AI prompts to increase the ROAS on performance marketing by 5x

New Delhi: In a session on Thursday at the India Affiliate Summit (IAS) 2024, Marisha Lakhiani, Chief Growth Officer of Mindvalley, shared insights on using ChatGPT prompts to optimise ad efficiency and significantly increase Return On Ad Spend (ROAS).

Highlighting strategies that led to Mindvalley’s 5X revenue growth during Black Friday from 2022 to 2023, Lakhiani outlined how AI empowers marketers to spend less while generating higher returns. 

Meta Advantage+ Shopping Campaigns: A game changer

At the heart of Lakhiani's strategy is Meta Advantage+ Shopping Campaigns, an AI-driven tool that's revolutionising performance marketing. "This is a meta-AI model that alone will 2X your return on ad spend. And what it does is it dynamically targets the right person with the right placement, with the right copy, and the right creative," Lakhiani explained.

To add, “It tests up to 150 different combinations so you don’t have to worry about testing. The automation extends even to audience expansion, allowing AI to identify and tap into new audiences that share similar characteristics with existing customers, potentially converting new leads more effectively,” Lakhiani quoted.

The key features of Meta Advantage+ Shopping Campaigns include:

Dynamic Targeting: The AI dynamically targets relevant audiences, ensuring that ads reach the right people at the right time.

Creative Combinations: It automatically tests up to 150 different creative combinations from all the images, videos, and copy you upload. This ensures that the most effective combinations are identified and utilized.

Audience Expansion: The tool automates audience expansion to find new potential customers, broadening the reach of campaigns.

Increasing ROAS

Mindvalley’s successful Black Friday campaigns were the result of using AI to improve Return on Ad Spend (ROAS). Sharing her methods Lakhiani said, “We combined all videos, images, and copy into one campaign and then blended high-performing and underperforming assets to create a campaign with a diverse range of content, optimised to serve different audience segments at different times.

We took videos, images, and copy, and we took five of our top-performing assets and five of our worst-performing assets because to make ASC work, you need diversity. You need different kinds of assets, different kinds of images, different kinds of copy. The more diverse, the better."

The blend of non-performing and performing assets allowed the AI to detect shifts in audience preference throughout the day. The non-performing assets, previously dismissed, engaged well with different segments, showing that their effectiveness depended on timing and targeting.

When Mindvalley found that generic creatives were not performing, they pivoted to product-driven creatives using AI. This shift resulted in an 800% increase in engagement rates and lowered Cost Per Mille (CPM), increasing their reach. "All ads were created insanely fast, in a matter of hours," Lakhiani added.

Real-Time data analysis with Google Apps Scripts

Another critical aspect of Mindvalley's strategy was the use of real-time data. Lakhiani commented on how it pulled data from ad platforms and payment gateways using Google Scripts and plugged it into Google Sheets in real-time. This approach provided a significant advantage, allowing the team to make data-driven decisions swiftly.

Sharing how she did it, Lakhiani said, "You pull data from ad platforms and plug it into Google Ad Script. Once you have it in a Google Ad Script, it plugs into Sheets. In Sheets, your Sheet basically turns into a dashboard where all of your transactions are being logged in real time. 

When we started using real-time data and stopped using dashboards and started using Sheets, we actually went from 100K on one campaign to 800K on the same campaign within three weeks," Lakhiani shared.

Smart Scheduling: Maximising budget efficiency

Another key strategy involved Smart Scheduling, where budget spending was concentrated during periods with peak conversion rates and low CPMs. Lakhiani explained how AI tools helped identify peak times for conversions, allowing them to focus ad spending when it would yield the highest returns. 

By analysing real-time data, Mindvalley identified the peak sales hours and adjusted their ad spending accordingly. 

"We found out the best times to spend our budget was between 11 pm and 3 am. Meta was taking our budget and spending it equally throughout the day, but actually, our sales are only happening at certain hours, so the rest of the time spent in my budget is actually wasted. Just by scheduling ad spends only during these peak hours the increase in ROAS was 400%,” said Lakhiani.

"When we tried Smart Scheduling, we literally went from 114% ROAS to 500% ROAS,” she further added.

Reducing CPMs with AI-driven comment management

To combat rising CPMs, Mindvalley employed an AI-driven tool to manage comments on their ads. This tool responds to comments, hides negative ones, and addresses customer queries, increasing engagement and reducing CPMs by up to 74%.

"When we started to use this tactic of AI comments using this tool called BrandFastian, it actually was able to drop our CPMs down to $16, which is crazy," shared Lakhiani.

Lakhiani's session highlighted the strength of AI in performance marketing. By leveraging tools like Meta Advantage+ Shopping Campaigns, real-time data analysis, and smart scheduling, marketers can significantly enhance ad efficiency and ROAS.

From the horse’s lips itself, Lakhiani said, "It's no longer acceptable now to get a 20% better. We are now expecting a 300, 400% better than we were doing before."