At this stage, nearly everyone has heard of OpenAI’s ChatGPT platform. Thousands of articles have been written on the topic, predicting everything from mass layoffs as AI makes human creativity obsolete through to ChatGPT and other tools like it being nothing more than a brief flash in the pan.
While I don’t think millions of people will be out of a job next week, the emergence of readily accessible & high quality AI generated creative is likely to have a huge impact on society and working culture moving forward. These tools are here to stay, but for now there’s still a major role for humans to play in guiding and overseeing the automation process to ensure the output is as effective as possible, whatever the end goal is.
So how can digital marketers use ChatGPT and tools like it to take their work to the next level? By remembering that output can only be as good as the input and if you put garbage in, you get garbage out.
Let’s look at an example.
Here, we’ve asked ChatGPT to write us an ad on the Google Ads platform. The input is very top level, all we’ve told the tool is the character limits for the ad, that it’s a retail business and their main product is athletic shoes. So what did we get?
As ads go, it’s OK. We’re within the character limits (assuming two description lines for 180 characters total), it’s calling out athletic shoes and has a decent CTA. I think we can do better though.
What if we gave more information about our target audience? How would it change if we let ChatGPT know that these shoes are made for men in their 30’s who have a goal of running their first marathon?
Now we’re getting somewhere. We’ve identified our target customer’s main goal and are calling that out both in the headline & in the description line. There’s still some improvements that we can make, for example mentioning that they’re men’s shoes, but a shopper seeing this ad will feel that it’s much more tailored to their specific goals.
We’re still not done though, at any given time there’s plenty of people training for a marathon. What other information can we give to create an automated ad that will set us apart from the other competitors in the search results?
After talking to customers, we know that a lot of new runners struggle with injuries as they build up to the marathon distance and it’s a major concern for them. What if we include that in the input?
Compared to the first ad, the difference is night and day. With just a few additional pieces of information we’ve gone from simply saying that we have athletic shoes for sale all the way to calling out how these shoes will solve a specific problem for a customer looking to reach a specific goal. The system itself didn’t change, the only difference was the quantity and quality of the input that we gave in order to create the ad.
ChatGPT has done 90% of the work here and now a human can step in to provide the final touches. We’ve already discussed making tweaks to specify that they’re men’s shoes and we can also use our own contextual knowledge to choose the best URL to send the page to along with the most relevant related products to associate with the ad.
Even with something as basic as athletic shoes, the difference in quality between a low input & high input automated ad is huge. As product complexity grows along with the value of a successful sale, ensuring input is of the highest quality possible becomes absolutely crucial to success. Without a human working to guide the process, any automatically generated ad will be lost among dozens of others in the search results.
Understanding this helps us understand the future of digital marketing and the ongoing shift from tactical to strategic thinking. Why spend time manually writing ads, when you can spend time building out the best customer profiles possible? Leave the tedious work to the machines and use human intelligence & creativity to design the process.
For automated creative, there’s ChatGPT, for everything else there’s Scope Digital. Get in touch.