Multi-dimensional Product Segmentation: Road to PMax success 

The tides in digital advertising are shifting. Highly automated, all-encompassing digital campaign types, like Google and Microsoft’s Performance Max (PMax), ring in a new era of streamlined marketing at the expense of granular user control.

A key challenge with PMax is its broad-brush approach to product catalogs. In its quest for quick wins, PMax tends to prioritize best-sellers, sidelining a vast array of products with less performance data that might still align more closely with critical business objectives. This can result in campaigns that boost immediate sales but fail to build long-term profitability.

How do we address this? Product segmentation.

For some time now, product segmentation has been a weapon in a marketer’s arsenal, but PMax and similar “Black Box algorithms” have increased the importance even further as other tactical levers are removed. This creates the challenge of how to optimally highlight product data without splitting the campaigns too much.

Some advertisers try to manage this by segmenting products based on historical ads data. The problem with that: Not all products have this kind of historical data.

So what if there was a more nuanced way to align your product catalog with your broader business goals?

That’s where multi-dimensional product segmentation comes in.

Understanding Product Segmentation

But what even is product segmentation exactly? At its core, product segmentation is the art of dividing a vast product catalog into smaller, meaningful segments based on distinct characteristics. These characteristics most often include: brand, product type and number of conversions.

When done right, product segmentation allows for a more strategic approach to marketing. This ensures that each product group is promoted in a way that resonates with the right business goals. Ultimately leading to more effective marketing strategies.

What does ‘done right’ mean?

“Doing it right” is crucial in today’s highly automated digital ads landscape. Digital ads providers are moving further and further away from product-level adjustments and item-level bidding. 

These “Black Boxs” have removed the manual control retailers once had over their product libraries. They tend to consolidate data rather than segment it. Favoring a more unrestricted environment for their algorithms to operate. All while disregarding any strategic business incentives retailers might have. 

On top of that, Returns on Ads Spent (ROAS) and budget are the only tangible levers with these ‘Black Boxes’. However, relying solely on ROAS falls short of meeting the nuanced demands of scaling a business. Or achieving specific targets such as profitability or boosting Customer Lifetime Value (CLV).

The issues with Performance Max

Besides limiting the retailers’ control over their product lineup to ROAS-level adjustments, PMax suffers from a few more shortcomings that complicate efficient product segmentation strategies. Frequent readers of our blog are likely very aware of these, but to summarize them briefly: 

  • Lack of granular control: PMax and similar campaign types restrict advertisers to setting broad campaign goals and adjusting budgets. All without the ability to make detailed product-level adjustments. 
  • Data scarcity: PMax and similar ‘black boxes’ lack detailed data about your business beyond basic sales and efficiency metrics. This hinders precise product segmentation and campaign customization for specific business goals.
  • One-size-fits-all approach: These highly automated campaign types often adopt the path of least resistance, focusing on top-selling products regardless of their strategic importance or profit margins. This can lead to a lopsided campaign that overlooks potential hidden champions in the product catalog.

In other words: These highly-automated campaign types are strategically blind. In their quest to achieve the fastest and easiest wins, they completely ignore any strategic goals a retailer might have – whether that’s long-term growth, pushing specific product lines or ramping up profitability.

Ideally, product segmentation counteracts these issues. But here’s the thing: the traditional way of segmenting products tends to categorize product catalogs in a very broad, two-dimensional way. Mainly driven by how well a product sells or how cost-effective it converts clicks into sales. 

The problem with that?

The old way: Two-dimensional Product Segmentation

Sure. Two-dimensional product segmentation is certainly better than just throwing your vast product catalog into one bucket to see what fits. But it comes with several notable issues.

As its name suggests, two-dimensional product segmentation mainly focuses on two metrics when categorizing product catalogs:

  • Sales Volume: This measures the quantity of products sold. It identifies which items are the most popular or in-demand within a given time frame.
  • Efficiency (ROAS): This gauges the cost-effectiveness of advertising efforts, highlighting which products generate the most revenue relative to ad spend.

The issues with two-dimensionality

Don’t get us wrong: Sales Volume and Efficiency (ROAS) are crucial elements of any solid product segmentation strategy. But relying solely on these two metrics is like having a narrow diet; it might keep things running for a bit, but it also comes with several side-effects:

  • Reliance on past performance: Segmenting products based on Sales Volume and Efficiency is heavily dependent on historical data. This poses challenges when pushing new or niche items, as they don’t have any historical data yet.
  • Conflated metrics: Sales Volume and Efficiency are closely linked together. The problem with that? Marketers are segmenting products in even fewer dimensions than this method suggests. This leads to a narrower, less varied categorization of products. 
  • Missed opportunities: Two-dimensional segmentation often neglects significant factors that might highlight a product’s strategic importance to the business or its appeal to specific consumers.  

Two-dimensional product segmentation methods tend to focus on immediate performance indicators. All while overlooking the potential of products that might not be current top-sellers but have a high chance to contribute to the overall bottom line.

Metric2D-SegmentationImplications
Sales VolumeYesIdentifies top-performing products, useful for quick wins.
Efficiency (ROAS)YesAims to make advertising spend cost-effective.
Product AttributesNoLeads to a disconnect between product offerings and market demands.
Consumer Behavior InsightsNoMisses targeted engagement and growth opportunities.
Market TrendsNoMisses emerging demands and market shifts.
Profit MarginNoAffects overall profitability by not prioritizing high-margin products.
Inventory LevelsNoRisks promoting out-of-stock or overstocked items.
CompetitivenessNoIneffective pricing and differentiation strategies.
An overview of metrics accounted for in two-dimensional product segmentation and their implications for digital ads strategies.

Transition into multi-dimensionality

Like written above, the traditional way of segmenting products tends to rely on a one- or two-dimensional analysis – typically considering metrics such as ROAS, number of conversions, or conversion value accrued over the last 30 days.

Advertisers typically look at only one or two metrics and segment products based on those, ignoring other relevant metrics altogether. And that’s where the problem lies: The importance of a product to your business cannot be determined by a single or even two metrics. The ecommerce reality is more complex than that.

For example: consider a product with high revenue and ROAS. While this may sound good, it is important to also consider the profit margin of the product. For instance, what if there is a new version of the product that has not been sold yet but could attract more shoppers than the previous version with a lot of conversion data? In this case, looking at one or two metrics alone is not enough.

Moving beyond two-dimensional product segmentation can fundamentally change your marketing strategy. And there is an impressive new way to do this. A method that allows retailers to incorporate a vast amount of diverse metrics for a more robust segmentation strategy. 

We call it …

Multi-dimensional Product Segmentation

Why limit yourself to a traditional, two-dimensional approach when you can leverage an advanced methodology that offers a multi-faceted strategy for product segmentation? 

Multi-dimensional product segmentation transcends the basic intersections of volume and efficiency. Instead, it integrates them with a multitude of metrics with the help of advanced automation software. This offers online retailers an unprecedented level of control over their Performance Max algorithms.

To get an accurate picture of the performance and hidden potential of their products, retailers need to not only look at more metrics, but the right metrics. This new way of segmenting product catalogs represents a significant step up from the traditional two-dimensional approach by encompassing a wider range of various factors.

Depending on your business strategy, these factors may include:

  • Traditional Ad Metrics: ROAS, conversions, etc.
  • AOV: Focus on products that increase cart values.
  • Brand Push: Highlight in-house or specific third-party brands.
  • Competitive Pricing: Promote attractively priced items.
  • Profit Focus: Aim for sales that maintain profitability.
  • Stock Management: Advertise products with ample stock.
  • Warehouse Preference: Prioritize products from specific locations.
  • Key Product Lines: Feature signature product ranges.
  • New Releases: Push the latest products or clear old inventory.
  • Seasonal Relevance: Spotlight season-specific items.

How to segment in
multiple dimensions

But how does multi-dimensional product segmentation work exactly?

To start off, it’s important to understand that different online retailers face unique challenges in the digital advertising space. Not one business is the same, that’s why multi-dimensional product segmentation considers the specific business goals of online retailers. 

Depending on the individual goals, this might involve identifying products that have:

  • Strategic Importance: Products crucial to the retailer’s brand identity or those with higher margins.
  • Niche Appeal: Products with unique features or appeal to specific customer segments or needs.
  • Seasonality and/or events: Products that are particularly relevant during specific seasons or in response to upcoming events. For instance, prioritizing skiing equipment during the winter season or camping gear in the lead-up to summer.

But why is this important?

By focusing on products that resonate with the brand’s business needs and customer expectations, businesses can foster stronger connections with their audience, driving loyalty and repeat business. Multi-dimensionality helps marketers align product promotion and business goals for sustainable growth and a competitive advantage

Multi-dimensional product segmentation not only enhances the efficiency of advertising campaigns but also ensures that every marketing dollar contributes to long-term brand development and profitability.

Multi-dimensionality with data-driven insights

Once the unique business goals of online retailers are set, multi-dimensional product segmentation focuses on giving automated campaign types, like Google’s Performance Max, a serious data boost. Turning data deserts into lush, insightful advertising landscapes. 

Unlike 2D-segmentation, multi-dimensionality goes beyond just looking at the big picture; instead, it looks at each product within the catalog individually to help retailers identify the best campaign opportunities. 

With multi-dimensional segmentation, retailers are able to feed algorithms with superior business data ranging from:

  • First-party data: This includes sales figures, customer engagement metrics, and other internal data that offer insights into product performance and customer preferences directly from the retailer’s perspective.
  • Second-party data: Often obtained through partnerships, this data provides additional market insights and benchmarks. Offering a comparative view that enriches the retailer’s understanding of their position within the industry.
  • Third-party data: External market research, consumer behavior studies, and trend analyses contribute a broader perspective. This overarching market dynamics that could influence product performance.
Predicting the conversion potential of products with little data? Almost impossible.
It’s difficult to predict the conversion potential of items with little data.

Uncovering product categories

After processing these pieces of data, multi-dimensional product segmentation can help retailers uncover various product categories that might have been previously underexplored. Or find categories that haven’t been fully leveraged yet. These categories can include:

  • Specialized products: Unique or specialized items hidden in the longtail that cater to specific customer interests or needs.
  • High-margin items: Products that offer a higher profit margin but may not have the highest sales volume. Making them valuable for targeted promotions.
  • High B2B share: Product types that have demonstrated significant success in driving B2B conversions.

These product categories give online retailers a superior understanding which products should receive more focus, budget, and exposure. All of this is based on the products’ strategic importance, potential for growth, or alignment with current market trends and consumer behavior.

On top of that, retailers can achieve a better understanding as to which products are not only performing well currently, but also have the potential to drive sustainable growth in the long-run. Allowing them to invest in areas with the highest potential for return and maintain a competitive edge in the market.

Multi-dimensional product segmentation: More data, superior results.
Multi-dimensional product segmentation: More data, superior results.

Finding the hidden champions

Multi-dimensional segmentation extends beyond basic categorizations that sort items into broad, singular categories based on one attribute (such as price range and product type). 

With multi-dimensional product segmentation, retailers are able to incorporate sophisticated AI models in their strategy that help them compare their top-performing products and their shared characteristics – like brand, seasonality, margin, trends, etc. – to pinpoint similar but underrepresented products

These products often possess a strong potential for conversion, provided they are given the opportunity to be highlighted from the more dominant items:

  • Niche specialties: Products that cater to specific, often underserved segments of the market. These could include items with unique features, specialized uses, or that appeal to particular hobbies or interests.
  • Strategic brand items: Products that are crucial to the retailer’s brand identity or strategic goals, such as proprietary or flagship products, even if they haven’t been top sellers historically.
  • Value products: Items that offer high value to customers, either through competitive pricing or superior features, but haven’t been adequately promoted or positioned.

For instance, through this analysis, an online retailer specializing in outdoor gear might uncover that certain camping tents, although not currently among the top sellers, share key attributes with best-selling items. Such as high customer ratings, compatibility with popular hiking gear, and preferred materials. These tents might also perform exceptionally well during specific seasons or promotional events.

With their hidden champions identified, marketers are able to tailor campaigns that ensure that not just the current bestsellers are promoted. They uncover also products that have the highest potential to appeal to specific customer segments and hit strategic business goals.

Decathlon Logo
DECATHLON achieved sustainable growth with
multi-dimensional product segmentation:

Decathlon ROAS uplift
Read the Case Study

The strategic value of
multi-dimensionality

Moving beyond the nuts and bolts of how multi-dimensional segmentation works, it’s crucial to understand the strategic advantages it brings to retailers. 

Multi-dimensionality is about getting down to the nitty-gritty of the entire product catalog, helping retailers make smarter decisions about which products to focus on and how to best market them. It’s about fundamentally transforming how businesses connect with their customers, position their products, and ultimately, enhance their bottom line.

  • Increased visibility of underperformers: Multi-dimensional product segmentation identifies products with scarce data or low current performance and enhances their visibility. This helps tapping into untapped growth potentials.
  • Refined product positioning: Multidimensionality considers a broad spectrum of factors to highlight products of strategic importance. This helps differentiate offerings from competitors.
  • Enhanced ROI: By fine-tuning marketing efforts, multi-dimensional analysis aids marketers in concentrating on impactful, highly-targeted campaigns. This ensures marketing budgets are allocated to campaigns with tangible results, thus maximizing profitability.

Unlike the old way of just looking at Sales Volume and ROAS, multi-dimensionality helps retailers match their products with what your customers are actually looking for. This means you can do a better job of getting the right products in front of the right people, leading to more sales and a stronger position in the market.

In other words: multi-dimensional product segmentation is not just about growing you business. It’s about growing your business more efficiently.

Not everything’s perfect

Here’s the thing, though: tackling the multi-dimensional product segmentation complexities without specialized software is simply not feasible. 

Sure, in-house teams can work with a plethora of filters and rules to label each individual product for a while. All the same, this leads you to spending hours and hours on error-prone manual tasks that might only scratch the surface of what is possible in multi-dimensionality:

  • Time consumption: The sheer volume of data and the granularity required for multi-dimensional analysis mean that marketing teams could spend countless hours on this task alone. Time that could otherwise be invested in strategic and creative campaign development.
  • Human error: With so many variables and such complex interrelations between them, the likelihood of oversight or miscalculation is high. These errors can lead to misinformed decisions, potentially overlooking hidden gems in the product catalog or misallocating marketing resources.
  • Data overload: Even the most adept marketers can find themselves swamped by the data deluge. The human brain alone simply can’t process the massive amounts of intersecting data that’s needed. We’re just not built that way.
Doing multi-dimensional product segmentation manually in not feasible - even with just three variables.
Doing multi-dimensional product segmentation manually in not feasible – even with just three variables.

Tech as a guiding rope

The good news is, a multi-dimensional software solution, not unlike the one we utilize at smec, can really save the day here. These tools regain control of any product segmentation tasks by feeding the ads platforms’ automated algorithms with superior data-points outright.

Make no mistake: Multi-dimensional tech solutions are not a replacement for the strategic insights of skilled marketers but a powerful amplifier of their capabilities.

These software solutions address the key pain points of manual segmentation by offering:

  • Speedy data crunching: They zip through tons of data quickly. Something that would take ages to do manually.
  • Smart connections: They’re great at figuring out how different data points relate to each other, giving retailers insights they might miss on their own.
  • Turning data into decisions: By taking over the busy-work of juggling sheer unlimited amounts of data, in-house teams have more time in their hands to tailor more engaging campaigns.

To sum it up

MetricTwo-dimensionalMulti-dimensionalImplications
Sales VolumeYesYesBroader analysis in a multi-dimensional approach enhances product promotion strategies.
Efficiency (ROAS)YesYesMulti-dimensional considers broader factors, optimizing beyond cost-efficiency.
Product AttributesNoYesEnables precise targeting and product differentiation.
Consumer Behavior InsightsNoYesSupports personalized marketing, improving customer engagement.
Market TrendsNoYesEnsures marketing agility and relevance to consumer trends.
Profit MarginNoYesAligns promotions with business profitability.
Inventory LevelsNoYesOptimizes stock management.
CompetitivenessNoYesInforms competitive pricing and product positioning.
Strategic ImportanceNoYesHighlights products essential to brand and strategic goals.
Niche AppealNoYesIdentifies specialized market opportunities.
Two-dimensional vs. multi-dimension product segmentation: Metrics and their implications for digital ads strategies.

Multi-dimensionality:
Your new best friend?

Let’s cut to the chase: Multi-dimensional product segmentation might be the game-changer you’ve been waiting for. Especially when dealing with those set-it-and-forget-it campaigns types that just can’t seem to grasp the full potential of your product catalog. 

When embracing a multi-dimensional product segmentation framework, retailers can get back control over their campaigns with new levels of precision in their marketing strategies. Leading to improved ROI and a deeper connection with their customers. 

The downside.

But while all of this sounds like sunshine and roses, there is a glaring downside to multi-dimensionality: It’s simply unfeasible to do it with manual labor alone. 

By leveraging smart technology, marketing teams can streamline this process, saving precious time that can be redirected towards more impactful and strategic tasks, enhancing overall efficiency and effectiveness.

The upside!

Lucky for you: That’s exactly what smec specializes in.

Putting online retailers’ strategic business goals into the forefront, smec’s multi-dimensional product segmentation solution is purpose-built to sideline PMax’s inherent shortcomings. Our software feeds the algorithms with superior business data provided by online retailers, to give them the granular, item-level control campaign types like PMax are denying them.

By combining strategic data integration, predictive AI insights, and a focus on uncovering your hidden champions, multi-dimensional product segmentation provides a level of insight and control unattainable with Performance Max out of the box. And we are more than happy to help you achieve this.

Enter the Multiverse

To truly harness the potential of multi-dimensional product segmentation in your digital ads campaigns, having a partner like smec can make all the difference. If you’re looking to refine your ecommerce strategy or maximize the impact of your paid search efforts, we’re here to help.

Our team of PPC experts is ready to dive into your current strategies, guiding you through the ins and outs of multi-dimensional product segmentation. Let us show you how to ensure that every ad spend aligns perfectly with your crucial business goals.

Wanna know more? Listen to our podcast!