MarTech Interview with Parry Malm, CEO at Phrasee

Hi Parry, please tell us about your journey in the technology. What inspired you to start Phrasee?

I like to think I was predestined to build something that combined language and technology: my mother was a language teacher, and my father was a physicist-turned-inventor who took an AI product to market. 

As for Phrasee: the idea originated from my hundreds of interactions with email marketers when I worked for an ESP. Whenever they asked me their #1 question – “what makes a good subject line?” – I’d always say, “try out a bunch of stuff and see what works”. It always felt like a cop out. I knew there had to be a better way – a more scientific way – and sure enough, there was. 

From brand-side marketer to leading an AI company as savvy as Phrasee – what inspired you to make the leap into the industry?

It dawned on me that marketers had traditionally relied upon their intuition and gut instinct when it came to choosing the best language to use. The marketing of the future is data-driven. Standing out from the crowd and keeping audiences consistently engaged across all channels, all while staying on brand, can no longer be left to the Don Drapers of the world. 

So we decided the market was ready to adopt advanced technology to optimize the language in marketing campaigns.

How does adding an AI element to your martech platform actually helps convince your customers and justify their faith in your platform?

In actual fact we have always been a parallelized AI/MarTech company – putting science behind language has always been our MO, and that science component has always been AI.

Natural language generation (NLG) is the AI that generates human language tailored to a brand’s style, tone, target audience, and campaign goals. Deep learning is how we leverage historical data and the results of automated experimentation. It all adds up to more optimized messaging and better marketing performance for our customers. We justify their faith with the numbers that we produce!

Could you tell us more about Phrasee’s offerings for the ecommerce industry and the unique features you currently provide to the Martech users within e-commerce?

It sounds fancy (because it is), but the math is simple. Your brand, plus Phrasee’s real-time language optimization, equals increased engagement across the entire customer journey. Phrasee generates high-performing brand language made personal and relevant to your target audience, in real-time and at scale. It helps you deliver the right message, at the right time, to the right people, in your unique brand voice.

What is Phrasee Score? What kind of data science engine drives Phrasee Score’s performance?

Phrasee Score confronts marketing’s dirty secret: that no existing individual email KPI is perfect. For instance, clicks and opens are often clouded by bots, click fraud and accidental or random opens. So rather than relying on these isolated metrics, Phrasee Score uses ensemble learning to aggregate multiple data points, producing one score that more accurately and reliably represents the overall impact of your email. 

Holiday Shopping Trends

During the COVID-19 pandemic, we learned how CMOs led their marketing teams through the accelerated need to crossover into Digital Shopping trends. Have you analyzed this unique trend in digital shopping behaviors that has helped or might help your customers meet the current demands of retail shopping?

Most data shows that the majority of customers still prefer in-store shopping – even the younger generations – but the blending of physical and digital shopping is accelerating rapidly. There’s now a need to balance continued digital innovation with a return to brick- and-mortar shopping. 

We’re seeing much more in-store sales impact from online marketing, so providing in-store experiences that pair neatly with the digital world could help differentiate your brand and increase customer loyalty through a better shopping experience. For example – imagine having a QR code next to the price tag that guests could scan to see reviews about a product before they purchase, or using that same QR code to access stock alerts. This is how retailers need to be thinking.

With the holiday shopping season just weeks away, what should be the action plan for retailers? How can retailers leverage Phrasee to drive increased CLTV?

Inboxes are of course extra-crowded during the peak retail season, so it’s key that you optimize your brand language enough to stand out. Our analysis of thousands of subject lines from last winter’s retail peak revealed that clear, concise, and direct language tended to perform better. Empathy and sensitivity were also key – as were emojis. That said, each brand is unique – what’s most important is taking what makes your business special and optimizing how you communicate that to your customers. Phrasee does this by distilling your brand’s style, tone, target audience, and campaign goals and generating optimized language that you can scale across your communications channels for maximum engagement.

AIOps, IT, Security and DevOps in Martech

You are highly acclaimed as an AI SaaS company for Digital Advertising and AI domains? Could you tell us how these two distinguished technology domains actually converge at Phrasee? We are very curious about learning more about the AI lab and how it works with the wider Product Development and Marketing team at Phrasee.

The process we go through with each customer combines the power of AI and great digital marketing. First, we build a language model tailored to a brand’s style, tone, target audience, and campaign goals. Phrasee can then produce consistent, natural-sounding and scandal-proof language, in real time, with the click of a button. Deep learning is how we then leverage historical data and the results of automated experimentation so that Phrasee is always learning from past performance and getting better with every send.

Hear it from the AI pro: Do you agree that Data Privacy is a huge concern for marketers? Tell us how AI-powered metric could help email marketers overcome challenges with Apple’s upcoming Mail Privacy Protection feature.

Yes, data privacy concerns are introducing new restrictions (and statistical noise) for email marketers – but contrary to some headlines, the sky isn’t falling in. Nothing has fundamentally changed about what makes an email work: you still need a killer subject line, a clean list, a good layout, a compelling offer, strong CTAs. What’s changed is that we all need to get smarter in how we track and optimize email performance. 

That’s why we created Phrasee Score – it solves the problem of noisy KPIs by combining information from multiple sources to give our customers one single, weighted score they can trust. This is how we equalize the extra noise introduced by Mail Privacy Protection.

IT support and customization of services are keys to adoption of any new technology, especially if it involves catering to a huge industry as huge as e-commerce and digital advertising. What kind of IT support do you offer to customers to ensure better customer experience (CX) and security on your platforms?

Very little IT customization is required to implement Phrasee – we have off-the-shelf plugins with all major marketing automation and CX systems, from Salesforce to Adobe.

Parry, this is my favorite part of the interview: Tell us about your AI Ethics policy. Do you think every CEO of a global tech company should have a very clear AI Ethics policy?

We’ve always believed AI should be used for good. Since we founded Phrasee in 2015, our focus on ethics has remained steadfast. Our AI ethics policy was one of the first of its kind but it’s also an ever-evolving document given the world doesn’t stand still. It’s important that our ethics policy is practical, specific and explicit. It’s publicly available on our website and covers everything from us not using data to target vulnerable populations to being willing to take action to avoid prejudice and bias, and always being open about what our AI does.

As for my views on other CEOs and how they should act, I think it would be grandiose for me to say what I think my peers should be doing. What I can say is that for Phrasee, our AI Ethics policy has been hugely impactful – both from a business focus and a cultural standpoint. Our team have all read the policy and know it, and trust in our aim as a company to adhere to it.

Any tips or advice for the young professionals looking to join or are already working in the Martech or AI Data Science industries in the coming months:

Don’t focus on the industry, focus on the people you’re working with, and the culture you’re entering. You can make money in a lot of ways these days – so why not do it with nice people, while having fun? 

Your plans for the upcoming holiday shopping season:

Buy as little as possible, because I’m not super into material things. Though I may get my sons some remote-control cars. 

What are you currently reading / any book that you would recommend to our readers:

The best book about business I’ve ever read is hundreds of years old. It’s about scarcity, and how scarcity drives humans to behave in interesting ways. It’s called Candide, by Voltaire – and I would recommend it as required reading for anyone in the business world, or for that matter, the world.

Which webinars / podcasts are you subscribed to: how do these help you stay on top of your goals

I am not a big consumer of business-focused media. The only podcast I listen to with any regularity is Business Wars, a podcast by Wondery, which pits two all-time competitors in a market sector and outlines the history of their competitive struggles. My favorite miniseries of theirs was the one on the Monday Night Wars – WWE vs WCW wrestling. Highly recommend! 

Thanks Parry, for answering all our questions!

Marketing Technology News: MarTech Interview with Max Nelson, VP of Marketing Communications at OpenX