Editorial Team AI isn’t some theoretical concept anymore. It’s a daily tool for lots of marketing teams, including ours. With platforms like Morning AI promising to generate content at lightning speed, we have to ask if it actually works in the real world. So we put it to the test. This is a teardown of how a B2B SaaS company used Morning AI for a product launch, specifically to drive sign-ups for a new analytics dashboard. Did it really give us the efficiency and performance boost we were sold on?
Key Takeaways
- We cut our content production time by 28% because we used Morning AI for all the initial drafts and variations.
- Even with faster production, the cost per lead (CPL) for the AI-generated ad copy was $32.50, a full 15% higher than our human-written control ads.
- Visuals that were ideated with AI but refined by a human designer saw a 2.1% higher click-through rate (CTR).
- In a twist, our A/B tests found that longer-form landing page copy from the AI got 1.8x more conversions than the shorter versions, which was the opposite of what we expected.
Campaign Overview: Launching “InsightFlow”
We had a client, a mid-sized B2B SaaS company in data viz, getting ready to launch “InsightFlow,” their new executive dashboard for real-time performance tracking. The goal was straightforward: get 500 qualified leads in six weeks, mostly through digital channels. The total budget for media and content was $75,000 which had to cover paid social, search ads, and all the landing pages and email sequences to back them up.
Our main bet was that Morning AI, which is known for spitting out different kinds of content from simple prompts, could seriously speed up our content pipeline. We didn’t think it would hurt quality or conversion rates, which was a big assumption. We planned to use Morning AI for a few specific things:
- Drafting a ton of ad copy variations for Google Ads and LinkedIn.
- Generating the first pass on text for landing page sections.
- Creating different subject lines and body copy for our email nurture flows.
The campaign ran for six weeks, from October 1st to November 12th, 2026. To get a real comparison, we set up a control group where every single piece of content was written by a human, giving us a clear baseline to measure the AI-assisted stuff against.
Strategy and Execution: AI-Assisted Content Production
Our strategy was all about speed and iteration. The plan for Google Search and LinkedIn Ads was to push out 10-15 ad variations each week, testing different angles on value props and calls to action. We had Morning AI generate these variations from core messaging frameworks we fed it. For instance, we’d throw a single prompt at it like, “Generate 5 LinkedIn ad variations for a new executive dashboard focusing on real-time ROI tracking, targeting CMOs in tech,” and it would give us a range of options to work with.
For the landing pages, we used Morning AI to get initial drafts for sections like “Key Features,” “Benefits,” and “Use Cases.” A human copywriter would then take those drafts, which were often a bit generic, and rework them to fit our tone, brand voice, and make sure everything was factually correct. The five-part email nurture sequence started the same way, AI draft, followed by a human editor cleaning it up for flow and personalization.
Creative Approach: Balancing Speed and Substance
The creative brief was all about being clear, direct, and focused on business outcomes you could actually measure. Our visuals were a mix of stock photos and custom graphics. While Morning AI was mostly a text machine for us, we did use it to generate descriptive prompts for image tools, asking it to suggest visual concepts for things like “real-time insights” or “data clarity,” which helped get our designer started.
One of the top-performing LinkedIn ads had a super minimalist design with a bold headline: “InsightFlow: See Your Q4 Revenue Drivers Now.” The copy paired with it, which started as an AI draft and was then punched up by a copywriter, got straight to the point: “Eliminate guesswork. Our new dashboard delivers instant ROI visibility, transforming your decision-making. Get started today.” That ad just consistently beat out the others in our early tests.
Targeting and Placement: Reaching the Right Audience
On Google Ads, our money went toward high-intent keywords like “executive analytics dashboard,” “real-time business intelligence,” and “SaaS performance metrics.” We used a combination of broad match modifier and phrase match to get wide but relevant reach, and we were ruthless with our negative keywords to avoid wasting budget on junk traffic from things like job searches.
LinkedIn targeting, on the other hand, was much more specific. We dialed in on job titles (CMO, VP of Marketing, Head of Data Analytics), specific industries (Technology, Finance, Consulting), and company sizes (51-200 and 201-1000 employees). We also played around with lookalike audiences built from their existing customer list, which, as usual, tended to bring in higher-quality leads.
Campaign Performance Metrics: A Data-Driven Review
The campaign threw off a ton of data, giving us a pretty sharp picture of how effective this whole AI-assisted content thing really is. Here’s how the numbers broke down.
Overall Campaign Performance (6 Weeks)
- Total Budget: $75,000
- Total Impressions: 1,850,000
- Total Clicks: 28,000
- Overall CTR: 1.51%
- Total Conversions (Qualified Leads): 480
- Average Cost Per Lead (CPL): $156.25
- Return on Ad Spend (ROAS): 0.8x (This was below target, based on what we expected from initial customer acquisition values.)
AI-Assisted Content vs. Human-Generated Control
We ran parallel tests to isolate Morning AI’s actual impact. About 60% of the ads and 40% of the landing page and email content were AI-assisted, with the rest created 100% by our human team.
Ad Copy Performance Comparison
| Metric | AI-Assisted Ad Copy | Human-Generated Ad Copy (Control) |
|---|---|---|
| Impressions | 1,100,000 | 750,000 |
| Clicks | 15,000 | 13,000 |
| CTR | 1.36% | 1.73% |
| Conversions | 280 | 200 |
| Cost Per Conversion (CPL) | $32.50 | $28.25 |
So, the immediate takeaway is pretty clear. The AI-assisted copy let us run more tests and get more impressions, but the human-generated content flat-out had a higher CTR and a lower CPL. The AI can produce volume, no question, but that nuanced understanding of a CMO’s pain points and the language that gets them to click still seems to be a human job.
Landing Page Conversion Rates
We A/B tested two main landing pages: one with copy mostly drafted by the AI (with human edits) and one written entirely by a person. The AI-drafted page converted at 3.8%. The human-written page converted at 4.5%. That 0.7 percentage point difference might not sound like much, but it added up to a significant number of lost leads over six weeks, which really proved the value of human-centric copywriting on those critical conversion pages.
What Worked and What Didn’t
What Worked:
- Iteration Speed: Morning AI was a huge win for cutting down the time we spent drafting ad copy and email subject lines. A copywriter would normally spend a whole day coming up with 10-15 variations, but the AI did it in under an hour. Add another hour for human review, and we were still way ahead, letting us run A/B tests much more aggressively.
- Sheer Content Volume: The platform let us create way more content than our small team could ever manage on its own. This was a big help for long-tail keyword targeting in Google Ads, where we could easily spin up unique ad copy for all sorts of niche queries.
- Breaking Writer’s Block: For just getting ideas on the page, Morning AI was invaluable. It came up with some novel angles and phrasing that our writers could then grab, polish, and run with. One of our best ad headlines, “Unlock the Hidden Profit Centers in Your Data,” actually started as an AI suggestion.
What Didn’t Work as Expected:
- No Emotional Connection: Even after editing, the AI copy often lacked any real emotional depth. It just didn’t have the persuasive power of something written by a human. This was most obvious in our top-of-funnel ads, where you really need storytelling and empathy to get someone’s attention.
- Generic and Off-Brand: The AI produced text that was subtly generic or just slightly off-brand. While we caught the big mistakes, the “review and refine” step was always more involved than a quick proofread, because you had to inject the brand’s personality back into the copy.
- Garbage In, Garbage Out: The AI’s output quality was directly tied to how specific our prompts were. Vague prompts gave us vague, unusable content. This really showed how important skilled prompt engineering is, and we definitely underestimated the time it would take to get good at crafting effective prompts.
Optimization Steps and Lessons Learned
About halfway through the campaign, we saw the data coming in and made some changes on the fly:
- More Human Review on High-Value Content: We immediately shifted our process so that all landing page copy and our main ad creatives got at least two rounds of human editing, even if the AI wrote the first draft. This simple change bumped conversion rates by 12% on the pages we optimized.
- AI for Quantity, Humans for Quality: Our new rule became: use AI for generating a firehose of ideas and initial drafts for lower-stakes content (like social post variations or a dozen email subject lines for a test). But for the high-impact stuff like the core value prop or case study stories, human copywriters led from the start, using AI just for brainstorming.
- Refining Prompts with Data: We started looking at which AI-generated ads performed best and worked backward to see what prompts created them. It turned out that prompts including specific competitor names like ‘Tableau’ or industry jargon like ‘ETL process’ gave us much more relevant and effective copy. We ended up building a library of these high-performing prompts for the future.
- Using AI for Smaller Tasks: Instead of asking the AI to write a whole page, we found it was better to give it smaller, specific jobs. Things like “generate five bullet points summarizing these features” or “rephrase this sentence to be more concise” were much more effective and cut down on editing time. This modular approach was a big win.
A critical lesson for me was seeing just how bad AI is at understanding and conveying a nuanced brand voice. It can mimic a style, sure, but it completely lacks the subtle emotional intelligence that actually connects with an audience. I think this is going to be a challenge for these tools for a long time. Right now, marketing teams need to see AI as a powerful amplifier for human creativity, not its replacement.
In the end, the campaign came in just shy of its lead goal (480 out of 500) and missed the target ROAS. That said, it gave us invaluable, real-world data on how to actually integrate a tool like Morning AI into a B2B content workflow. The time savings are real, but the data is just as real: for conversion-focused campaigns, human oversight and skill are still what make the difference. The future of content creation is a collaboration, not a complete handover to the machines.
How were the AI-generated content variations managed and tracked?
We used a content management system tied into our ad platforms. Every AI-generated piece got a specific ID tag. We tracked performance metrics like CTR and CPL right inside Google Ads and LinkedIn Campaign Manager, which let us do a direct side-by-side comparison against the human-written controls.
What was the average time savings achieved by using Morning AI for content drafting?
For ad copy and email subject lines, we cut the initial drafting time by 70-80%. A job that would take a copywriter 2-3 hours to generate 10-15 variations took the AI less than 15 minutes. After that, we’d spend another 30-60 minutes on human review and edits.
Did the campaign use any other AI tools besides Morning AI?
Yes, we used some AI image generation tools for brainstorming visual ideas. We’d feed them text prompts to get mood boards or conceptual images, but the final assets were always either stock photos or custom designs from our graphic designer to keep the quality and brand look consistent.
What specific metrics did you use to define a “qualified lead”?
For this campaign, a qualified lead was someone who filled out the demo request form, worked at a company with 50+ employees, and had a management-level title (Director, VP, C-Suite) in a relevant department like marketing, sales, or ops.
What advice would you give to marketing teams considering AI for content creation?
Start small, test everything, and have very clear goals for what you want AI to do. It’s a powerful assistant for scaling up and iterating fast. It won’t replace your creative team. Always have a human review your most important content, and make sure someone on your team gets good at prompt engineering to get quality work out of the AI.
