Skip to main content

How to Plan for Segment Creation

Reading time estimate: 8 minutes

Updated over a year ago

Intro

It can be difficult to start creating your own segments without a useful idea of where to start or how to judge effective results. The sequence of beginning with no segments and ending with effective segments can be broken down into simpler steps that are much easier to achieve separately.

Turning the plain language of "What are my individual definitions of good and bad products?" into a spreadsheet showing X% of spend saved and Y% of ROAS increased is accessible through determining what success means on your terms.

This guide is intended for onboarding purposes.

Read the full documentation here: Product Insights: Segments.

Segment Planning

While it can be possible to approach segment creation through iteration by exploring different combinations of metrics in various contexts, it's often easier to start at the conceptual stage.

The ROI Hunter product strategies are widely useful with default settings, but can offer even more benefit when modified for your personal interests.

You may even be interested in integrating your custom data in order to have more nuanced insights for your own purposes. Read more here: Business Data Integration.

Questionnaire Walkthrough

Try going through the steps below to build an idea of how you can start segmenting your products. Start by considering answers to these questions:

Step 1: Define your success

  • What are your overall business goals, both short term and long term?

  • What are your individual KPIs (e.g. metrics) tied to these goals?

  • Do you have any specific target you want to achieve in a given month, quarter, year, etc?

User A - Revenue

User B - Profitable growth

My team's main business goal is revenue growth. We measure it with the overall ROAS of our marketing activities. We should reach an average ROAS of 3 across all our Meta dynamic ads by the end of 2024.

My team's main business goal is profitable growth. We measure it with the revenue generated by our dynamic ads on Meta plus the gross profit generated (or margin percentage). We want to grow 110% YoY with 30% margin on those revenue gains.

Step 2: Based on your goals, define your good and bad products

  • What are your good products, that support your business goals?

  • What are your bad products, that block your business goals?

  • Is there a specific group of products that might be interesting for your particular type of business, such as hero products from the brand perspective, 1st party products, high margin products, etc?

User A - Revenue

User B - Profitable growth

  • Good products generate good revenue compared to their spend (they have high ROAS).

  • Bad products don't generate enough revenue, but waste our spend (they have low ROAS).

  • Good products generate high revenue and high margins (my biggest profit opportunities).

  • Bad products have negative margins, or are the ones not generating any revenue.

Products with potential for me include:

  • Probably products with High ROAS but low spend (I can explore those products, try to invest more budget into them, and see how they behave).

  • New Arrivals have big potential to catch the interest of my customers since they are brand new.

  • Discounted products may help me to raise ROAS because they have attractive price tag, as I'm not optimizing towards profitability now.

Products with potential for me include:

  • Hidden Gems that generate a lot of revenue, but are not promoted in my dynamic ads.

  • New Arrivals have a lot of potential to catch the interest of my customers as they are brand new, and they are also sold at full price so they represent a big profit opportunity.

  • Trending products might often be young products currently in season so they are barely discounted, and my customers are interested in them. These can be products with high Product Detail Views from GA or High Impressions from Google.

Step 3: Convert these goals and insights into product filters

  • In the ROIH app, navigate to the Product Insights tool

  • Use the Metrics and Attributes filtration to outline your good and bad products

User A - Revenue

User B - Profitable growth

I want to reach an average ROAS>3 across all my dynamic campaigns.

  • Good products filtration: Meta ROAS≥3

  • Bad product filtrations: Meta ROAS<3, Spend>100 EUR (you do not want to disqualify products which haven't had an opportunity to shine yet)

I want to reach an average 30% profit of growing revenue.

  • Good products filtration: Revenue top 20% and Margin>30%

  • Bad product filtration: Margin<0%, Revenue bottom 10%

Products with potential for me:

  • Probably products with High ROAS but low spend, so ROAS≥3 and Spend bottom 10%

  • New Arrivals filtration: Product Age<14 days

  • Discounted products filtration: Discounted by X%

Products with potential for me:

  • Probably products with high revenue but no spend, so Revenue top 20% and Spend bottom 20%

  • New Arrivals filtration: Product Age<14 days

  • Trending products filtration: Product Detail Views top 20% or Google Impressions top 20%

Step 4: Test different values to find the best definitions

If there are not enough products filtered after your initial definitions are calibrated, then experiment and play around with the filtration criteria.

  • Maybe your product inventory is not ready to jump straight to ROAS 3, and you need to start lower with 2 or 1.5 and increase your filters in time once you encounter first positive changes.

  • Good product segments should be big enough so they can make a significant impact if you invest more budget into them.

  • Bad product segments should be small enough so they won't disturb your BAU product performance if you limit them from your campaigns.

  • Remember to check the charts to get to know your segments better.

  • Don't be afraid to mix and match segments or to use more than one filter.

Step 5: Save the filters and apply them to your campaigns

Follow the next article to finish your workflow.


Next up, learn more about how to convert your insights into campaign applications.

Previous page:

Current page:

How to Plan for Segment Creation

Next page:

Did this answer your question?