# American Trench started texting customers after delivery. Four weeks in, they were buying again 37% more often

> In a randomized A/B test, thousands of customers who received a Signals text after delivery repurchased 37% more often after four weeks than a holdout that got nothing. Among customers who replied, the lift was 74%. All this without the need for discounts or ad spend.

<div class="not-prose my-2">
  <p class="text-sm uppercase tracking-[0.14em] text-tertiary mb-3">American Trench × Signals</p>
</div>

[American Trench](https://americantrench.com) makes clothes meant to last twenty years: ventile trench coats, waxed chore coats, shaker sweaters, selvedge denim, all made in the USA. But selling high-end apparel online comes with a common challenge: how do you provide the best possible experience to every customer when every interaction is completely remote, and how do you make sure they come back to buy again and again?

Many brands reach for marketing SMS discounts and boatloads of retargeting spend, tactics that can make the brand feel cheap and less exclusive. Signals gives brands a different path centered on useful customer conversations. We engaged with American Trench customers over iMessage in two ways: a check-in right after delivery (outbound), and a "Text us" button on the website (inbound). Both drove significant new revenue: the outbound outreach increased the 28-day repeat purchase rate by 37%, and the inbound texts converted some of the largest orders (on average over twice the typical AOV) by giving high-spending customers confidence in what they buy.

<div class="not-prose my-8 grid grid-cols-1 sm:grid-cols-3 gap-3">
  <div class="rounded-2xl border border bg-accent-info/10 px-5 py-5">
    <div class="text-xs uppercase tracking-[0.12em] text-tertiary mb-2">Everyone texted</div>
    <div class="text-3xl font-heading text-primary">+37%</div>
    <p class="text-sm text-secondary mt-2">28-day repeat purchase lift across every customer who received the check-in, replier or not.</p>
  </div>
  <div class="rounded-2xl border border bg-accent-info/10 px-5 py-5">
    <div class="text-xs uppercase tracking-[0.12em] text-tertiary mb-2">Repliers</div>
    <div class="text-3xl font-heading text-primary">+74%</div>
    <p class="text-sm text-secondary mt-2">Customers who replied repurchased 74% more often than the holdout within 28 days.</p>
  </div>
  <div class="rounded-2xl border border bg-accent-info/10 px-5 py-5">
    <div class="text-xs uppercase tracking-[0.12em] text-tertiary mb-2">Reply rate</div>
    <div class="text-3xl font-heading text-primary">37%</div>
    <p class="text-sm text-secondary mt-2">Customers texted who wrote back, on a check-in with no discount and no survey.</p>
  </div>
</div>

<div class="not-prose my-10 border-l-4 border-accent-primary pl-6 sm:pl-8 py-2">
  <p class="text-xl sm:text-2xl text-primary leading-snug mb-4">&ldquo;American Trench has always been a brand built around the details. Signals is another way for us to carry that same attention to detail into how we communicate with our customers.&rdquo;</p>
  <div class="flex items-center gap-3">
    <div>
      <p class="text-base font-medium text-primary">Jacob Hurwitz</p>
      <p class="text-sm text-secondary">CEO and Co-founder, American Trench</p>
    </div>
  </div>
</div>

## A small change to the customer journey

For outbound, we randomly split American Trench customers into a treatment group that received post-delivery check-ins and a control that only received the pre-existing post-purchase email flow. The check-in opened with a low-pressure service question such as "Hi Alex, this is Olivia from American Trench, how is your new jacket fitting?" Signals treated these brand-initiated follow-ups as marketing and evaluated channel, consent, suppression, and restricted-state signals before each send.

Eligible customers either had written marketing consent or confirmed iMessage service with no restricted-state signal. We ended the outreach after the first message when a customer didn't reply. 37% of American Trench customers replied, some with support questions and others asking about products they could buy next.

We compared both the number of orders and the total spend by customers who we texted (treatment group) and who we did not text (control group). To see if we're simply moving forward sales that would have happened anyway or creating true incremental sales, we also stratified uplift by weekly maturities (e.g. 1-week maturity means repeat purchases made within 1 week of enrollment, for customers who were enrolled more than 1 week ago).

After 4 weeks, the treatment group had 37% more repeat purchases than the holdout, the strongest lift across the 4 maturity windows. The lift reached 35% after 2 weeks, measured 29% after 3 weeks, and rose to 37% at week 4. That persistence across the month supports the conclusion that the check-ins created incremental purchases.

**Repeat purchase lift vs. holdout**

| Cohort            | 1 week | 2 weeks | 3 weeks | 4 weeks |
| :---------------- | -----: | ------: | ------: | ------: |
| Everyone texted   |    +9% |    +35% |    +29% |    +37% |
| Replied customers |   +45% |    +70% |    +64% |    +74% |

<div class="not-prose my-8 rounded-2xl bg-surface-elevated border border px-6 py-6 sm:px-8 sm:py-7">
  <p class="text-xs uppercase tracking-[0.12em] text-tertiary mb-4">Example conversation with an American Trench customer</p>
  <p class="text-sm text-secondary mb-2"><span class="text-tertiary">Signals:</span> hi aaron, this is olivia from american trench. saw your shorts arrived. how's it all so far?</p>
  <p class="text-sm text-primary mb-2"><span class="text-tertiary">Customer:</span> I literally haven't taken off the shorts since they arrived</p>
  <p class="text-sm text-primary mb-2"><span class="text-tertiary">Customer:</span> I wish the brown and mauve striped mesh were still available. Got any secret stashes?</p>
  <p class="text-sm text-secondary mb-2"><span class="text-tertiary">Signals:</span> no secret stashes, unfortunately! those are completely gone.</p>
  <p class="text-sm text-primary mb-2"><span class="text-tertiary">Customer:</span> No, I like the stripes. I'll try your pants out this fall</p>
  <p class="text-sm text-secondary mb-2"><span class="text-tertiary">Signals:</span> when fall rolls around, you might really like the keystone pant in hickory stripe. they're a straight-fit cotton twill made in los angeles that can take a serious beating. want to see a photo?</p>
  <p class="text-sm text-primary mb-2"><span class="text-tertiary">Customer:</span> Sure</p>
  <p class="text-sm text-primary"><span class="text-tertiary">Customer:</span> Love em</p>
</div>

<div class="not-prose my-8 rounded-2xl bg-accent-info px-6 py-6 sm:px-8 sm:py-8">
  <p class="text-xl sm:text-2xl font-heading text-inverse leading-snug mb-3">&ldquo;They&rsquo;re great! I love American Trench. I&rsquo;m literally wearing the jeans right now.&rdquo;</p>
  <p class="text-xs text-inverse/80">&mdash; Jeremy, American Trench customer</p>
</div>

## The "Text us" button became the biggest-ticket channel

The "Text us" button provided another way for customers to talk to us. Oftentimes customers texted us inbound because they wanted to make a big purchase but had hesitation. Our agent was able to navigate the customer's questions and help them find the SKU that is right for them.

The result: 30% of all inbound conversations ended with the customer buying something, with the average order value being twice the store's overall AOV. When their orders arrived, we followed up, closing the loop and building a relationship.

<div class="not-prose my-8 rounded-2xl bg-accent-info px-6 py-6 sm:px-8 sm:py-8">
  <p class="text-xl sm:text-2xl font-heading text-inverse leading-snug mb-3">&ldquo;I&rsquo;ve had other AI responses, and they are terrible, so well done.&rdquo;</p>
  <p class="text-xs text-inverse/80">&mdash; David, American Trench customer</p>
</div>

## The same thread does sales and support work

What is special about Signals is that we are genuinely helpful. In the same thread we would help set up an exchange for a customer with a jacket that ran large, figure out a birthday gift for their loved one next month, and alert them when an item that was out of stock got back into stock. When a policy exception was necessary, we would alert American Trench customer support personnel and they would reply all in the same iMessage thread as our AI agent, providing customers a truly seamless experience.

<div class="not-prose my-8 rounded-2xl bg-accent-info px-6 py-6 sm:px-8 sm:py-8">
  <p class="text-xl sm:text-2xl font-heading text-inverse leading-snug mb-3">&ldquo;Overall it&rsquo;s really (really) tough and well made. The hardware is sick and I got a great deal. Can&rsquo;t wait to wear it this fall!&rdquo;</p>
  <p class="text-xs text-inverse/80">&mdash; John, American Trench customer, on the Waxed Chore Coat</p>
</div>

## Why this mattered for American Trench

American Trench wanted to keep its brand character and build long-lasting relationships with its customers. Genuine two-way conversations at scale gave the brand a way to do both. Signals helped increase repeat purchases by 37% in an A/B test, and it converted 30% of inbound conversations into sales at twice the typical AOV.

Many conversations that started over a month ago are still going on. Customers come back and talk to Signals because they know they now have a friend and advocate at American Trench, and that is priceless.

<div class="not-prose my-8 rounded-2xl bg-accent-info px-6 py-6 sm:px-8 sm:py-8">
  <p class="text-xl sm:text-2xl font-heading text-inverse leading-snug mb-3">&ldquo;Thanks for asking. Always good clothes.&rdquo;</p>
  <p class="text-xs text-inverse/80">&mdash; Luke, American Trench customer</p>
</div>

<div class="not-prose my-12 rounded-2xl bg-surface-elevated p-8 sm:p-10 text-center">
  <p class="text-xl sm:text-2xl font-heading text-primary mb-3">Want results like these for your brand?</p>
  <p class="text-secondary mb-6 max-w-md mx-auto">Signals helps apparel and lifestyle brands turn post-delivery check-ins into real customer conversations and measurable repeat purchase lift.</p>
  [Book a Demo](https://cal.com/alejandro-zaniolo/30min?overlayCalendar=true)
</div>