Running ads without proper testing is a bit like throwing darts blindfolded – many people make this mistake, but the biggest brands in the world consistently test their ads.
A/B testing is a crucial practice that helps you save time and money by revealing exactly what works and what doesn’t, allowing for optimisation without guesswork.
Doing this manually can be a huge hassle, which is where AdCreative.ai steps in. This platform helps you generate multiple versions of your ads and connects directly to your ad accounts to streamline the A/B testing process.
Here’s a step-by-step guide on how to A/B test ads using AdCreative.ai
Step 1: Create Your Account and Connect Ad Platforms
- First up, create an AdCreative.ai account, using a free trial link if available.
- Once logged in, navigate to “Platform Connections” from your initials in the top right corner.
- It’s absolutely crucial to connect your existing ad accounts, such as Meta Ads, Google Ads, LinkedIn, or Pinterest. This connection allows AdCreative.ai to pull in vital data like conversion information, which is essential for effective A/B testing and for the AI to learn from your past ad performance.
Step 2: Plug in Your Brand Details
- After connecting your ad accounts, head over to the “Brands” section.
- Here, you can input details about your business or brand, like your domain, which helps the platform find your brand.
- You’ll also have the option to select your brand logo and brand colours.
Step 3: Select Ad Creative Type and Define Your Goal
- The platform offers a range of ad creation options, from basic image ads and product photo ads to videos, ad text, and social creatives.
- Before proceeding, clearly define your A/B test goal: Are you aiming for link clicks, conversions, or likes? Your goal will guide your choice of ad type, such as post-size for social media, landscape, story, portrait, or display.
Step 4: Generate Ad Copy (Headline, Punchline, CTA)
- You can manually write your ad’s headline, punchline, or Call to Action (CTA).
- Alternatively, you can leverage the platform’s AI to write the ad text for you. This AI becomes smarter and more specific based on the performance data from your previously connected ad accounts.
- When using the AI, you can specify details like the product, desired tone (e.g., friendly), language, product description, audience (e.g., triathletes), and CTA (e.g., “shop now”).
- The platform will generate multiple versions of the text, and a key A/B testing tip is to test different headlines to see which resonates best.
Step 5: Select Your Visuals
- You have the flexibility to import your own photos or search for relevant images directly within the system. For instance, you could search for “triathlon swimsuit”.
- You can upload various images and videos to enable testing of different visual approaches.
Step 6: Generate Ad Versions and Review Convergence Scores
- Once you’ve provided the details, the platform will generate a multitude of different ad versions for you.
- Each ad will display a “convergence score” above it. This score is derived from extensive historical tests the platform has run, indicating how well similar ads have converted in the past.
- As you scroll down, the scores might decrease for ads that historically performed less effectively. It’s important to remember that every brand is unique, so what worked for others might not work for you, reinforcing the value of your own A/B testing.
Step 7: Publish Ads and Initiate A/B Tests
- From the platform, you can directly publish your selected ad versions onto your connected ad platforms to begin your A/B test. You also have the option to download them.
- Before publishing, you can fine-tune ads using a “quick edit tool” or a “creative Studio editor”. The Creative Studio editor offers a user-friendly, drag-and-drop interface for modifications, meaning you won’t need separate design software like Adobe or Figma.
- When starting out, it’s perfectly fine to A/B test broadly with completely different ads. However, as you gain insights, it’s recommended to focus on minor adjustments to a single ad (e.g., tweaking the description, visual, colour, or headline). This approach helps you pinpoint exactly which variable influences performance, rather than having too many uncontrolled variables.
Step 8: Monitor Performance and Optimise
- When running your A/B tests, allow them to run for a sufficient duration; avoid stopping them prematurely after just a few days. A minimum of seven days is suggested to gather enough comprehensive data, as people’s shopping habits vary significantly across different days of the week.
- Keep a close eye on key performance indicators (KPIs) like click-through rates (CTR), cost per click (CPC), Return on Ad Spend (ROAS), and the overall conversion rate.
- Recognise that different ads serve different purposes: ads that bring in a lot of traffic but fewer conversions can be valuable for brand awareness, while ads with high conversion rates should be prioritised for driving sales and further closes.
- The platform also offers predictive tools, including a “creative scoring AI” to forecast ad performance and a “compliance check AI” for regulatory peace of mind.
- Finally, continuously test your ads, as what works effectively today might not yield the same results in the future. A/B testing is not just a ‘nice to have’ but a fundamental “must” for creating effective, converting ads.