March 8, 2019
Email on Tap Episode 2, featuring Victoria Peppiatt, co-founder and COO of Phrasee.
250ok’s Director of Product Marketing Anthony Chiulli was a busy guy at eTail West in Palm Springs! After his chat with Jen Capstraw from Iterable, he found time to meet with Victoria Peppiatt, co-founder and COO of Phrasee for another edition of Email on Tap.
In this episode, Anthony and Victoria discuss how brand personality, tone, and AI can come together to create more powerful email, plus some thoughts on email marketing as a whole.
Check out the video below for the full conversation.
(Keep scrolling for key timestamps and even a full transcript. Plus, find links to our podcast version!)
Total Run Time: 8 minutes
00:20 – Story behind how Phrasee started and its business model
2:35 – How Phrasee generates content and subject lines to match a brand’s unique tone
3:53 – Phrasee’s approach to using AI and machine learning in the right way
4:27 – The two forms of AI: General AI and Narrow AI
6:22 – How Victoria’s artistic background in music and art influenced her role at Phrasee
Listen and subscribe on your favorite platform:
0:00 Hi and welcome back to another episode of Email on Tap. I’m your host, Anthony Chiulli. We’re here at eTail West in Palm Springs, California, and today my guess is Vic Peppiatt, cofounder and COO of Phrasee. Vic, thanks so much for being on the program.
0:17 Thank you for inviting me.
0:18 So Phrasee burst onto the AI scene, and recently ranked as one of the top eight AI startups in the UK. Tell me how this idea started out as a business.
0:27 So we founded Phrasee four years ago, February 2015. And it was myself, Parry Malm, and Dr. Neil Yager. It’s actually the brainchild of Parry, our CEO. He’s been working in the email marketing world, and had been working for one of the main ESP. One of the biggest questions he’s constantly got by all of his clients was, What’s the perfect subject line? Is there a science to it? Should it be long, should it be shot, should it have an emoji? What’s the sentiment? Should include the offer, or the discount, or free delivery? And it was the one question that he always said, you need to test. You need to test it. But the issue with that is, that as humans, we’re unable to quantify why one piece of short-form language may outperform another, and as human beings we relate to language very, very differently to from person to person. It’s very, very hard to be able to quantify and justify why some language outperforms another. So he came up with the idea of, what if we could build a machine that took away personal bias, and instead, we could actually base it on data, real live data from split testing. That’s where you bring in Dr. Neil Yager. Neil and Parry studied at university 20-25 years ago and they majored in computer science. They got back in touch about five years ago and Parry was talking about it through with Neil, and Neil said, “Well I could definitely help you.” He’s an expert, he’s got a PHD in machine vision, and he’s our expert in AI. At this point, they both went, Right, we could build a beta product. We’ve got a great idea, but we need help in setting up the business and getting it going, and that’s where I came in. So I’ve been working with Parry previous to Phrasee. I ran a creative agency for 11 years, and Parry had been one of our clients so, I’ve known him. He came to me and said, “Vic, I’ve got this amazing idea,” and told me about it, and I said, “Wow that is amazing. I want in.” That’s how we founded it.
2:34 So brands have unique voice and tone. Talk to me about how the machine learning and phrasing matches and personalizes subject line suggestions based on your brand’s unique tone.
2:45 Yeah. So it’s a really important point and actually coming from a creative and branding background, I really understood about the importance of making sure that you’re generating language at scale. It needs to sound like the brand. So we work with global brands such as eBay, Domino’s, Virgin, and everyone’s got a very distinct tone of voice, and it’s really really important. But if you’re using AI, it needs to be indistinguishable to that written by a human. So, when our clients start working with us, we want to make sure that Phrasee is generating language as if they were a copywriter in your marketing department. And the way that we do that is that we, when we onboard new clients, we take on the brand guidelines, that tone of voice, and we have a team of professional linguists, language technicians, who then build these bespoke language algorithms on a brand-to-brand basis. So at the touch of a button, you can literally generate thousands of variants, basically saying the same thing, but it sounds like someone from your marketing department and that’s a really unique thing we can do.
3:54 There’s a lot of talk about AI and machine learning here at this conference, and it’s certainly a buzzword and trend this year as well. But I think a lot of marketers get a little bit scared about marketing with machine learning and AI. How does Phrasee aim to disrupt this in the right way, and the morals and ethics behind machine learning?
4:15 This is a really key point for us. There’s a lot of scare-mongering going on about AI at the moment. And I think that’s an educational piece; really understanding the power the AI. So artificial intelligence is actually this umbrella term, and it can be lots of things depending on whether you’re a marketer or a mathematician or a scientist. And, you know, we’ve got two forms of AI. We’ve got what we call a general AI, which is where it’s the Terminator coming to replace you, take your jobs, take over the world, and they’re gonna just, you know, yeah, take us. Yeah, right. Which is quite terrifying, sure, but actually, in reality, we’re quite far away from reality. We’ve got what we call narrow AI, which is what Phrasee does. And that’s where it’s really, really powerful, when you’ve got a specific problem, that you’re using AI to solve that problem. So where we bring AI in, it’s about increasing email open rates, so we always are measuring against the human control, what you would normally be doing, and we can then see the actual power of using AI to generate language into generating optimized language, which is really key, is where then we can see the uplift. We then work with brands and copywriters and we onboard them to work together. That’s really, really important; copywriters to not think that we’re coming to replace them. They’ve got to think we’re working well with them. As email marketers, we spend so much time thinking about the content, designing an email, building it, making sure it’s responsive, and actually, when it comes to sending it, sometimes the email subject line is the last thing that everyone thinks about.
5:55 But it’s also the first impression.
5:58 And it’s the one thing that you’ve got full control over, because your brand name will always stay the same. And so why wouldn’t you put your best foot forward and use optimized language to get the highest open rate? And how we position it with copywriters is, we are helping your audience read your content. So we’re trying to get the highest level of engagement so that your customers can read the juicy bit of content that your copywriters are creating.
6:54 One last question that I have for you is, you studied music at your university and you play piano. And I believe flute, correct? How has that artistic background in music helped you launch Phrasee?
7:08 So I think for me, studying music and technology at university means that I’m probably a little bit braver going into tech while most others would fail, because I’ve never felt that there is a boundary or barrier between technology. And I think the creative aspect is fundamentally, you know, the short-form copywriter. So we’re empowering brands throughout the world with AI-powered copywriting. So as a creative element, it’s understanding that brand side that we discussed, about the tone of voice, and I think that creative element has given me a better understanding of what’s required and what our customers want, you know, that it can’t be just generic language, a one-size fits all.
7:53 Vic, thank you so much. This has been very insightful. I appreciate you taking time to be a guest. Thanks, everyone for watching. And we hope to see you next time on another episode of Email on Tap.
You may also like...
It’s hard to believe we are nearing the end of yet another exciting year in email, and 2019 proved to be one of the most momentous and active years to date. Over the past year, the number of new technologies, mergers and acquisitions, mailbox provider (MBP) announcements, news, and highlights is evidence of the versatility […]
I recently gave a presentation on global privacy regulations to a post-graduate marketing class and one of the things I noticed while preparing was that even within a single country, privacy is complicated. On a global scale, it is really complicated. For example, Canada has one federal private sector privacy law, three similar provincial laws, […]
*Update: This article was featured on email influencer Jordie van Rijn’s emailmonday blog! To see it in action, plus a great round-up of other articles and thought leadership looking forward to the future of email, click here.* The Black Friday emails are deleted, marketers’ email lists are checked twice, we pretty much know which senders […]