If you attended Opticon '23, you saw first-hand how Optimizely has been investing in AI. Optimizely introduced Opal, an AI assistant designed to accelerate the entire marketing lifecycle. Opal is ever-present across Optimizely One, providing generative AI, smart insights, and recommendations to transform how our customers create, test and personalize digital experiences.
Now, our latest AI capability is here: copy generator. Available for all Web Experimentation customers, copy generator helps experiment authors expedite the ideation and creation of test variations.
What does it do?
copy generator leverages generative artificial intelligence to create a list of phrasing suggestions based on a site’s text elements like headlines, product descriptions, or call-to-action (CTA) wording, ultimately making it easier and faster for experimenters to plan multiple variations for their tests, which can be quite time-consuming.
Who is it for?
Based on our research, around 30% of experiments include text changes. So, experiment authors like optimization managers or digital marketers are spending a lot of time ideating/brainstorming multiple versions of the original copy to decide which should be tested. copy generator empowers users to add more variations in an experiment, which we strongly suggest after our Experimentation Benchmark research found that experiments with more variations (4+) tend to see higher win rates and return higher uplifts on the metrics tracked.
Cool...but generative AI is popping up everywhere, why does it matter here?
- Directly embedded into our UI: No separate tools or tabs to click out to...No typing out a prompt to a chatbot…just click the text element you want suggestions for, and click "generate." All interaction stays within our Visual Editor.
- Reduce time and effort in variation ideation: Shorten the time it takes to come up with new experiment variations, allowing experiment authors to get more time back into their day.
- Optimize each variation in an experiment: copy generator provides unbiased and creative alternatives to experiment authors so they can make sure that each variation is different enough to avoid duplicative messaging, yet effective enough captures visitors' attention.
- Increase a test's chances of winning: Our Benchmark research shows that experiments with 4+ variations are ~90% more likely to win than experiments with just 2 variations. copy generator helps experiment authors create more variations, leading to higher lifts.
- Fine-tune brand positioning: Improve existing headlines, product descriptions, CTA buttons, and more, ensuring a consistent and impactful brand message across digital properties.
Increase a test's chances of winning
This outcome is important enough to highlight a second time. Mentioned earlier, we know from our Experimentation Benchmark research that tests with more variations (4+) are more likely to produce a winning (statistically significant) result versus a traditional A/B test that pits a baseline (original version) against a single variation. copy generator can help experiment authors get into the habit of testing more variations and producing more winning results.
Optimizely is committed to continuous innovation and improvement. Potential enhancements for copy generator include generating suggestions for other content types like images, icons, HTML, and CSS, as well as giving users more control over output fine-tuning, such as adjusting length, tone, and other fields.
At the end of the day…
Optimizely's copy generator is a simple yet powerful feature that empowers experiment authors to create more effective and winning experiments. By harnessing the power of artificial intelligence, this feature saves time, optimizes variations, and fine-tunes brand positioning, ultimately leading to better results, stronger brand presence, and an effortless workflow.
Want more info? If you're an existing customer, ask your account manager about copy generator, and if you're a future customer, contact us to learn more.