Hacking Mobile Distribution and Deployment via SMS

As someone who worked on consumer mobile products for a good number of years (and who carries the scars of the endless search for mobile distribution), the chatter about new mobile services using text messaging as the interaction layer has been fascinating to watch. Kent Goldman from Upside has some of my favorite tweets about the emergent trend, remarking “Good old SMS is the new UI” and Jonathan Libov of USV had a longer post (click here) about how this trend developed as a UI and where it could go in the future.

With the release of Magic and Luka.ai, why is all of this happening at the same time? When Twitter launched back in 2006, its signature 140-character limit was originally created and imposed in order to ensure tweets fit inside text messages, which were limited to 160 characters. Since then, users could always tweet via text through Cloudhopper by syncing their phone to their Twitter account and texting their tweet to 40404. I still do it often when I’m on the go and don’t want to open TweetDeck but have a question I want to ask or thought I need to share in the moment.

Here’s a brief recap of how we got here:

  • Dearth Of Mobile Distribution Dampens Developer Libido: If you’ve read this blog over the years, I am a broken record on this topic — mobile distribution is so hard to engineer, yet while the overall market size of mobile users creates the largest technology market we’ve ever seen. For an enterprising mobile developer who has tried his/her hands with a few apps only to never get distribution, these conditions can force mobile app makers to look for other ways to capture consumer attention on phones.
  • Global Tidal Wave Of Messaging Apps: We all understand this, I trust — the massive footprint of mobile messaging apps which will eventually turn into commerce, interaction layers, and much more.
  • Avoiding App Store Noise and Taxes: Kent pointed out releasing a service via SMS also allows developers to avoid the morass which are the mobile app stores. This also reduces developer libido, working so hard over an app only to see the store flooded with copycats, black box featuring schedules, and drastic changes in UI which reduce search functionality and browsing. I’d add to Kent’s point that for digital transactions for services rendered through the app, some of these new SMS services could provide a bypass around Apple’s 30% toll in the app store.
  • Reduced Cognitive Load Requirements: It’s hard to design a transactional mobile app where the user doesn’t have to make lots of decisions. Uber is one of the few. But, if I could just text “I want X delivered a bit after 8pm” it’s just easy, and I don’t have to search for an app, open it, wait for it to load, and then find the input areas, etc.
  • Easier User Onboarding: Asking to download a mobile native app has many steps. Find the app. Push button to download and/or use TouchID to verify download. Wait for the software to arrive. Open the app, wait for it to load. Then, sign up, register, perhaps go through a tutorial….ah, screw it, I just want my burrito and some organic juice. Another friend, Robert Stephens, pointed out it’s an easier request of the user’s time and attention, and after a few interactions, an easier upsell to download the native app. But, to start, asking a user to just “Send us a text to 70707 to get started with Sandwich Valet” is a pretty easy call to action.
  • Immediately Cross-platform: Kent also pointed out going text-first on mobile also makes the service omni-platform from Day 1, even for a Windows Phone. That’s a joke. But, not really, think of Xiaomi in China or other OS and app ecosystems emerging. Whereas many startups are thinking about going iPhone-first, some of these new services can service everyone right out of the gate.
  • Conversational UI: This is sort of an aside, but related….My friend Max shared a screenshot of Luka.ai and called it “Conversational UI,” which is clever — delivering a service (in this case, in a mobile native app) in a format where users feel like they’re texting with their friends.
  • Gateway To Mine Our Conversational Patterns: Last year, we started to see apps deployed which mined our emails to build up our profiles and preferences for eventually delivering a predictive or AI-based solution. Like email, getting into our native messaging clients like SMS provides developers another fertile ground for mining our conversational habits and preferences. People have been wondering aloud about “How can Magic scale if humans are providing answers?” and the answer is, often, “AI.” Eventually, machines will be communicating with us, and we will likely not have any idea.
  • Potential Drawbacks To This Strategy: This all sounds great, but what are the drawbacks? It’s hard to build a bond with the user without the mobile app. Not having a mobile app blocks the service from leveraging core phone sensors (mainly GPS and camera). Some commerce functions require a browsing interface or creating a basket of goods, in which case typing out the basket is inefficient. There’s also a business model consideration for services which launch here but are on top of other services — additional tips and fees could conceivably add 3x to COGS of an item’s true cost. And, well, we must only believe it’s a matter of time before Messenger, Whatsapp, Snapchat, Instagram, maybe Twitter, and others do some version of this, too, inside their own “chat” experiences.

Haystack is written by Semil Shah, and is published under a Creative Commons BY-NC-SA license. Copyright © 2018 Semil Shah.

“I write this not for the many, but for you; each of us is enough of an audience for the other.”— Epicurus