206: The Right Way To Run Facebook Retargeting Ads With Reza Khadjavi Of Shoelace

Podcast: Download (Duration: 45:02 — 62.1MB)

Sequential Facebook retargeting means showing a visitor a different ad each day after they leave your site, rather than repeating the same dynamic product ad until they tune it out. Roughly 98% of visitors leave without buying, and the standard approach of one ad on repeat produces banner blindness and eventually irritation with your brand.

Reza Khadjavi founded Shoelace, which builds automated retargeting journeys for ecommerce stores and runs millions of dollars a month in retargeting spend. The company started when merchants he was interviewing about a different idea kept saying their retargeting was set up badly and they knew it.

This episode walks through the three levels of retargeting sophistication, a real four-stage journey from a children’s shoe brand, the frequency caps worth watching, and the average order value threshold below which retargeting stops working.

Key takeaways

• Three levels of retargeting: blanket ads to all visitors, behavioral segmentation by funnel stage, then sequential journeys that vary the ad over days.
• Segment by what visitors did. Someone who only saw your homepage should not get the same aggressive discount ad as someone who abandoned a cart.
• An audience segment needs roughly 500 to 1,000 people before a dedicated journey makes sense.
• Mix ad types across the journey: dynamic product ads, lifestyle carousels, objection-handling content, brand video, and bestseller carousels.
• Judge the journey holistically rather than demanding immediate return from each individual ad.
• Keep frequency in the mid-to-high single digits, roughly 5 to 9, and check which placement is driving it.
• Average order value under $30 makes strong retargeting returns very difficult. Above $60 to $100 it gets much easier.
• Match journey length to purchase consideration time. Impulse products need short journeys, considered purchases need longer ones.

What is retargeting and why does most of your traffic need it?

Retargeting shows ads on Facebook, Instagram, and around the web to people who visited your site without buying. Roughly 98% of visitors leave without converting, and only a fraction of those leave an email address you can follow up with.

For everyone else, retargeting is one of the only remaining ways to reach them. It is not about acquiring new traffic, it is about extracting more from traffic you already paid for.

What are the three levels of Facebook retargeting?

Level one shows every visitor the same ad regardless of what they did. Level two segments by behavior. Level three sequences different ads across days.

The blanket approach is what most stores run, and it is better than nothing when your traffic is small. Everyone who touched the site sees one ad for some number of days after leaving.

Behavioral retargeting maps the customer journey and splits people by where they dropped out. Homepage, collection page, product page, add to cart, purchase, and repeat purchase are all distinct signals.

Sequential retargeting adds time. Instead of one ad repeating, the visitor moves through a varied sequence of ad experiences over several days.

Why segment retargeting audiences by funnel stage?

Message intensity should match how far someone got. Aggressive “come back and buy now” copy makes no sense to someone who saw your homepage and never viewed a product.

That same copy works well on someone who added an item to their cart. They have demonstrated real intent, so a discount offer can close the gap.

The success metric changes by stage too. For a homepage-abandon audience, moving people to view a product is the win, rather than an immediate purchase.

Retargeting also does not stop at non-buyers. Past customers are a distinct audience worth targeting with repeat-purchase and loyalty messaging.

How large does an audience need to be for its own journey?

Roughly 500 to 1,000 people is the practical minimum before a segment justifies its own sequential journey. Below that the audience is too thin to serve a multi-day sequence.

Shoelace builds the audiences anyway and monitors them. One day, two day, three day, five day, seven day, and 30 day windows across each behavior type get created from the start.

That lets the system watch each segment grow and notify the merchant when a new journey becomes viable. Retargeting complexity should scale with traffic rather than being fixed at setup.

What does a real sequential retargeting journey look like?

Wee Squeak, a Shopify brand selling children’s shoes, runs a four-stage journey lasting about seven days for people who viewed at least one product.

Day one serves two ads simultaneously. A dynamic product ad based on what the visitor looked at, plus a carousel of lifestyle images pulled from their Instagram feed and tagged with products.

The second stage handles objections and builds the brand. One ad links directly to their sizing chart, addressing the obvious hesitation about buying shoes for a child online, and the return policy sits on that same page.

The other second-stage ad is a short video of a toddler wobbling along in their shoes. It is roughly ten seconds, genuinely charming, and its job is establishing what the brand feels like.

The third stage is a bestseller carousel, hand-assembled from that month’s top-performing products rather than dynamically generated. The copy names it as such: here is what people are buying right now.

The final stage returns to a dynamic product ad tied to the visitor’s specific browsing behavior. The later stages each run a couple of days longer than the early ones.

Do lifestyle images outperform dynamic product ads in retargeting?

It depends on your vertical, your average order value, and where your audience spends time. Reza deliberately declines to give a universal answer despite running retargeting at large scale.

The pattern he can identify is audience-driven. If your customers live on Instagram, lifestyle imagery resonates and produces better click-through rates than white-background product shots.

Dynamic product ads work best for stores where the product itself is the entire proposition. Trendy gadgets and general merchandise stores fall into this category.

Brand-driven stores are the opposite case. When people buy from you because of what you stand for, product-only ads never communicate that, which is exactly the gap sequential journeys fill.

Running both is the sensible default. That is why day one of the Wee Squeak journey serves both formats at once.

Where should a lifestyle retargeting ad send traffic?

Send lifestyle ad clicks to a shoppable gallery rather than a generic product page, so the image the person clicked stays visible alongside purchasable products. Shoelace integrates with Foursixty, which builds these shoppable Instagram galleries.

Without that setup, link directly to the product page. The important thing is continuity between what they clicked and what they land on.

How do you measure a sequential retargeting journey?

Measure conversion across the whole journey rather than demanding immediate return from each ad. A brand video or a sizing chart ad is doing a different job than a bestseller carousel.

Individual ads still get evaluated on their own terms. Click-through rate, engagement, and video watch rate tell you whether a specific creative is working within its role.

Reza’s benchmarks are practical rather than precise. Facebook relevance scores of seven to nine are a strong signal, click-through around 1% or higher is acceptable, and 2% to 4% is excellent.

Near-zero click-through is a clear failure. Retargeting audiences should also outperform your cold traffic on click-through, since these people already know you.

For a ten-second brand video, expect a very high watch rate, around 90%, with most viewers seeing most of it.

Why measuring ad-set-level return on ad spend is misleading

Optimizing purely for return multiple pushes you toward the smallest, hottest audience, which caps your total sales. Narrowing everything to add-to-cart audiences produces spectacular multiples on a tiny revenue base.

That is the tradeoff nobody states explicitly. Higher multiple, less absolute revenue.

Different merchants sit differently on that curve. Some want the minimum spend that clears 3x, and others will accept a lower multiple for meaningfully more sales.

Agreeing on the success definition before comparing approaches is the necessary first step. Otherwise two people comparing retargeting methods are measuring different things.

What frequency cap should you use on retargeting ads?

Mid to high single digits, roughly 5 to 9, is a reasonable frequency target for retargeting. Anything above 10 starts warranting attention.

Check which placement is generating that frequency before you panic. A high number driven by right-hand-column desktop placements is far less concerning than the same number driven by mobile newsfeed.

The reason is disruptiveness. Right-hand-side placements barely register with users, while feed placements are genuinely interruptive.

Running four to six varied ads is itself a frequency management technique. Each individual creative gets seen fewer times while total exposure stays useful.

What average order value do you need for retargeting to work?

Below $30 average order value, strong retargeting returns get very difficult. Above $60 to $100 it becomes substantially easier.

The math is straightforward. Ten sales at $20 and ten sales at $150 represent identical ad performance and wildly different returns.

Traffic quality is the other variable and it matters more the lower your order value goes. Excellent traffic at a low order value can still work if enough of it converts.

The combination to avoid is low order value plus poor traffic quality. Retargeting only advertises to people who already visited, so bad traffic in means bad retargeting out.

That makes traffic auditing a prerequisite. Fixing the source of your visitors comes before optimizing what you show them afterward.

Where should you start with retargeting?

Start with an add-to-cart audience, since it is small, cheap to reach, and highly engaged. If you cannot generate a strong return there, something is wrong upstream.

That audience becomes your baseline. Once you know what add-to-cart retargeting returns for your store, you can compare higher-funnel audiences against it as you expand.

Poor performance on an add-to-cart audience points at the site or the messaging rather than at the ads. These are people who chose your product and stopped at checkout.

How long should a retargeting journey run?

Match the journey length to how long the purchase decision actually takes. Impulse products need short windows and considered purchases need longer ones.

A t-shirt or a novelty gadget gets bought quickly or forgotten. Nobody deliberates for a week over it.

Jewelry or anything in the mid-hundreds is different. Buyers compare competitors, think it over, and come back days later, so a journey that ends after 48 hours misses them entirely.

Frequently asked questions

What is sequential retargeting?

Showing a visitor a varied sequence of different ads across several days after they leave your site, rather than repeating one ad. It reduces ad fatigue and lets you tell a brand story across multiple touchpoints.

How is behavioral retargeting different from standard retargeting?

Behavioral retargeting segments visitors by what they did on your site, so homepage abandoners and cart abandoners receive different messages. Standard retargeting shows everyone the same ad.

How big does a retargeting audience need to be?

Roughly 500 to 1,000 people before a dedicated journey is worth building. Smaller segments cannot sustain a multi-day sequence.

What is a good click-through rate on retargeting ads?

Around 1% or better, with 2% to 4% being excellent. Retargeting should outperform your cold traffic click-through since the audience already knows you.

What frequency is too high for retargeting ads?

Above 10 warrants investigation, and 5 to 9 is a reasonable range. Check which placement is driving the number, since right-hand-column impressions are far less intrusive than feed impressions.

What average order value do you need for profitable retargeting?

Under $30 makes strong returns very difficult. Above $60 to $100 the economics work much more easily, assuming reasonable traffic quality.

Should each retargeting ad drive immediate sales?

No. Content ads, brand videos, and objection-handling ads serve different roles within a journey, so measure conversion across the whole sequence rather than per ad.

Where should beginners start with retargeting?

With an add-to-cart audience. It is small, inexpensive to reach, and highly engaged, which makes it a reliable baseline to compare other audiences against.