is easy in the tasting room. Staff can tailor an experience in real time and sell more because of it. In the inbox, customers get the same email because turning data into something usable can take days that busy teams don’t have.
For Argyle, this newfound customer segment isn’t going to produce 58× revenue per recipient at scale. The teams will test and refine further, but the segment isn’t the win so much as evidence that the next one is a few prompts away.
Experimentation is the path to innovation. Giving smart, creative people the ability to experiment faster seems like a good idea.
Find the buyers your best emails are missing.
Try asking AI: “Isolate web channel orders and organize customers by average price per bottle, bottles per order, and wine club status.”
Free 14-day trial ∙ No credit card required
Method
Campaign revenue counts paid, non-club, Commerce7 web-channel orders. Engagement comes from Klaviyo. The second experiment compares the original full-list send with the targeted resend five days later, using five-day and three-day measurement windows respectively.
About this case study
Argyle Winery and Premier Cru used New Vintage with Claude to analyze Commerce7, Klaviyo, and RedChirp data. The reported 39× and 58× comparisons use the displayed revenue-per-recipient figures. This result is evidence for further testing; it is not an expectation of 58× at scale.
Argyle Winery × Premier Cru Solutions × New Vintage
Argyle case study · Commerce7 web-channel orders