Latest / Marketing Automation with Fexingo: HubSpot, Marketo, and Automated Customer Journeys / Why Marketing Automation Ignores Customer Voice Assistant Shopping
Transcript
- Lucas: Let me ask you something, Luna. When was the last time you bought something using a voice assistant—like an Amazon Alexa or Google Assistant? Luna: Honestly, never. I've asked it to add things to my shopping list, but I've never actually hit 'buy.' I think I'm not alone there. Lucas: You're definitely not alone. But here's the thing—according to a 2025 report from Insider Intelligence, US voice commerce hit about forty-two billion dollars last year. That's up from roughly twenty-eight billion in 2023. Luna: Forty-two billion is real money. So who's buying? And how does that affect marketing automation? Lucas: Great questions. The buyers skew younger—Gen Z and Millennials—and they're buying repeat purchases: groceries, household supplies, pet food. Things they've bought before. But here's the marketing automation blind spot: almost none of those voice purchases feed back into a brand's automated customer journey. Luna: Because the platform—say, Amazon—keeps that data inside its own walled garden. Lucas: Exactly. If you're a brand selling on Amazon and someone buys your laundry detergent via Alexa, that transaction lives inside Amazon's ecosystem. Your own marketing automation platform—HubSpot, Marketo, whatever—has no idea it happened. So you can't trigger a follow-up, a replenishment reminder, or a cross-sell based on that voice purchase. Luna: That's wild. I mean, we've talked about how automation misses moving events or hobby changes. But voice shopping feels like a whole other level of disconnection. Lucas: It is. And it's not just Amazon. Google Assistant purchases through the Google Shopping platform have a similar problem. Even if you get some data via Google's purchase reporting, it's aggregated and delayed. You don't get the real-time trigger you need for a timely follow-up. Luna: So what's the workaround? Can brands set up their own voice skills or actions to capture that data directly? Lucas: Some are trying. If you build an Alexa Skill that handles ordering directly, you can pipe transaction data into your CRM. But the friction is high—you need developer resources, and you're competing with the default 'Alexa, buy X' experience. Most consumers just use the built-in voice reorder, not a brand-specific Skill. Luna: Right, because the whole point of voice is convenience. Nobody wants to remember a specific Skill name. Lucas: Exactly. So the data gap persists. Now, Salesforce and HubSpot are starting to offer voice integration modules. Salesforce's Einstein Voice lets you log data via voice commands, but it's more for sales reps than for capturing consumer voice purchases. HubSpot has a voice app too, but again, it's oriented toward input, not ingestion. Luna: It feels like the real solution might be on the platform side. If Amazon or Google exposed voice purchase data via an API that marketers could hook into—with user consent—that would change the game. Lucas: Privacy is the elephant in the room. Voice data is sensitive. Consumers are already wary of smart speakers listening in. So any API would need explicit opt-in. But if you could get that consent, imagine the trigger possibilities. Luna: Like, 'Hey, we noticed you just reordered your coffee pods via Alexa. Want a subscription at ten percent off?' Lucas: Exactly that. Or a replenishment reminder that lands in your inbox three weeks later, timed to when you usually run out. That's the kind of automated journey marketing automation is supposed to deliver. But right now, it's a black hole. Luna: And it's not just about the sale. It's about understanding customer behavior. If someone switches from buying via a website to buying via voice, that's a signal. Maybe they're in a hurry. Maybe they're cooking and their hands are full. That context could change how you communicate with them. Lucas: That's a great point. The channel itself carries meaning. A voice purchase might indicate a preference for low-friction, high-trust transactions. So the automated follow-up should be equally frictionless—maybe a push notification, not an email with five paragraphs. Luna: But without the data, marketers are flying blind. They're sending the same post-purchase flow to everyone, regardless of how the purchase happened. Lucas: And that's the core miss. Now, there are some early-stage startups trying to bridge this. Companies like Voiceter and SoundCommerce are building middleware that connects voice platforms to marketing stacks. But adoption is slow. Most brands don't even know the gap exists. Luna: So what's the takeaway for a marketer listening right now? Should they start building voice skills? Lucas: I'd say, first, audit your data sources. Do you know which purchases come from voice? If you sell on Amazon, you can see 'Alexa' as a device in your order reports, but it's not a clean field. Start tracking it manually or with a simple script. Then experiment with one brand-specific voice Skill for a repeat-purchase product. Measure whether you can capture that data. The ROI might be small initially, but this channel is growing. Luna: And if the conversation about voice shopping has sparked something you've actually tried in your own marketing, or if you just find these deep dives useful, consider supporting the show. It keeps us independent and ad-free. You can do that at buy me a coffee dot com slash fexingo. Lucas: Yeah, that really helps. And we appreciate it. So back to voice—I think the bigger point is that marketing automation is still very screen-centric. It assumes every interaction happens on a web page or in an app. Voice, like IoT and offline events, is a whole category of data that's mostly invisible to our systems. Luna: And ignoring it means you're missing a growing piece of the customer journey. Especially as smart speaker penetration keeps climbing—it's now in over forty percent of US homes. Lucas: Right. So the question is, will marketing automation platforms evolve to include voice triggers natively? Or will it always require third-party middleware? I think we'll see native support within the next two to three years, especially as consumer voice commerce crosses a hundred billion dollars. Luna: That feels optimistic. But I hope you're right. Because the alternative is that marketers keep treating voice purchases as blind spots, and customers get generic experiences that don't match how they actually shop. Lucas: Exactly. And that's the opposite of what automation is supposed to do. It's supposed to make the experience smarter, not dumber.