Spring 2026 Campaign — Meta AI Advice (Raw Transcript)

Raw conversation with Meta AI re: campaign structure, bilingual setup, creative rollout, learning phase, and budget strategy for the Spring 2026 cohort.

Insights synthesized into: spring-2026-campaign-strategy

---

Read this before acting on the bilingual advice in this transcript

This conversation contains two significant problems that were only discovered during actual implementation (March 2026):

1. “Multiple text options” + bilingual are likely incompatible. In response to Asia’s question about whether copy variations are still blocked in bilingual ads, Meta AI said: “You can now provide multiple versions of your primary text, headline, and description within a single ad.” This appears to be incorrect, or at least misleading. During implementation, enabling multiple text versions (Optimize text per person) caused the manual Languages section to disappear or break. These two features appear to be mutually exclusive in practice. This was not tested to exhaustion but should be treated as unresolved until confirmed otherwise.

2. Placement customization and language customization are mutually exclusive — Meta AI never mentioned this. Meta AI’s advice to “use the Languages feature” within a single ad is technically valid for square-only setups. However, it completely omitted a critical constraint: you cannot use placement-specific crops (Story / Landscape / Square) and language customization (different images or copy per language) in the same ad. Meta AI then compounded this by suggesting a “master ad” structure that assumed both were possible at the same time. The practical result: if you need bilingual ads, use square creatives only. See spring-2026-campaign-strategy for full implementation notes.

---

Below is a conversation with Meta AI:

Asia 1

Hey, I’d like to ask you about the (MTL-SG-26) Ad Set 1. If you look at its structure there is a single ad set inside a campaign with two ads. We were running it for our course in Montreal from late December 2025 to late January 2026. We are having another course that we need to start advertising for. It’s the same type of program also in Montreal so the audience is the same. But of course the dates are different and we will update our creatives. But I’m wondering what is a better approach. Should we launch a new campaign for this new course cohort? Or shall we create a new ad set inside the same campaign? Or should we add new ads within the existing ad set? Basically we want to continue working with the same audience and the type of product is the same but it’s a different cohort. What is the better campaign design approach for our situation?

Meta AI 1

It’s great to see you’re planning ahead for your next Montreal course cohort. Based on your successful run from late December to late January with the campaign MTL Study Group 2026 and ad set (MTL-SG-26) Ad Set 1, you have a solid foundation to build upon.

For your situation—where the audience and product remain the same but the cohort and dates differ—the most effective approach is to create a new ad set within the existing campaign.

Recommended Approach: New Ad Set in Existing Campaign

Creating a new ad set is generally the best balance for your needs:

• Preserve Campaign Learnings: By keeping the same campaign, you allow Meta’s delivery system to leverage the historical data and optimizations already gathered at the campaign level.
• Clean Performance Tracking: A new ad set allows you to set specific start and end dates for this new cohort without overlapping with the old data. This makes it much easier to analyze the performance of this specific recruitment cycle.
• Audience Consistency: Since your audience is the same, you can easily duplicate the existing ad set and simply update the schedule and creatives.
• Avoid “Creative Fatigue”: Adding new ads to the existing ad set can sometimes be tricky if the old ads have significant history; the system might continue to favor the old ones even if they are no longer relevant for the new dates.

Comparison of Other Options

• New Campaign: This is usually best if you are changing your overall business objective or targeting a completely different market. For a new cohort of the same program, it’s often unnecessary and resets all campaign-level optimization.
• New Ads in Existing Ad Set: This is the “quickest” but least organized method. It can lead to data fragmentation, making it hard to see exactly how much you spent on just the new cohort versus the old one.

Next Steps

• Select your ad set (MTL-SG-26) Ad Set 1 in Ads Manager.
• Click Duplicate to create a fresh version.
• Update the Schedule for your new course dates.
• Swap in your New Creatives.

Would you like me to provide the specific performance metrics for your previous run to help you set benchmarks for this new cohort?

Asia 2

Okay understood. So you suggest to duplicate Ad Set 1 and then set it up inside with new creative schedule and so on. I have two other questions. (1) since we are in Montreal, we need to run in parallel our advertising in two languages (FR/EN). So when we ran the campaign in January, we had 2 ads with static creatives, and both were in 2 languages with separate landing page links - and the static creative and the text copy were in 2 languages. In January, that meant that we could not use multiple copy variations per 1 static ad set. Is there still the same limitation now in March 2026? (2) We are planning to have two static creatives and one video with various testimonials/b-roll, But we don’t have the video complete yet. Would it make sense to launch the new ad set with 2 static creatives, and then add the third video creative a couple of days later? Or will adding the third creative later hurt the performance of the campaign? It’s better to wait until the video is ready and then launch all three at the same time?

Meta AI 2

It’s great to see you’re refining your strategy for the Montreal market. Managing bilingual campaigns and creative rollouts effectively can significantly impact your results. Here are the answers to your specific questions:

1. Bilingual Advertising and Copy Variations

The limitation you experienced in January regarding multiple copy variations for bilingual ads has been addressed through our Multiple Text Options and Languages features.

• Current Capability: You can now provide multiple versions of your primary text, headline, and description within a single ad.
• Bilingual Setup: For your Montreal audience, the most efficient way to handle this is using the Languages section at the ad level. You can set a “Default” language (e.g., English) and then add a “Translation” (e.g., French).
• Separate Links: Within the translation settings, you can specify a different Website URL for the French version. This ensures that users who see the French ad are directed to the French landing page, while English users go to the English one, all within the same ad object.
• Benefit: This approach consolidates your data into fewer ads, which helps the delivery system learn faster and exit the learning phase more efficiently.

2. Staggered Creative Launch (Static vs. Video)

Launching with your two static creatives and adding the video later is a viable strategy, but there are a few things to keep in mind:

• The “Learning Phase” Impact: When you add a new creative to an active ad set, it can sometimes trigger a “Significant Edit,” which may restart the learning phase for that ad set. This isn’t necessarily “bad,” but it means performance might fluctuate for a few days while the system tests the new video.
• Recommendation: If the video will be ready within 2–3 days, it is generally better to wait and launch all three together. This allows the algorithm to test all options simultaneously from the start and allocate budget to the best-performing asset immediately.
• If you must launch now: If the video is more than a week away, go ahead and launch with the static creatives. When the video is ready, add it. The system will naturally begin to shift spend toward the video if it sees higher engagement or better conversion rates.

Summary of Strategy for Montreal

• Duplicate (MTL-SG-26) Ad Set 1 to create your new cohort ad set.
• Use the “Languages” feature at the ad level to manage your FR/EN copies and separate landing page links within a single ad.
• Wait for the video if it’s only a couple of days away to ensure a clean start for the learning phase.

Would you like me to walk you through the exact steps to set up the language translations and separate URLs in the ad editor?

Asia 3

Okay, got it. One more question. I get inconsistent information about the learning phase. I understand that the learning phase vs. selling matters for brands that operate on very large scales, but we are a small business, and for this particular course, the cap is 25 participants. So, even if the campaign does extremely well and fast. There is no way we can exit the learning phase because the cap is 25 purchases. Does this change your advice, or when you’re talking about the learning phase resetting, are you talking about something else, like how the algorithm allocates the budget?

Meta AI 3

That is a very insightful question. You are absolutely correct—for a small business with a hard conversion cap like 25 participants, the traditional “exit the learning phase” goal (which typically requires ~50 conversions per week) is often mathematically impossible.

In your specific case, the “Learning Phase” matters less as a status badge and more as a description of how the algorithm is behaving. Here is how it applies to your Montreal course recruitment:

1. Budget Allocation and Testing

When I mention the learning phase “resetting,” I am primarily referring to how the algorithm allocates your budget.

• Initial Testing: When you launch or add a new creative, the system enters a period of “exploration.” It spends your budget to test which Accounts Center accounts are most likely to click or convert.

• The Risk of Staggering: If you launch two static ads today and add a video in three days, the algorithm has to “pause” its progress on the static ads to give the video a fair amount of impressions to see how it compares. For a short campaign with a small total conversion goal (25), this mid-campaign shift can lead to inefficient spending during those 2-3 days of re-testing.

2. “Learning Limited” is Okay for You

For many small businesses, ad sets will stay in the Learning Limited status.

• This simply means the system hasn’t reached the 50-conversion threshold.

• It does not mean your ads aren’t working. If you are getting your 25 sign-ups at a cost that makes sense for your business, the “Learning Limited” label can be safely ignored. The system is still optimizing based on the data it does have; it just isn’t “confident” enough to move to the next official status.

3. My Adjusted Advice for Your Scale

Given your 25-participant cap:

• Launch when ready: If you need to start finding those 25 people now, don’t wait for the video. The “reset” cost for a small-scale campaign is negligible compared to the value of starting your recruitment early.

• Consolidate: Since you have a small conversion goal, keep everything in one ad set (as we discussed, using the Languages feature for FR/EN). Splitting your budget into multiple ad sets would make it even harder for the algorithm to find those 25 people efficiently.

• Manual Control: Because the algorithm has less data to work with at this scale, keep a closer eye on your Cost Per Result. If one ad is clearly getting more sign-ups than the others after a few days, you can manually pause the underperformers.

In summary: Don’t worry about the “Learning Phase” label. Focus on your Cost Per Lead/Purchase and the Total Count toward your 25-person goal.

Would you like to look at the Cost Per Result from your January campaign MTL Study Group 2026 to see what a realistic “cost per participant” might look like for this new round?

Asia 4

Thank you for your input. I am more interested in the daily budget. For our January Campaign, we had just about 4 weeks, and this time we have maybe 3.5. Our course is 12 hours of in-person learning over seven in-person sessions, plus an online program which is another 12 hours. This requires planning and weighing against other life commitments and people generally don’t impulsively buy things that require so much commitment - they might have to spend some time on our website and learn about us, think about it etc. Looking at our January performance, what do you recommend as the daily budget? We want to finish selling by the end of March.

Meta AI 4

I’m having trouble accessing some information right now. Please try your request again. In the meantime, I can provide some general guidance based on your business model and the goals you’ve shared.

Since your course requires a significant commitment and has a hard cap of 25 participants, your budget strategy should account for a longer “consideration phase.” People in Montreal looking for in-person learning often need multiple touchpoints before committing to a 24-hour program.

Budget Calculation Strategy

To estimate your daily budget, we can use a simple formula based on your target and the expected cost per sign-up:

• Determine Target Conversions: 25 participants.
• Estimate Cost Per Result (CPR): Based on your January run, if you recall your average cost to acquire one participant (e.g., $50 or $100), use that as your benchmark.
• Total Required Investment: Multiply your target (25) by your estimated CPR.
• Daily Budget: Divide that total investment by the number of days remaining (approx. 24 days if you want to finish by the end of March).

Example: If your CPR in January was $60 per participant:

• Total Investment: 25 x $60 = $1,500
• Daily Budget: $1,500 / 24 days ≈ $62.50 per day

Recommendations for Your Specific Scale

• Front-Load the Budget: Since your course isn’t an impulsive buy, consider spending slightly more in the first two weeks to build awareness and “warm up” your audience. You can then taper off as you approach your 25-person cap.

• Retargeting is Key: Because people need time to think, ensure you are reaching those who have already visited your website or engaged with your ads. This is often where the actual “buying” happens for high-commitment products.

• Monitor “Frequency”: In a local market like Montreal, keep an eye on your Frequency metric. If it gets too high (e.g., above 4 or 5), it means the same people are seeing your ad too many times, and it might be time to refresh your creative or slightly lower the daily spend.

Once you have a moment to check your previous Cost Per Result in Ads Manager, I can help you refine these numbers exactly. Would you like to know how to find that specific metric for your January campaign?