Click Here & Buy ZATO Owner, Kirk Williams' newest book on Google Ads - Ponderings of a PPCer: Revised & Expanded.
Kirk Williams
Google Shopping

Stop Segmenting Shopping Ads by Margin or ROAS

Stop Segmenting Shopping Ads by Margin or ROAS

10/25/19 UPDATE: Hello Facebook Agency Visitor Person!  We’re delighted to have you visit this awesome post. About a year ago, ZATO stopped offering Facebook Ads solutions so we could focus solely on what we do best: Google Ads. Because of this, we’re always interested in partnerships with great Social Advertising agencies (like yourself, wink wink!) and we offer referral fees for signed clients!  Anyway, back to it, and happy reading…

Post Summary

  • To Segment, or Not to Segment... That is The Question
  • So let's talk segmentation by Margin or ROAS in Google Shopping Ads. Here are 4 Issues With It.
  • So, how should we segment Shopping Ads then? 

For years, I felt like a neanderthal in the online Shopping Ad discussion space because every time the "Top 10 hottest ways to segment your campaigns" for Shopping Ads came out, I would find ROAS or Margin segmentation near the top of the list and would feel like I was missing something.

I just don't like segmenting by ROAS / ROI / Product Margin, but hadn't ever really lined out my reasoning before.

That, friends, is about to change in this post!

I'm ready to go on the record against ROAS segmentation with this thought piece (and a little data for support, but it will be private so you'll have to collect your own and gather your own opinions mkay sorry not sorry).

To Segment, or Not to Segment... That is The Question

Let me start with a brief philosophical meandering into the realm of segmentation. Why do we segment our campaigns within Google Ads (really, in any channel)? 

Have you ever thought of that at length? Like, most people know some form of segmentation (breaking campaigns or ad groups into unique entities to target unique people) is probably a good idea. But what is the primary reason we do this? 

I would submit that the only reason to segment out your campaigns, is because it should help to better guide the targeting algorithm (through audiences, bidding, location, etc) in showing the right ad to the right person.

Anything you are doing to segment your campaigns, that does not do this, is pointless busy work. You're further complicating something what shouldn't be complicated (remember K.I.S.S.!).

But, importantly, if there is a way to better aim the Google targeting algorithms at the right people for your products, then by all means, segment away!

So let's talk segmentation by Margin or ROAS in Google Shopping Ads.

This is the practice by which advertisers group products according to their internal product margin, and then target their Google Shopping Campaigns accordingly.

One problem with this is, when you voice this strategy to others, it just sounds like a really great idea. Like, it's a strategy that already has fans simply because of how it sounds. It's the equivalent of Uncle Lester's "hold my beer" firework ideas at the 4th of July where the idea is a good one until it fully comes to fruition.

I mean, what operator's ears don't perk up at the idea of matching your ROAS targets to your product margin!


Conserve cost!

Focus on your most profitable products!

What could possibly be wrong about this? 

Let's go back to what I described before about segmentation, campaign segmentation isn't primarily about internal segmentation, it's about external consumer targeting.

In this case, you're not segmenting so Google can better match the right products to the right searcher, you're giving Google random products.

Think of it like this, let's say you create a Custom Label for ROAS groupings based on historical product performance. You lump your 15,000 products into three separate margin-focused Custom Labels: 

  • High_Margin
  • Med_Margin
  • Low_Margin

Then you build SSC (well, now PMax) campaigns around these three and sit back with your hands behind your head and a Mona Lisa-like smile on your lips, just waiting for the customer adulations to roll in.

But here are some issues with that: 

  • Marketing is by nature big picture focused, so you will always have some products that lose profit and some products that carry great deals of profit, and the savvy advertiser *balances those two together* across the account to actually grow the account while maintaining top-level ROAS or MER goals. In other words, you can actually kill your account by being overly focused on directly tracked ROAS. What if Products you have in your Low Margin custom label compose 50% of your sales and killing them would actually kill top-line revenue... even if your overall profit actually increases? Losing gross revenue can negatively impact a lot of other things like carrier shipping rate deals, etc. Profitability AND volume both typically matter (the exception being aggressive goals due to a specific season for any giving business). This strategy focuses solely on maximizing profitability (and rather poorly at that, keep reading).
  • The second, related issue with this strategy, is that specific products that lead to ad clicks aren't always the products people eventually purchase... think of some low margin products as doors to new worlds rather than the final destination. If the low margin $15 product brings people in successfully, and then they purchase $75 AOV of more, better products, then perhaps the solution isn't for you to filter those successful products out of the feed. Rather, it should be to focus on CRO and implement suggestions on your Landing Pages to better target people to the right products, upselling them in a better website experience. Good ole "loss-leader" products have been around for ages for a reason.
  • Another issue that often occurs when this strategy is employed, is you end up lumping multiple products together that would never be purchased by the same audience. This is crucial to understand and goes back to my previous thought about why we segment in campaigns in the first place. If you are selling a $20,000 watch with the same "margin" as a $20 watch on your website, then you've likely lumped two products together that should never be served the same ad creative or be targeted to the same people. With some exceptions, those are likely unique people and should be segmented accordingly. Now, admittedly, you can segment your campaigns/ad groups/asset groups further with other segmentations, but the point remains that margin alone doesn't really do anything helpful here.
  • The final issue is that... I just don't see this strategy actually working in accounts it's been implemented in! We've taken over a number of accounts running this setup, and we've tested variations of this setup as well and the funny thing is, I don't think I ever remember a time when I could see the Margins custom labels performing in a way that clearly matched the margins themselves and/or caused clear evidence that the margins themselves were leading to better purchasing in the desired high margin products. It's all over the map, like IMO, it just doesn't really actually DO ANYTHING to target margins individually. Perhaps someone has seen something differently, or segments in additional ways to aid targeting along with these custom labels, and that's fair. I'm just noting that from personal experience, I haven't seen evidence that segmentation by margin actually helps to incrementally improve performance. It just kind of looks cool and the client feels great because they feel like their Shopping Ads are now segmented by their internal goals when that's not the point of segmentation (though certainly, internal goals should be taken into account).

So, how should we segment Shopping Ads then? 

There are lots of ways to segment, and I would begin by pointing you to my first section above. The purpose of this post isn't to outline all of the ways you can segment (it's complicated and needs to be individualized to your account). Do the hard work to analyze, think, and observe how your customers are actually converting. Then consider how to better guide Google's algorithms (especially in PMax with different asset groups, though I'm frustrated by the lack of complete targeting assistance in PMax Campaigns as I've written here) to how people are actually searching and purchasing your products.

Product Type

For instance, Product Type is STILL an underutilized feed attribute, but it mirrors the site category system *which almost always mirrors how people are searching and buying on that site!!* Product Type is ready-made audience segmentation by how they will search and purchase! It's at least a good place to start.


The custom label I actually like to test in similar "grouped" ways, is Price-point (not margin). The reason I say price-point, is I think this better guides how people purchase. They don't care about your margin, but they are drawn to their own budget and your product price. Going back to the watches, whatever the margin, a person in the market for a $20 watch is not the same person (at least at that moment) as a person actively shopping for a $20,000 watch. They should receive different bidding targets, possibly different geolocation targeting and bidding, and especially different ad creative since the audience types typically interested in those two objects have different reasons and motivations for purchasing (Save money! vs Look the hippest at the club! 👈 don't use either of those, they're dumb examples).

Top Products

If you focus on your Top Selling products instead of margins, you are getting the products people most like to purchase (already) in front of even more people. This has been a long-standing beloved Custom Label for me (one I've shared often in my SSC Course on Udemy back when SSC wasn't murdered by Google) and it's because you're able to use Google's algorithms to get the products that lead to the most sales in front of more people. Other variations of this Custom Label are: Products-Most-Likely-To-Sell-Other-Stuff (these products aren't themselves top sellers, but still lead to a lot of conversions on the site) and Top-Reviewed products (grab your highest rated product IDs from your reviews feed and test that in its own campaign or ad group).

BTW, in terms of the Products-Most-Likely-To-Sell-Other-Stuff (we really need a better name for that... is "loss leader" the name? I mean, only if it loses money itself), Liam Wade had a great thought on LinkedIn I'm sharing:

"They’re called PLAs for a reason, they’re ADS *not* PRODUCTS. They’re the gateway to your full offer, not just the SKU (your website, delivery, brand). Here’s an example you’ll love…One of our brands has an outlier SKU with an AOV 5x the average, and a low margin % >>> It has one of the highest click volumes >>> Google Ads UI says it’s one of the top 5 revenue drivers.Can you guess the punchline? They barely sell any of it. It’s an exciting product that acts as a gateway to the more sensible ones that are then purchased on it!And don’t get me started on acquisition or CLTV…! PPC is fun, isn’t it? 🙃"

So, there is my diatribe against the much-loved margin-based custom label segmentation in Shopping Ads. I hope that this post gets you to think. What about you? If you love it, how have you gotten around the issues I bring up here? 

Come join the conversation on Twitter or LinkedIn!

Want more free content like this delivered directly to your inbox?
Subscribe Here
Kirk Williams
@PPCKirk - Owner & Chief Pondering Officer

Kirk is the owner of ZATO, his Paid Search & Social PPC micro-agency of experts, and has been working in Digital Marketing since 2009. His personal motto (perhaps unhealthily so), is "let's overthink this some more."  He even wrote a book recently on philosophical PPC musings that you can check out here: Ponderings of a PPC Professional.

He has been named one of the Top 25 Most Influential PPCers in the world by PPC Hero 6 years in a row (2016-2021), has written articles for many industry publications (including Shopify, Moz, PPC Hero, Search Engine Land, and Microsoft), and is a frequent guest on digital marketing podcasts and webinars.

Kirk currently resides in Billings, MT with his wife, six children, books, Trek Bikes, Taylor guitar, and little sleep.

Kirk is an avid "discusser of marketing things" on Twitter, as well as an avid conference speaker, having traveled around the world to talk about Paid Search (especially Shopping Ads).  Kirk has booked speaking engagements in London, Dublin, Sydney, Milan, NYC, Dallas, OKC, Milwaukee, and more and has been recognized through reviews as one of the Top 10 conference presentations on more than one occasion.

You can connect with Kirk on Twitter or Linkedin.

In 2023, Kirk had the privilege of speaking at the TEDx Billings on one of his many passions, Stop the Scale: Redefining Business Success.

Continue reading

Find what you're looking for here: