How Smart Brands Are Using AI to Win

Real, actionable strategies for DTC brands to get ahead with AI

Pietro Saccomani
February 20, 2025

Just about every tool is now “AI-powered” or “AI-assisted.”

But let’s cut through the hype: Is AI actually making your business more profitable? Or is it just another flashy trend?

The short answer: a little of both (but leaning more towards the first camp, if you know how to use it right).

The long answer? That’s the subject of this week’s email (it’s not actually that long).

This deep dive start from our experience at the New York Fashion Ecommerce Summit, where our Head of Growth, Nihal Mandanna, chaired the Fashion Forward Technologies panel, featuring a collection of industry leaders from global retail brands.

The insights they shared were incredible (Nihal wrote about the panel here).

On the back of that, we started thinking about more real-world implementations of AI for DTC and retail operators, big and small.

We wanted to uncover whether AI can really make a difference, and if so, what’s the right way for brands to use it?

We came up with these five takeaways that sum up how AI and DTC work together to produce results.

1. AI Lets Lean Brands Compete (And Win)

We all know the current state of DTC ecommerce.

Low margins, high CAC, and it seems like every change that iOS, Meta, Google make is designed to make it harder for brands to make a profit.

Big, VC-funded brands are burning cash just to break even. AI is letting lean, bootstrapped brands move faster, smarter, and more profitably.

✅ AI-generated ad creative → More iterations, lower cost per test.

✅ AI-powered customer service → Instant answers, fewer support hires.

✅ AI-driven inventory forecasting → Cut dead stock , avoid overordering.

💡 Takeaway: AI is a force multiplier. If you’re not using AI to automate low-value tasks, you’re leaving money on the table.

2. AI Should Support (Not Replace) the Customer Experience

AI doesn’t change the end goal of everything you do – which is to provide an awesome customer experience.

Make sure it plays a supporting role in your CX strategy, not taking it over completely.

You should be using AI to either:

Provide the same level of customer experience, with fewer resources.
Provide a better customer experience, with the same resources.

At the end of the day, your AI-driven workflow is working towards the wrong goals if it makes the customer’s experience worse (even if margins initially go up).

Here’s an example: completely automating customer service with AI.

AI can be incredible when it comes to automating customer service, reducing the load on your CS team, and providing customers with immediate answers.

Replacing your whole customer service team with AI might reduce expenses and increase net profit in the short term…

But you’ll have issues with more complex questions or problems, which will degrade your brand’s reputation over time, ultimately hurting your business, and creating a problem that’s hard to reverse.

The best brands use AI to:

✅ Automate the repetitive, humanize the rest. Chatbots handle FAQs, while human reps build real relationships.

✅ Enhance, not replace. AI recommends products, but real people handle VIP customers.

✅ Make personalization smarter. AI curates offers based on browsing history, not just discount spam.

💡 Takeaway: AI should lower costs without killing CX. If your AI implementation makes your brand feel robotic, you’re doing it wrong.

3. The Human Element Is Now a Competitive Advantage

Customers are already tuning out AI-generated content.

Your customers can spot AI-generated ads. After all, they’re probably using AI in their own jobs as well.

AI is everywhere - and don't think your customers don't notice it.
There are amazing operational use cases for AI, and it can help lean startups compete with VC-funded brands.
But on the customer side, there's going to be greater demand for human-first content.
Do more of:
✅… x.com/i/web/status/1…
— Pietro Saccomani (@psaccomani)
1:34 PM • Feb 13, 2025

Automating content that’s clearly spun up by AI and putting it out into the world will tank engagement so fast.

As a direct response to how much AI content is online, we’re seeing customers respond far better to brands with a real human face and voice.

The best-performing ads, emails, and content today are:

“Ugly” UGC that feels real.
Emails that sound like a human, not ChatGPT.
Messaging that leans into real human connection.

There was an excellent feature in the Chew on This newsletter recently about Barry Hott’s system for creating “ugly” ads.

Barry built a whole system around unpolished, low production value, genuine and authentic ad creative.

These ads consistently deliver better engagement, because they are human.

I loved this quote from Barry:

❝

People hate ads. But they don't hate all ads. They hate ads that feel like ads.

There are so many examples right now of brands killing it with raw, unpolished branding, which resonates with customers because it feels real.

💡 Takeaway: The best brands today are blending AI’s speed with real human storytelling. Never discount the importance of having a human face on your brand.

4. AI Can Revolutionize Sizing & Slash Return Costs

Returns have always been a margin-killer for fashion and apparel brands.

As much as you can try to reduce the impact of returns, it’s long been accepted as a cost of doing business.

One of the hot topics at the panel in New York was how AI could transform the sizing experience in ecommerce, and put cash back in brands’ pockets.

They talked about the common customer practice of “bracketing” (buying multiple sizes and returning the ones that didn’t fit).

The main problem is that few customers actually look at size charts, and even when they do, having to measure yourself and guess which size will fit best is a clunky way to do it.

Marie Prune de Batz, Performance Marketing & Merchandising Director at Adore Me, summed up the struggles just about every fashion brand has with first-time customers.

❝

The most challenging part for us is fit. If you're not sure about your size or have no idea how a garment will look on your body, it's hard to shop for certain product categories like intimates online.

Our partners at Bold Metrics are doing some cool stuff to solve for this.

Their AI-modeling makes better fit estimates without asking a customer to go through the tedious process of getting exact measurements.

Brands like Canada Goose, New Balance, have seen a massive lift in customer satisfaction while drastically reducing return rates with their tech.

Finding a way to bridge your digital CX with a delightful in-person brand interaction simply sets the stage for long-term retention.

💡 Takeaway: Sizing and fit is one area where brands can both enhance the customer experience (less anxiety and doubt for customers over selecting the right size) and increase profits (more accurate sizing = less returns).

5. The 10/80/10 Rule for AI Execution

As great as AI can be, don’t fall into the trap of letting it run entire processes.

This runs the risk of:

Putting things out that harm your brand’s image (e.g. incorrect information).
Key information slipping through the cracks (product performance, supply chain issues that don’t get seen).
Creating a disconnect between you and your customers.

As Marie from Adore Me mentioned in the panel:

❝

Today, our merchandising is made automatically but with supervision from merchandisers that know the products. Tomorrow, if it's 100% driven by AI, we won't have any control, and if there's an issue, we won't know.

The capabilities of AI are amazing, but it lacks human intuition and creativity, which is essential for building a successful brand.

Mike Coble, IT Director Digital and CRM Technologies at Eileen Fisher‍ had this to say:

❝

Think of AI like hiring an intern. You never have them churn out the final product. You say this person's going to help me get stuff done quicker.

Sabrina Wong, Head of UX at Tapestry‍ echoed a similar perspective:

❝

AI is assistive technology meant to help you with your work but never replace your creativity. It always has to amplify human creativity and human connection.

We recommend a 10/80/10 approach:

10% human input – Strategy, direction, big-picture thinking.
80% AI execution – Automate repetitive, data-driven tasks and legwork.
10% human refinement – Final review, brand consistency, polishing.

💡 Takeaway: AI is a tool, not a replacement. Use it for scale, but keep human oversight where it matters.

Final Thought: AI Won’t Replace You – But AI-Powered Brands Will

DTC may be getting harder, with higher CAC, slimmer margins, and more competition for your customers’ attention.

And it’s only going to be harder if you don’t take advantage of AI.

But take care that you’re using it the right way.

The brands that win are not using AI for AI’s sake, but injecting AI into their workflow to get more efficient, more profitable, and more human.

How are you adopting AI in your business? 📧 Reply and let us know any unique tips, tricks or use cases you’ve found.

Quick Hits

Here’s some of the best content we’ve read from in and around the ecom world.

$150M CEO On Why Brand is Crucial

Dude Wipes is one of the best examples out there of building a business on the back of a brand with a unique voice and image.

CEO Sean Riley is also one of the best follows for DTC & ecomm insights.

In this 3-minute video (posted by Preston Rutherford, another well of incredible ecommerce knowledge), Sean gives a masterclass on why your business should be investing in brand (and a little on how to do it).

You can check out the full interview here (37 mins), or check out the LinkedIn post for a super quick, abridged version that you can watch while waiting for your coffee.

Chat GPT Deep Research for DTC Brands

On the back of what we’ve been talking about in this newsletter, here’s an awesome look at the capabilities of OpenAI’s latest tool, Deep Research.

This model of Chat GPT goes far further than any of the previous models (such as o1, o3, 4o), and does (as the name suggests) deep research of the web, collating its findings in a structured report.

It even asks follow-up questions to ensure it gets the assignment right.

It’s like a proper research assistant in your pocket.

This video shows some specific Deep Research use cases for DTC brands – such as competitive analysis, looking for market gaps, and uncovering insights for product development.

Why the First Message is the Most Important Message

First impressions matter. And with email, SMS and push, you’ve got one chance to make this first impression with a new subscriber matter.

The welcome message you send to someone just after they subscribe to your list may just be the most important you’ll ever send. It’s a crucial opportunity to turn this subscriber into a customer, and a long-term fan.

This article dives into why welcome messages are so important, and helps you craft the ideal ice-breaker (with real examples from successful DTC brands).

Do Upsells Always Lead to More Sales?

Upsells in the cart drawer are standard operating procedure for most brands.

But in this post, Samuel Hess from Drip Agency shares how they A/B tested removing this upsell from the cart drawer… and increased revenue per user 2.68%.

Their hypothesis that fewer distractions would = more completed purchases turned out to be true.

Does that mean you should remove your own cart drawer upsells?

Not necessarily. In fact, probably not. These work for most brands. The results of this test could be an outlier, due to the market, or the type of product.

But what you should take away is the value of testing everything. Rather than blindly following “best practices”, let the data guide you.

The Shoppers Have Spoken

What to know what really matters to your customers?

Yotpo (maybe you’ve heard of them) analyzed 51 million reviews to put together a bumper guide on what’s behind positive reviews, negative reviews and returns right now.

It gives a data-driven look at 2024’s most loved products, across 8 industries, from Fashion to Beauty to Outdoor.

That’s all for now.

We’d love to hear what’s working and what’s not for your brand in terms of AI.

Just hit reply – we read every message.

Also, keep an eye on RetentionEdge.co for our brand new podcast, coming soon, where we’ll be sitting down weekly with the brightest CX and retention minds in ecommerce and retail.

Until next time!

Pietro and The Retention Edge Team

Someone forward this to you? Get The Retention Edge delivered weekly →