By Tom Wilson

May 22, 2023

AI-powered marketing firm Phrasee talks of creative impersonalization; their response to the challenges that brands and marketing agents face in accessing reliable third-party data. Jasper Pye, VP of Product at the firm, speaks to diginomica about how they help their clients mitigate the perils of unreliable data and deliver targeted - yet generalized - campaigns. Phrasee also explains how they seek to utilize generative AI powerhouse Chat GPT to diversify their offerings in the future. 

Creative Impersonalization and Data

To begin, we wanted to understand exactly what the team at Phrasee means by ‘creative impersonalization, and how they are using AI to deliver an alternative marketing strategy to their clients. Jasper Pye, VP of Product at Phrasee, says: :

The concept comes from this notion that individual brands don't necessarily, or aren't able to, accurately track the amount of data that will be required to effectively personalize to a user. 

It's better to hedge your bets, and is also more interesting as a marketer, to come up with the clever hooks that we were used to years and years ago when it came to marketing. So, sending out less personalized, more generic campaigns, which are designed to get across your brand message - but at a global scale across your audience

Personalization driven by the analysis of vast swathes of data holds the power to diversify user experiences and widen the scope of marketing campaigns. However, the drive behind so-called creative impersonalization is the acceptance that marketing agencies and brands do not have access to the amount of data that tech giants such as Apple and Meta command. On the difficulty that brands face in accessing customer data, Pye says:

It's hard to get data right, and it's hard to have clean data. Especially with things like Apple's privacy controls, they're making it ever, ever harder for brands to be able to track how people behave across communications. So if you go to a CRM or a marketing conference, what you'll find is a lot of what the brands are talking about is zero party data.

In the face of growing press and government attention on privacy and data security, brands are turning to new solutions to move with the times in their marketing campaigns. ‘Zero party data’, data freely and deliberately shared with a brand, is posited as the solution to the need to develop diverse campaigns without access to valuable gathered data that is usually controlled and sold by major tech corporations.

Phrasee’s customers

This outlook manifests as an AI-powered marketing strategy that generates various permutations of the same material and split-tests it across their clients’ customers. Jasper Pye explains:

You use our platform to generate a series of short-form messages to test out on an audience. So, they might take the form of subject lines or headlines or push notifications, basically any short-form content that enables you to AB test, and then what our platform will do is automate the process of testing that content on your audience.

For their customers, Pye was enthusiastic about their ability to track in real-time the ROI for each campaign. He adds:

Since the platform enables automated optimization and we use a control in all of our experiments, the customer is continuously able to see the uplift that is generated by Phrasee and translated into real ROI. The platform provides performance reporting that tracks this performance and presents it back to the user. In general, our customer stories always focus on this demonstrable ROI as a key point about why they like and continue to use the tool.

Using AI to bolster the more traditional practice of demography in marketing is perhaps a refreshing move to maintain the long-developed skill sets that have dominated the industry historically; understanding audience trends on a human level.

The future for Phrasee

Creative Impersonalization is how Phrasee operates today. However, Pye expressed a keen intention to embrace the concepts of optimized personalization in the future. Moving forward, the company hopes to use rapidly evolving AI tools, such as ChatGPT, to generate more permutations and automate this testing process further. Large language modules, according to Pye, hold the potential to massively increase the capacity to test marketing content. On where Phrasee wants to take their products next, he says: 

The idea is that historically we've optimized for the whole audience, but we recognize that our customers, in the pursuit of personalization, are getting a smaller and smaller, and more targeted audience. And it means that optimization doesn't make sense. 

So we're building a feature which is going to do the opposite of that. So rather than having 10 headlines or subject lines that you're looking to find the best one for the audience, we would generate potentially 1000s of outcomes, and then we would serve the best outcome for an individual based on what we've tracked about that individual. 

When prodded on where the likes of ChatGPT will take the company, Pye makes it clear that their proudly-held belief in generalized marketing may well subside in favour of further personalization, especially as rapid feedback and content generation become easier and easier with the latest AI offerings.

My take

Like most questions lately, this boils down to AI and its recent rapid development. In its nascent stages, generative AI perhaps provided a novel way to garner insightful analysis in a fraction of the time. In this sense, creative impersonalization is a welcome departure from incessant personalization that often misses the mark and lands the brand in muddy waters with its customers. However, like with most use cases of the technology, its exponential evolution leaves us at a tricky crossroads. Where before it might have underpinned understanding-based marketing strategies, now it may well be the catalyst for the ultimate personalized experience. I would argue, however, that the next clever hook that launches a global brand will first be thought, rather than generated.