Marketers are really good at detecting marketing BS.
In the marketing field, AI is all the rage. There’s excitement around new AI capabilities. There’s heightened interest in AI-driven efficiency. And there’s an ongoing stream of new opportunities that are springing from AI innovation. That might lead even the best of us to get carried away in claims of AI “magic.”
At the same time, there’s also growing competition in the growing field of AI solutions and AI has become a buzzword, a catch-all term. That often leads to companies feeling pressure to wave the AI flag and dive immediately into the AI details. But that can spark more questions than enthusiasm.
This episode of FIVE focuses on the AI sales pitch to help marketers hone their intuition, find the legit powerful solutions, sniff out the gotchas… and answer the question, “What’s the catch?”
The FIVE List
Join us for episode 7 for the all the details behind the FIVE:
- If you’re selling solutions powered by real AI, then you should really try to avoid pinning your story on AI.
- If you’re a buyer, spend most of your time diagnosing your problem and far less on finding the solution.
- Sellers should sell performance first, not technology.
- Buyers and sellers should expect clarity and transparency from each other.
- If you’re thinking about adopting an AI-powered solution, it’s better to take a step and fail than not try at all.
Three impressive, insightful guests to explore best practices for vendors selling AI-powered solutions. And, best practices for marketers looking to implement AI-powered solutions.
Parry Malm, CEO @ Phrasee the most advanced AI-Powered Copywriting tech on the planet, a world-leading natural language generation system that writes marketing copy that sounds human – and fits your brand’s voice. It’s a deep learning engine that can predict what language will and won’t work better than any human. The tech extends across all digital channels, including email, push, paid social, paid search, display, and web, giving you a consistently high-performing brand voice everywhere. They quite literally wrote the book on this technology called “The Language Effect”.
Liz Miller, VP and Principal Analyst at Constellation Research, Inc. where they are passionate about how business models can be transformed by disruptive technology. Their goal is to help clients realize the Art of the Possible.
Don Fluckinger Senior News Writer, Search Customer Experience and Search Content Management at TechTarget. He covers CX management and its enabling technologies (CRM, service/support, marketing automation, sales automation, e-commerce, call center and digital experience) for SearchCustomerExperience. A music diehard and vinyl blogger, Don cranks up the rock, jazz, blues or funk — and rips off the knob.
And of course, Jeremy Lockhorn, Global Head of Partner Solutions at Ericsson Emodo, speaker, mobile marketing expert.
The Five podcast is presented by Ericsson Emodo and the Emodo Institute, and features original music by Dyaphonic and the Small Town Symphonette. Social media and promotional content was composed and conducted by Lyon Solntsev. This episode was edited by Justin Newton and produced by Robert Haskitt, Liz Wynnemer, and Jake Moskowitz.
Transcript for AI E7: The AI Pitch and Catch
If you keep automating your bad processes, all you’re getting are really fast bad process, right? And so AI is never going to solve your bad process. It can accelerate it.
Let’s talk AI. Welcome to FIVE, the podcast that breaks down AI for marketers. This is episode seven, the AI pitch and catch. I’m Jake Moskowitz.
You know that this is a podcast for marketers. It’s in the title. It’s in the intro. It’s in the conversations we have in every episode. Well, if there’s one thing we marketers have in common, aside from, you know, marketing, is probably this, marketers are really good at detecting marketing BS. If they’re not buying what you’re selling, they either don’t need it, or they’re smelling the spin. But in the marketing field, AI is all the rage. There’s excitement around new AI capabilities. There’s heightened interest in AI driven efficiency. And there’s an ongoing stream of new opportunities that are springing from AI innovation. That might even lead the best of us to get carried away in claims of AI magic.
At the same time, there’s also growing competition in the growing field of AI solutions. And AI has become a buzzword, a catch all term. That often leads companies feeling pressure to wave the AI flag and dive immediately into the AI details. But that can spark more questions than enthusiasm.
This episode of FIVE focuses on the AI sales pitch to help marketers hone their intuition, find the legit powerful solutions, sniff out the gotchas and answer the question, what’s the catch? So let’s talk about what we’re talking about.
When marketers are looking for solutions. It’s not like they’re out there shopping for a good deal on AI. Like, hey, how much for a large AI? Or where can I get more of that AI? Marketers aren’t buying AI. So if marketers aren’t buying AI, and the term AI is kind of nebulous, what are they buying? Exactly. It’s getting a little murky. Let’s clear it up. We’ve talked about a lot of AI’s virtues, misconceptions and challenges, like how AI models are bringing groundbreaking change to marketing, how they’ll impact the way we’ll work in the future, how they’ll drive efficiency in retail, the importance of super solid training data, and the challenge of unconscious bias. All of these things are important, and they can certainly be decision factors. But ultimately, when the rubber meets the road, how do AI powered solutions get bought and sold?
In this episode, I check in with three impressive insightful guests to explore best practices for vendors selling AI powered solutions, and best practices for marketers looking to implement AI powered solutions. We’ve got a number of things to talk about. When I do the math here. Let’s see a couple plus a few. Looks like that number is five.
Number one, if you’re selling solutions powered by real AI, then you should really try to avoid pinning your story on AI. Number two, if you’re a buyer, spend most of your time diagnosing your problem and far less on finding the solution. Three, sellers should sell performance first, not technology. Four, buyers and sellers should expect clarity and transparency from each other. And finally, if you’re thinking about adopting an AI powered solution, it’s better to take a step and fail than not to try at all.
Listen for these themes throughout the interviews. Also, this episode is packed with a whole bunch of other insightful stuff from trust to integration to culture. My guests share a ton of great advice.
If people claim to use large amounts of AI, it’s always worth verifying that because what you don’t want to do is get into bed with somebody who’s basically just catfished you, and there are a lot of vendors out there who will catfish you.
That’s Perry Malm, CEO of Phrasee, a company that provides AI powered solutions that simplify and optimize marketing copywriting.
So if you want to find out, you know if people are using actual AI or not, there’s a few different techniques which can befall, we’ve got a bunch of resources for stuff like that on our website. But basically, if people can’t answer direct questions, and if they can’t sort of explain the data which drives their models, and if they can’t, sort of, you know, justify the specific types of AI which are being used, and it’s off, you skated with all sorts of buzzwords, and this and that and the other, then you know what they say, if it walks like a duck, and it talks like a duck, well, it’s probably a duck in it.
So I hear you that if a company is able to explain the data they have access to and that they’re using for their training data, that’s a good sign, I get that. But then you also said, you don’t need to be an AI expert, or you don’t need to dive too deep into what the AI is, in order to know if something’s a good solution. So how do I balance those two statements? Because they seem contradictory.
Yeah, I got you, I got you. And I guess the point that I’m trying to explain is, Caveat emptor, right? You don’t want to start working with a vendor who has been dishonest to you upfront. So while your problem may get solved with non AI technology, and that’s awesome, if they’ve been dishonest to you upfront, they’re clearly not to be trusted. And you should hold vendors to a higher account. So I think my point about you know, whether or not its AI stands, but if people are flying the AI flag, make sure they aren’t bullshitting you, because that’s a sign of bad things to come down the road.
How do I know if they are BS’ing me?
Yeah, well, it’s tough to be honest. But there are a few telltale signs, if they can explain exactly what data you need to give them and what they’re going to do with that data, that’s a good starting point. If they don’t need much data, then they’re probably not working with very much AI. If they talk at length about this sort of magical system where you just turn a key and it does everything for you, that’s probably not right. Like, here’s the dirty secret about AI, pretty much none of it is turnkey black box off the shelf, all of it needs, you know, a high level of skilled customization to actually work for you and your brand. And if I’m perfectly honest, this is where talking to somebody who knows about this stuff can probably help. And like, there’s no hard and fast rule to it, but you would perform the same level of diligence, you know, if you were bringing in a new accountant, you would ensure that they were, you know, qualified to count the beans.
The term AI is used so broadly, it’s kind of hard to know what it means. Whether you’re a buyer or seller, is it even a useful term at this point?
So here’s the thing about the phrase AI, Artificial Intelligence. It doesn’t actually really mean very much. What it actually is, is it’s an umbrella term that covers off myriad technologies, you know, things like machine learning, and deep learning, which are all the rage right now. But it’s also, you know, older technologies, like ocular character recognition or like expert systems. When you get down to it, though the actual underlying technology, to a practitioner, like to a business person doesn’t actually matter, you can leave that to the AI nerds, right? What actually matters is, it’s computers doing intelligent stuff. And that’s all that actually matters. So basically, when people ask me what AI is, my answer is really simple. It’s computers doing stuff that previously relied upon human intelligence or intuition.
So to summarize, AI isn’t necessarily changing marketing just by making new things possible. But rather, it’s improving upon things that humans have done previously.
Yeah, like, I mean, if we think about in the last sort of four or five years AI has been, you know it’s become a real like buzzword. And the reason for that is like, three things happened all around the same time. So number one, and Vidya came out with GPU, a graphical processing unit, which happens to be very good at processing multidimensional mathematics at speed and at scale. It’s originally made for like 3D video games and stuff. It’s almost like factor number one, factor number two, to make these you know deep learning systems work, you need loads of data going into them. And like we’re not short of data. I mean, just look at you know, the 7 million articles or something on Wikipedia, like we’ve got data coming out the Yin Yang, so that’s no longer a barrier. And then the third thing was the democratization of deep learning algorithms. So when you combine these three things, so better hardware, more data, and access to better software, you’re then able to start applying AI in situations that was never possible previously, so that’s why we’re hearing about it now.
Perry also hit on a topic that turned out to be thematic across all my conversations with this episode’s guests. Where to focus most of your time and energy when considering marketing solutions. Here’s a hint, it’s not looking at vendors.
So like the advice which I give here is this, if you have an hour to solve a problem, spend 55 minutes thinking about the problem, and only 5 minutes worrying about what the solution is. Because what you don’t need is a whole bunch of technology on top of technology on top of technology, what you need is a solution to your problem. And once you deeply and fundamentally understand what your problem is, the right vendor, the right technology will become obvious.
So Perry, before we wrap, can you tell us a little bit about your company?
So Phrasee is AI powered copywriting. So think about it like this, the way that brands used to get coffee to use in their advertising campaigns, they would write a brief and they would send it to someone like Don Draper. And he’d think for a bit then he’d say, here’s the language that I think you should use. And then it would go back to the marketing team, and they’ve used it in various campaigns. The problem with that is that you’re entrusting millions of dollars of advertising spend on the gut feeling of one person. So first of all, you need to test it to make sure it works and make sure it’s right. And second of all, who’s to say that Don Draper is right? So the new way of working is what Phrasee does, instead of briefing Don Draper, you brief Phrasee’s AI system, and the AI generates the language for you to use in your campaigns. And we beat humans nearly every single time a campaign is sent out in a heads up test. And if you get you know, emails, or if you get served ads from brands like eBay, brands like Groupon, Walgreens companies like that, then you’ve experienced Phrasee’s technology without knowing it. And you’ve probably bought stuff as a direct result of machine generated language. So I don’t know if I should be saying congratulations, you’re welcome or apologizing. But I’ll leave that for you to figure out.
Don Fluckinger is a Senior News Writer at TechTarget, a company that is 80 technology websites. Don’s beats our customer experience and content management, in business solutions, like sales and marketing automation platforms.
I’m wary of companies that anthropomorphize their AI with mascots, you know, dead white guys like Einstein, and Magellan and Watson and all that, because really, all AI is, it’s machine learning bots that save you time and get you to insights that you didn’t know before and couldn’t have possibly figured out for yourself. We don’t have Einstein and Da Vinci or a genius inside our machine, we have a 1000 Little bots that are doing this work for us. And the companies that sort of, don’t necessarily brag on it, are probably doing the right thing on the back end. I know we’re not talking about names a lot. But Microsoft too, Microsoft does not really make AI the main selling point of anything. But if you ask them what AI is doing, they’ll tell you, but they’re not trying to sell that. They’re not trying to make it a big differentiator. It’s like everybody has AI. It’s kind of like table stakes, right?
Another point I wanted to make is that an analyst told me recently that marketers they’ve heard of AI, but they don’t know what it is, how it works or what it does. And this analyst’s particular problem was making up surveys to ask marketers about AI. They know so little about AI and what it does, that they can’t even answer questions about it. So when the analyst goes to make out his surveys for the quarter, to talk to marketers about AI, he has to talk about the specific use cases. Because if he puts, How do you need AI in your marketing stack? People won’t answer it. But if he says something like, Do you need an AI powered product recommendation engine? He’ll get tons of answers. Yes, we need that. Or would you use AI to help with your marketing segmentations to make more effective groups to market to? They say yes, but if you have general questions about AI, they either don’t know, not applicable, or they don’t even answer the questions. So I think that informs a lot of the AI discussion with marketers, you have to sell function, you have to sell performance. You don’t sell the technology. You don’t sell the whiz bang data tools that you’re making, it’s all about the results.
That’s where the clarity and transparency are so important. In order to get those results, the marketer and vendor have to be in sync. As Perry pointed out earlier, AI isn’t turnkey off the shelf stuff.
There’s a couple other things you have to think about too. And that’s how hard is this going to be to integrate into our current technology stack? Because all companies have data that isn’t quite ready to feed into these systems. So how much cleanup, how much normalization and cleansing is going to need to happen, right? So marketers aren’t data scientists, and they don’t know how the backbone of their systems work, right? So they should be consulting with their IT people. You know, with every buying decision, the IT people don’t know what the marketer is trying to sell or how the funnel works. And, you know, the marketing people don’t know how the backend data systems work, and how the CRM is glued together to the marketing system and all that stuff. So not only KPIs, but also how the data integration works, as well as talking to other customer references from companies that look like yours, if you’re a b2b or b2c, and of a certain size, that vendor should be able to say, hey, somebody who does sort of what you do, uses our tool and you talk to them in private, right?
And so also, another thing, another KPI is privacy compliance. How does a tool provide compliance with things like the European GDPR Privacy Rule, the California Consumer Privacy Protection Act? That’s a mouthful. And you know, there are a bunch of states are coming out with consumer protection rules too, that are in various stages of being approved by legislature. So a vendor needs to prove to a customer that they’re going to be able to comply with all these things, and much more importantly, document compliance, because, God forbid, I, as a marketer get in trouble for a compliance violation for consumer privacy, I should be able to document everything I did to make sure that I was complying with it. And we did a good faith effort to do everything, all this patchwork of consumer protection laws force us to do.
Remember what I was saying earlier about marketing to marketers? Don definitely has a perspective on that.
The other thing is these technologies, you’re marketing to marketers, you’re marketing to salespeople who are out there marketing and selling all day. So you have to bring your A game, you know, you can’t play a player, right? So if you start getting into some confusing jargon, or whatever, the marketing and sales, people who will be using this technology can smell that very quickly. So you’re right on that market. And all they care about is metrics, revenue, measuring it, and immediate turnaround. And I think it’s fair for salespeople and marketing people to demand that, because if I’m going to spend a bunch of money with you, I have to prove that that was a good investment.
So I, Liz Miller, I’m Vice President and Principal Analyst with Constellation Research. We are a boutique analyst firm here in the heart of Silicon Valley. There are lots of companies that are good at crunching data, it’s harder to find companies that actually value data. And by value I mean, like respect it, like it, love it, not be terrified of it, want to go have fun in it, right? Organizations that get so mired in what to do with data and are like, we’re going to be good at it, we’re going to be good at data, there’s probably not a whole distance, that they’re going to be able to cross with AI because they’re not actually looking to do anything different. They’re just looking to do something faster. So I would say companies who have a culture that embrace data, and all the good and the bad, and the weird and the ugly that come in between that, those are the companies who are probably going to get much further in a journey towards and it kind of an autonomous enterprise where AI is running, you know, a lot of those kind of, you know, fundamental rote tasks and allowing their people to embrace something completely different and something new and exciting and creative, and all of those things.
And Liz has some really poignant things to say about looking inward before looking to AI to solve your problems.
I think that AI has become a really important tool when used correctly and properly. I actually think though, when used to accelerate random acts of marketing, it has become an unintentional tool that automates bad behavior, rather than helping empower best behaviors. And I always think about something that a friend once said to me about automation, which was, you know, if you keep automating your bad processes, all you’re getting are really fast bad processes, right? And so AI is never going to solve your bad process. It can accelerate it, it can combine it to a whole lot of other bad processes. But it can’t magically fix the broken machine.
Where I think AI has been incredibly powerful, especially for marketers coming into marketing organizations that feel broken, is that they’ve been able to identify those points of friction, and identify where the marketing process or where marketing operations might be broken, so that you can go and then fix it, right? But it’s always going to take that strategic mind, that leader, the change agent, who’s going to go in and be like, Oh, okay, I see what our problem is, you know, the system is telling me that after looking at all this data, these three things don’t align. Why is that? AI is not going to go on to automatically align them for you, at least to a degree that you’re expecting. But I think that it has been able to point out a whole lot of stuff that we wouldn’t have necessarily been able to get to, had we not had access to the AI or the ML tools that we have access to today.
Does that mean we have to fix the broken machine, as Liz puts it, before doing anything in AI? Liz points out that failing is an important part of the process. And that a fail fast corporate culture can ultimately help companies find greater success.
Great. So when you start to realize that the process that you were automating, through whatever automation system you were using, was actually creating a negative impact or a negative effect, right? It wasn’t the smartest thing you could have been doing. And you could have been doing something better. And I think in a lot of corporate cultures, that turns into an opportunity for punishment and repercussion. So as much as I say that, you know, a culture that values data, is going to excel in this world of AI, a culture around exploration and a culture that embraces almost in values, that kind of concept of strategic design, where you fail small, fail fast and fail forward, right? And with that, we’re using these tools. And we’re using these opportunities to find ways to continue to advance and explore and challenge and experiment. I think those cultures are the ones that are going to be able to not only identify those processes that can be automated, but also identify those processes that can be changed, and should be changed for the better.
It’s fitting to wrap up with the customer relationship, keeping that relationship top of mind at all times as we test and implement AI solutions.
You better have some really crystal clear ideas of the goals you want to achieve. So that you can look at all of these implementations of AI, and all of these implementations of technology to understand how all these goals fit together to benefit the customer and to benefit the brand, right? If you don’t have that clearly defined, if you don’t have that vision there, again, you are now just accelerating random acts of marketing, and the risk you run is alienating your customer.
Before we go, I’ve invited Jeremy Lockhorn back to the pod to reflect on the perspective shared by this episode’s guests, and share some personal experiences of his own.
I spent a very long time on the agency side and a lot of my role was what I flippantly described as sitting in the gatekeeper chair to evaluate either new technologies or new vendors, new potential partners who are coming in and trying to sell their wares. So I got really good at quickly separating the good from the bad and you know, there’s just, you know, a lot of themes and best practices that I’ve sort of developed as a result of all of that, and I’m going to break them down into two categories. One is kind of the pre pitch, you know trying to get the meeting. And then the second is actually what happens during the actual pitch discussion.
And so let’s talk about the pre pitch first. One of the most annoying things for me was you know, preponderance of buzzwords either in an email or on a website or a one shooter, you get to the point where you played buzzword bingo, and there’s just so much there that you can’t separate what it is that the company really does. And that’s especially true when it comes to complex things like artificial intelligence that you know, is a relatively new and emerging field that not everybody has a great handle on.
The other one that’s kind of funny is just really bad personalization going awry. So like, I literally have gotten several LinkedIn outreaches when I was on the agency side that said things like, hello, bracket, first name, and bracket, you know, I’ve been doing research on bracket, company name bracket, and I think you might benefit from our tech, you know, are you are you free for an initial discussion tomorrow or next week, or whatever it is. And you’re like, really, so you know, just kind of buttoning up those really simple things is clearly the best practice.
The last annoyance in the pre-pitch phase for me was around lead gen content that promised the world and delivered no value, you know, you’d find a really interesting report that seemed like it had a bunch of great data and analysis and insight in it. Based on you know, the story that they’re telling to get you to enter your email and contact information. And then, you know, you go through all that process, you download the PDF, or whatever, and it’s just garbage. Would put those kinds of vendors on the bottom of the queue, for sure.
So then, like, once you actually get into your pitch meeting, that initial discussion, I think one of the things that’s super important is to be crystal clear about the purpose of the meeting you’re about to have, and to calibrate your story and manage expectations accordingly. And specifically, what I’m talking about is, there were times when I would be looking for a very specific solution to solve a very specific problem that my client had, right? You know, in those cases, I have a clear picture of what exactly I need or the problem that I’m trying to solve.
And then there were other times that were just purely exploratory, right? Like, you know, I don’t have a specific need to solve at this time. But, you know, my client, trust me to keep my finger on the pulse of what’s out there, and be curious sort of on their behalf. And then, you know, filter what’s noise and bring them, you know, the interesting things and the interesting trends, you know, those are obviously two very different kind of meanings. When I would set those up on the agency side, I always tried to be really clear with the potential partner that was coming in about my objectives, and sort of which category of meeting that this would fall into. And I was shocked at how frequently a potential partner would come in, pitching the opposite way, right? Like trying to get to a sale when I’m in an exploratory meeting, or vice versa. And so I think that’s a really interesting dichotomy.
I’m curious, looking back, where did you have most of your big wins? Which kind of meeting?
The exploratory kind, yeah. I frequently would find something that was kind of surprising. Like, that’s where you find the breakthroughs, I think, more often than not anyway, because it kind of comes out of left field sparks an idea. And you know, you’re kind of off to the races. So in the pitch meeting, or the live meeting category, that the second thing I would say, was always a nuisance, or an annoyance is poorly done homework. You know, a potential partner coming in with assumptions made based on what they see from a particular brand as a consumer, for example, right?
And if you think about what’s consistent with really great brands is that their strategy doesn’t necessarily show, right? Like they sort of transcend the obvious, right? And then they come in and the pitch is sort of tailored against those assumed strategies. And more often than not, they’re completely off base. You know, it may be that they assume, you know, the ultimate objective from a communications perspective incorrectly. It’s a shame because you can discern the strategy in much more accurate ways than just trying to assume a strategy, from what you perceive as a consumer.
The takeaway is, it’s not that you shouldn’t assume what the strategy is, it’s just that you should take in more training data before making your assumption.
That’s a fantastic way to say it. One of the other things that I’ve seen frequently is telling more than asking. So a potential partner may come in with a fantastic story. But if they don’t align it to my specific needs as an agency or my client’s needs, it’s difficult to make that story resonate, right? And so it kind of comes back to the previous thing about doing the homework, you can do some of that in the first few minutes of the meeting by just asking really smart questions. And then sort of adjusting the rest of the discussion based on the answers that you hear.
I have not spent any time on the agency side. So I can’t speak specifically about my own experiences. So I’m going to take this and the exact opposite end of the spectrum, which is let’s go much higher level. As I heard these suggestions from our guests, including the suggestions that you just provided us, by the way, I keep thinking about just general life truisms that I’m reminded of that are true about life just as much as they are about selling technology to agencies. And it’s kind of true for all of them. Like for instance, when we hear Perry say, if you’re selling real AI then you don’t really need to talk about AI. Like to me, that is just basically the same as you know, braggarts are hiding and securities or confident people don’t need to brag. And when you hear about how you know, it’s more important to spend most of your time diagnosing your own problems internally within your company, before looking to external solutions to solve your problems.
I’m reminded that the same is true of human relationships, romantic and non romantic. You know, you read a lot, and you hear a lot about how you need to love oneself before having a great external relationship. You can’t depend on other people for happiness. I mean, very similar sort of truism, when you hear that buyers and sellers should expect clarity and transparency from each other. I mean, that’s just a human relationship. And then lastly, this thing that Liz talked about, of fail fast, fail forward, as she put it, this concept that if you’re thinking about adopting an AI powered solution, it’s better to just take a step and fail, rather than not do anything at all.
What could be more true about life? You know, as Yoda said, there is no try only do or the general truism of better to have loved and lost than not to have loved at all just this concept of just put yourself out there and do it and figure it out from there. So I just noticed that so many of these things we heard are themes that are generally about life, not just about selling technology, and not just about being a marketer.
I’d like to thank my guests Perry Malm, CEO of Phrasee, Don Fluckinger, Senior News Writer at TechTarget, Liz Miller, Principal Analyst at Constellation Ventures, and of course, our own Jeremy Lockhorn. To learn more about my guests, check out the show notes for this episode in your podcast app or emodoinc.com. Hey, and if you like the show, please write us a comment and give us a rating on your favorite podcast listening platform. Especially if you’re listening on Apple podcasts. We’d be super grateful. It definitely helps more people discover the show. Thanks for joining us.
The FIVE podcast is presented by Ericsson Emodo and the Emodo Institute and features original music by Dyaphonic and the Small Town Symphonette, original episode art is by Chris Kosek. This episode was edited by Justin Newton and produced by Robert Haskitt, Liz Wynnemer and me. I’m Jake Moskowitz.