Meta Ads Structure for B2B SaaS – Retargeting Part

B2B buying journeys are long. Building trust is key, but you also want to effectively showcase the product value through various use cases. Long-term communication with your already engaged audience is crucial.

Situation

The goal is to get our ads in front of as many people in the audience as possible, ideally with different messages at a reasonable frequency. The restuls are not quick sales as in B2C but in incremental increase of hand-raisers over time.

This is the second part of my Marketing Festival presentation recap. (check the first part focusing on structure for acquisition). For retargeting, I use a completely different campaign and ad set structure.

The principle is quite simple: set the campaign objective to reach, and use multiple ad sets with strict frequency caps.

With this setup, the Meta algorithm does not optimize ad delivery based on conversion data. Therefore, it is not suitable for larger acquisition targeting but is effective for retargeting or smaller audiences that have shown some level of intent (up to a maximum of 80k users in the audience).

The Structure Logic

In the reach campaign, I use multiple ad sets, all with exactly the same targeting (essentially duplicates), but featuring different ads representing distinct messages.

For example, if there are 4 ad sets with different message:

1. Social Proof (case study demonstrating how your product helps a big brand)
2. Product Value (use case showcasing the problem your product solves)
3. Thought Leadership (educational content)
4. Action (request a demo)

Each ad set contains ads with the same message, which can include various formats such as video ads, carousels, or static images. The key is consistency in delivering the message to the user.

Frequency

Frequency capping is the key because you can regulate that the messages are split evenly.

For instance, with these 4 ad sets, you can aim to show the message twice per month (frequency cap of 1 impression every 15 days). The maximum impressions one person can receive is 8 per month in total but most of these are different.

Number of Ad Sets

You can’t have too many ad sets.

Look at this example: when a user enters your targeting audience, any of the ad sets can show the ads since the user hasn’t seen any ads from the ad sets (none of the ad sets have reached the frequency caps).

If the budgets were unlimited, the user would see ads from all active ad sets on the first day, with no ads shown on subsequent days (depending on the frequency limit, as in our example, a full 15 days without ads).

With fewer ad sets and a shorter frequency cap, new users wouldn’t see as many ads on the first day, and there wouldn’t be long gaps between impressions. So don’t go over 20 ad sets.

Budget

Similarly, the budget can influence ad distribution. With a limited daily budget (and consequently lower bids), not all ad sets would win the ad auctions. Only the ad set and ad with the highest ad rank would enter the ad auction, smoothing out peaks and gaps in ad delivery.

Setting the budget at the minimum ($1 a day) or adjusting it based on estimated reach (10-15% of the audience size) and real frequency data observed in the account can help optimize ad delivery.

Audience Size

I aim for an audience size of 20k-50k, combining engagement audiences with website retargeting or customer lists. However, some projects may have smaller audiences, such as 5k, while others may reach up to 80k (or my max 100k for less active IG/FB audiences).

Combinations

This was a simplified example for easier understanding. In reality, you can manipulate these four variables to achieve your desired outcomes.

In planning, I consider extremes:

• First-week experience: How many ads someone can see on the first day and how the first week would look.
• Monthly maximum impressions per day: When someone is extremely active on the platform and could potentially see all our ads, what is the maximum distribution.

Initially, it’s essential to test and monitor frequency on a weekly basis, comparing estimates with results. Having used this strategy for almost two years, I can estimate how the average user scenario would unfold based on audience size, number of ad sets, frequency, and budgets.

Stay tuned for concrete examples from different projects demonstrating how this strategy is implemented!