Latest / Marketing Automation with Fexingo: HubSpot, Marketo, and Automated Customer Journeys / How Marketing Automation Workflows Ignore Voice Commerce Signals
Transcript
- Lucas: If these marketing conversations have sparked something you've actually used in your own workflows, there's a new angle that's been quietly building steam — and most automation platforms are completely ignoring it. Luna: Let me guess — voice commerce? Lucas: Bingo. We talk about triggers based on email opens, page visits, form submissions. But what about a customer saying 'Alexa, reorder my laundry detergent' and then nothing else happens? That's a signal. A strong one. And most workflows treat it as a dead end. Luna: So the workflow sees a completed purchase — which is usually an exit trigger, not an entry trigger. Lucas: Exactly. The order fires, the workflow closes the loop. 'Customer bought — done.' But with voice, the purchase itself is often the beginning of a conversation. The customer didn't browse, didn't compare, didn't add to cart. They just said a phrase. That phrase contains intent data that a click never can. Luna: What kind of data? Give me a concrete example. Lucas: Take a CPG brand that sells through Amazon. They set up a voice reorder campaign — 'Alexa, reorder Brand X paper towels.' The customer says it, the order goes through. Standard workflow: send a thank-you email, maybe a cross-sell in seven days. But what if the customer said it in a rushed tone? Or what if they added a qualifier — 'Alexa, reorder the big pack this time'? That's a signal that their usage pattern changed. Most platforms don't capture that nuance. Luna: So you're saying the workflow should branch based on the voice query itself, not just the transaction outcome. Lucas: Right. And there are platforms that can do this — if you route the voice interaction data through a CRM that logs the query text and any sentiment flags. But the typical marketing automation platform doesn't ingest that feed. It's still built for web and email behavior. Luna: What about Google Assistant? That's a different ecosystem. Do workflows handle it any better? Lucas: Generally no. Google's Shopping Actions program — which has been rolled into Merchant Center — does pass some order data back. But again, it's transaction-level. You don't get the audio context. And the trigger logic in Marketo or HubSpot can't differentiate between a voice order and a typed order. It's the same 'order placed' event. Luna: So the opportunity is in the gap. The voice interaction itself is a richer trigger than the purchase event. Lucas: Exactly. And there's a real revenue impact. Voice commerce is projected to hit $80 billion by 2027, according to Juniper Research. That's up from about $5 billion in 2022. The growth is real, but the marketing infrastructure hasn't caught up. Luna: Let's talk about what can be done today. If I'm a marketer using, say, HubSpot — what's the first step? Lucas: First, you need to capture the voice query as a custom property. If you're on Amazon, you can use Amazon's Voice Interoperability Initiative or the Alexa Skills Kit to log the utterance text. Then pipe that into a custom field in your CRM. That alone lets you segment anyone who used a voice reorder. Luna: Right — then you can build a workflow that triggers on that custom field being populated. Lucas: Yes. And then branch based on keywords. If the query includes 'big' or 'bulk' or 'more', you can trigger a different follow-up — maybe a subscription offer or a bulk discount. If the query includes 'again' or 'same as before', that's a loyalty signal. Send a thank-you with a referral incentive. Luna: What about sentiment? Can you get that from the voice platform? Lucas: Some platforms offer sentiment analysis on the audio — Amazon's Lex can return sentiment scores. If you have a custom integration, you can store that as a numeric score. Then in your workflow, you can route angry or frustrated queries to a customer service follow-up instead of a marketing email. That's a huge win for retention. Luna: That's smart — using voice sentiment to prevent churn before it happens. A frustrated voice reorder might mean the customer almost switched brands. Lucas: Exactly. And the workflow can intervene immediately. Send an apology, a discount, or a direct call from support. The typical email automation would just send a 'thanks for your order' and miss the emotional context. Luna: I want to push back a little. How many brands actually have the capability to pipe voice data into their marketing automation? Isn't that still niche? Lucas: It's niche today, but it's growing fast. Any brand selling on Amazon with Alexa skills can do this — it's just a matter of setting up the data flow. And for brands using Google Assistant, the Merchant Center API can pass order origin data. The main blocker is that marketing teams don't think to ask for it. They treat voice as a sales channel, not a data channel. Luna: So the mindset shift is the real barrier, not the technology. Lucas: Exactly. And that's where marketing automation can lead. Instead of waiting for the platform to add 'voice trigger' as a native option, marketers can build it themselves with custom fields and API integrations. It's not trivial, but it's doable with a decent MarTech stack. Luna: Let's talk about a concrete example. You mentioned a CPG brand earlier. Walk me through the workflow they'd build. Lucas: Sure. Let's call them Brand X. They sell paper towels on Amazon. They have an Alexa skill that handles reorders. When a customer says 'Alexa, reorder Brand X paper towels', the skill captures the utterance and the sentiment score. That data goes to a custom object in their CRM — say, a 'Voice Order' object with fields for query text, sentiment, and product. Luna: Then the workflow triggers on new Voice Order records. Lucas: Right. The workflow checks: sentiment score above 0.8? Send a thank-you email with a cross-sell for the jumbo roll. Sentiment below 0.3? Send a 'we're sorry' email with a coupon and a phone number for support. If the query contains 'big' or 'bulk', add them to a high-volume usage segment for future bulk offers. Luna: Okay, so that's three branches. What about the content of the emails? Should they reference the voice interaction? Lucas: Absolutely. Personalization matters. The email could start with 'We heard you — you asked Alexa to reorder your paper towels. Here's a tip: you can save 10% if you switch to the 12-pack.' That makes the customer feel like the brand is listening, not just transacting. Luna: That's a big difference from the standard post-purchase email. And it builds loyalty because the brand acknowledged the specific context. Lucas: Exactly. And you can measure the impact. Compare the repeat purchase rate for customers who received a voice-aware workflow versus the standard post-purchase flow. I'd bet the voice-aware group has higher retention. Luna: I can see a potential pitfall though — privacy. Voice data is sensitive. Customers might not want brands logging their exact words. Lucas: That's a fair concern. But the data is already being captured by Amazon and Google. The brand is just accessing it through APIs. As long as they're transparent in their privacy policy and allow opt-out, it's no different than tracking web clicks. And the value exchange — better follow-ups, relevant offers — is clear enough that many customers will opt in. Luna: So the key is to be upfront and give control. Lucas: Right. And start with a small test. Pick one product category, one voice platform, and build a simple workflow. Measure the lift in repeat purchase rate or customer satisfaction. If it works, scale it. Luna: You know, this kind of thinking — finding signal where others see noise — is exactly why people listen to this show. Lucas: It is. And if these conversations have sparked something you've actually used in your own work, there's a simple way to help keep the show ad-free and accessible. A couple of dollars a month is genuinely what keeps these going — buy me a coffee dot com slash fexingo, if you've gotten something out of them. Luna: Yeah, it really does make a difference. We don't run ads, and listener support is what makes that possible. Lucas: So back to voice commerce. The bigger picture is that marketing automation needs to evolve from event-based triggers to intent-based triggers. Voice is just the first wave. As more interactions happen through conversational interfaces — chatbots, smart speakers, in-car assistants — the workflows that ignore context will feel increasingly robotic. Luna: What's the next frontier after voice? Visual search? Gesture control? Lucas: Probably ambient computing — where the device acts without explicit commands. Think a smart fridge that detects you're low on milk and automatically adds it to your shopping list. The trigger is a sensor reading, not a spoken word. That's even further from current automation capabilities. Luna: So the takeaway: start experimenting with voice data now, even if it's clunky. Build the muscle before ambient computing becomes mainstream. Lucas: Exactly. The brands that figure out how to capture and act on non-click signals today will have a massive advantage when the next wave hits. And the good news is, you don't need a new platform. You just need to extend the one you already have.