Want to get more out of your product campaigns? Start with segmentation

In this article, we will show you:

• How to prepare for product segmentation
• How to segment products
• How does product segmentation work in practice?
• What comes after creating segments — working with data over time

Using concrete examples and tools (e.g. Mergado), we’ll show you how to take control back into your own hands. Efficiently, automated, and with better results.

The problem? A catalog without a strategy

When you advertise your entire catalog at once, you’re paying just as much for the top products as you are for the ones that don’t sell. It’s like trying to win a bus race — everyone’s driving, but no one’s going extra fast.

This is the most common problem in e‑commerce campaigns: one strategy for everything. Regardless of performance, margin, price or season.

The result?

• You spend money on ineffective advertising,
• you don’t keep track of what’s making money and what’s not,
• and the ad systems’ algorithms are working blindly.

The solution? Segmentation that gives campaigns direction.

With segmentation, you can:

• Pull performance from top products,
• cut budgets from weaker products,
• and most importantly, divide your strategy so that each product group gets the care it deserves.

Segmentation is the foundation of the performance structure.

With it:

• you set different ROAS targets for different segments,
• optimise bidding and budget allocation,
• and prepare campaigns that can be managed, not just hoping they hit something.

And how do the agencies themselves approach segmentation?

Most often, it’s based on the following data and criteria:

• Performance metrics (e.g. ROAS, number of conversions, “zombie” products),
• standard attributes from the feed (brand, category, availability),
• business metrics such as margin or purchase price,
• external data sources (e.g. competitor pricing, Analytics or Facebook data),
• seasonality and promotions (e.g. Black Friday products),
• suppliers whose products need to be promoted,
• custom labels created e.g. in Mergado,
• or custom scripts that automate more complex logic.

So each segment can be built a little differently, depending on what the online store needs and what data it works with.

How to prepare for product segmentation

We already know that you can’t do without segmentation in today’s performance advertising. But for the whole process to work well, you need to prepare for it thoroughly — not only technically, but also strategically.

It is crucial that the e‑shop and the PPC specialist are clear at the outset about what they want to achieve with segmentation. Do we want to increase sales? To improve the margin? Get rid of non-selling goods? Promote new products?

👉 Clear goal = meaningful structure. If it’s missing, the whole segmentation will be a complication without benefit.

Here’s an overview of what you need and how to proceed:

✅ 1. Determine why you want to segment

There is no point in segmenting just for the sake of segmenting. First, be clear about what goal you want to achieve by segmenting:

• Increase ROAS?
• To limit spending on loss-making products?
• To increase visibility of new products or seasonal items?

The objective will then determine how you segment — perhaps by performance, margin, availability, season or brand.

✅ 2. Start simple but systematically

There is no need to create ten segments right away. Two is enough to start with — for example, high-performing vs. low-performing products. Once you get a feel for the system and get results, you can expand the segmentation.

In practice, you often work with different terminology and methodologies. Some divide products according to the BCG matrix, which creates categories like Stars, Question Marks, Cash Cows and Dogs.

Find a system that makes sense to you — the key is to make it understandable, repeatable and tied to data.

✅ 3. Make sure you have good product data

The foundation for success. In your feed, you should have at least:

• product ID and name
• price (including promotional)
• availability/​stock availability
• categories (yours and Google’s)
• brand, images, description
• identifiers (EAN, MPN…)
• parameters such as size, colour, etc.

✅ 4. Get advanced data

For more advanced segmentation, you will need data outside of the normal feed, for example:

• margins or purchase prices
• performance data from Google Ads, Analytics, Heureka, etc. (conversion, PNO, ROAS)
• competitor data (e.g. from assortment report)
• own company data (e.g. sales, labels like “new”, “sale”, “season”…)

✅ 5. Prepare the right tools

Google Ads alone is not enough for segmentation. You will need:

• A feed editing and management tool
• analytics tools and access to performance data
• spreadsheet tools to prepare and import data
• a system for feed auditing, additional feeds, error checking and tagging
• (We’ll talk about how to do all this with Mergado in the next section.)

✅ 6. Prepare audience and advertising materials

Especially for Performance Max, it’s important to have the materials (images, text, videos) and signals for each segment ready, so the system knows what and who to show.

How to segment products

Now that you know what you need for segmentation (data, tools, clear logic), let’s look at what you can specifically segment products by.

Segmentation logic is always based on business goals and available data. Here are the most common and practical criteria used in advertising:

🏷️ 1. Segmentation by product characteristics

You’ll find this data in your product feed — and if you don’t, you should get it there as soon as possible.

• Category: basic segmentation by e‑commerce category. Allows you to manage bidding by product type — different strategies for electronics, different for apparel.
• Brand: products segmented by brand — perhaps for margins, exclusivity, or targeted marketing.
• Price: Different price levels require different strategies — cheap products are often impulse sold, while more expensive ones need more thoughtful promotion.
• Margin: If you have access to margin data, this is a great basis for a ROAS strategy. The higher the margin, the more room for investment.
• Inventory: divide products into those you have in stock (and want to sell now) and those with longer availability or zero stock.
• Parameters: colour, size, material… you can use for example, to promote specific variants or seasonal campaigns.
• Supplier: When you need to promote products from a particular supplier, perhaps because of agreed terms and conditions or higher margins.
• Special labels from the online store: New item, sale, bestseller — labels you already have in your system but need to get into the feed and use in advertising.
• Identifiers: EAN, GTIN, MPN or custom IDs — use for precision targeting or data matching.

⚡ 2. Segmentation by performance

Here you are already working with data from advertising systems, analytics or Mergado exports.

• General performance (clicks, conversions): split into products that sell and those that just draw budget.
• ROAS/PNO: Ideal basis for a budget strategy — profitable products can get a larger portion of the budget, loss-making ones are better to cut back.
• Sell-through: How the product sells across channels, not just from PPC. Helps to pick out best sellers.
• Zombie products: those that don’t show up or don’t perform — often a hidden treasure or unnecessary stowaway in a campaign.
• Competitor data: Use to, for example, push products with better pricing more than the competition.
• Products without data: For new products or those that are unfamiliar to the system, a more cautious strategy is needed — perhaps a data collection campaign.

💼 3. Segmentation by business strategy

This is where you go beyond performance and work with a broader context.

• Seasonality: for example, summer vs winter goods — segmenting by season helps you be flexible.
• Events and promotions: special campaigns for promotional products, like Black Friday or sales.
• New arrivals and sales: New products need visibility, old stock needs clearing out.
• Banned products: Products that can’t be advertised (e.g. due to Google rules) need to be effectively hidden.
• Reporting and analytics: segmentation helps you report clearly on the performance of different parts of the range — ideal for agencies and internal teams.

And don’t forget — you can combine segments. For example: performance + stock + high-margin products = ideal target for more aggressive advertising.

How is product segmentation done in practice?

Segmentation is a process that connects your business goals, available (often augmented) data and Mergado’s capabilities. Here’s a step-by-step guide on how to do it:

1. Figure out what data you need for segmentation

A basic feed will usually give you ID, name, price, brand, category, availability, parameters. But for more advanced segmentation, you often need additional data — and you need to either add or import that data.

Most common external data:

• Purchase prices/​margins (for segmentation by profitability)
• Performance metrics from Google Ads / Analytics (e.g. ROAS, number of conversions)
• Competitor data (e.g. from Bidding Fox or Heureka Assortment Report)
• Inventory/​sales from online store or ERP
• labels from the online store (e.g. promotions, bestseller, new)

2. Prepare a data file for import into Mergado

Download data from other systems (e.g. CRM, accounting, GA, Google Ads, online store) into a file — usually CSV or Google Sheets. The key is that it should contain:

• a matching element (typically a product ID)
• and columns with the data you want to add (e.g. margin, roas, supplier, stock_pieces…)

3. Upload data to Mergado (Import data file)

In Mergado, you use the “Import data file” rule:

• specify the source of the file — it can be a static file or a URL (e.g. a published Google Sheet)
• for proper matching, set the table so that the first column contains the matching element (e.g. product ID),
• determine which elements the data will be stored in (create them if they don’t already exist)

4. Create segments and custom labels using rules

Once you have your data in Mergado, create your product selections (queries):

• e.g. high margin, stock more than 5 units, ROAS > 600%, bestsellers
• Use rules like “Bulk rewrite by query” to populate your custom elements or directly custom_​label 0 – 4 to get them into Google Ads.

5. Segment your Google Ads campaigns

In Google Ads, use custom labels to segment your product groups:

• different campaigns/​sets = different strategies, bidding, goals, budgets
• e.g. high-performing vs. low-performing products, high-margin goods, Black Friday assortment…

What happens after creating segments – working with data over time

Segmentation is not a one-time task. It is the beginning of a cycle that requires regular monitoring, evaluation, and optimization. It is at this stage that the true power of data and automation becomes apparent.

1. Monitoring segment and campaign performance

After deploying the new structure, monitor how individual segments are performing:

• In Google Ads, examine campaign performance by custom labels – i.e., by the segmentation you created in Mergado.
• Track metrics such as conversions, revenue, ROAS/PNO, CTR, as well as the number of impressions and clicks.
• Also pay attention to individual products: which ones are performing well, which ones are stagnating, and which ones are generating costs without results (“zombie” products).

2. Regular data and segmentation updates

Performance data changes every day – in order for segmentation to reflect the current reality, you need to regularly update the imported data in Mergado.

• Set the reports from which Mergado draws data to update automatically – for example, monthly or every 14 days, depending on the type of product. Add-ons or scripts can help.
• This is the only way to ensure that products remain correctly classified according to current margin, ROAS, or availability.

3. Bidding and budget optimization

• Increase budgets and bids for high-performing segments (e.g., “index” or “near-index”).
• You can reduce investment in ineffective segments or tighten ROAS targets (“under-index”).
• If you have data on inventory or prices in Mergado, you can use it for more advanced bidding logic.

4. Dynamic adjustment of segment content

• Products move between segments — e.g., due to stockouts, changes in demand, or deteriorating performance.
• Automatic segmentation in Mergado ensures that the product is moved to another segment based on current data.
• But be careful: too frequent switching (e.g., daily) can disrupt campaign stability.
• You can “lock” key products by manually overwriting the label in Mergado or Merchant Center.

5. Expanding and refining the structure

• Start simple: two campaigns (e.g., high-performing vs. low-performing).
• Once the structure has proven itself, you can add other segments (e.g., “no-index,” “under-index,” “seasonal”).
• But remember: the more segments you have, the more data you need — ideally at least 50 conversions per campaign in the last 30 days.
• If you don’t have that many, stick to a simpler structure that will be stable.

6. Special segments – zombies, new products, seasonal goods

• Zombie products: few impressions, no clicks – give them a chance in a special campaign, perhaps with bidding to maximize clicks.
• New products, promotional goods, seasonal assortment: distinguish them with labels and create your own campaigns where you will treat them differently (e.g., higher bids for new products).
• Mergado will help you create labels automatically – e.g., based on the date of stocking or type of promotion.

7. Connection to business goals and broader optimization

• Segmentation gives you a tool to bend the machine learning of advertising systems to achieve your business goals.
• Consider tracking POAS (profit margin return) instead of just ROAS if you have margin data.
• You can also use the results of segmentation outside of campaigns – for example, when deciding on your product range or pricing.
• And don’t forget: campaigns aren’t everything. Even perfect segmentation needs a strong foundation – product cards, names, descriptions, images, promotions.

Segmentation is not a goal, but a journey

Effective campaigns don’t start with advertising systems, but with quality data and good segmentation. Once you divide your products according to performance, margin, or other relevant parameters, you can work with your campaigns in a much more targeted way. And most importantly, you’ll stop wasting your budget on products that currently have no chance of converting.

But keep in mind that segmentation is a living system. Product performance changes, seasons come and go, and the market is different every day. That’s why it’s important to have a tool that allows you to respond quickly and automate adjustments. Mergado can be a powerful partner in this — from segmentation and rules to connecting data from different sources.

Want to get started with segmentation but need some help? Sign up for a webinar on performance segmentation or book a free online consultation right away – we’ll be happy to advise you on your specific case.