Powering Value (and Product) Through Conversation

Apr 1 · 8 min read

The Simple Power of SMS

By Trevor Tarakjian

Have you ever had an engaging conversation with your favourite app?

At Hopper, we like to challenge what it means to engage our users, how engagement can lead to new experiences, and what driving experiences through personal conversations with an app that you rely on for accurate travel predictions might look like.

The potential value behind what we dub “Business Chat” is hazy. Business initiated discussions are rare. Sure, Spotify might tell you that a new album from your favourite artist is available to stream, or perhaps Uber is forcing you to vacate your couch to come pick up your food delivery, but outside of these small, purpose-driven interactions (let’s not call them conversations just yet), lengthier discussions that provide a meaningful experience, or add to an existing one, are few and far between.

Does anyone really want to talk at length with an app? Surprisingly, there are quite a few who do.

Over the course of the last few months, we’ve been having thousands of independent conversations with our users to determine how we can optimize the time our users spend between app experience and the flight experience. We wanted to create a balance between continuous learning and providing value.

Originally, we were using the accessibility and high read-rates of SMS to achieve what we didn’t have the resources to do in-app at the time. It provided a cheap, effective solution to signal user intent: to gauge user interest and to capture early feedback on new product offerings.

Ideas lived or died through a text.

With SMS we didn’t need a fully fleshed out chat platform, and we didn’t need experienced agents online all the time. Since SMS is asynchronous, we could leave and pick up the conversation without worrying about how long a user was waiting in-between responses. We actually saw that intent through push notifications with similar messaging had roughly the same conversion rates for the same product offering through SMS, but with the first few hundred conversations that we were having, we quickly found that customers wanted to continue a conversation we had started just as a simple probe, and we could actually sell users on other relevant and useful products (like trip insurance and adding bags ahead of time to avoid incurring day-of airport charges) that they may not ever have engaged with had we never kept the discussion going — something we couldn’t do through push notifications alone.

So, how do you successfully complement your mobile flight and hotel booking platform with a separate, text-based user journey without alienating your user base in the process? It comes down to what we already do well.

How do we communicate?

At Hopper, our users are already talking to us. They receive notifications when the prices drop or increase for the flights and hotels they watch and take action based on our recommendations. We wanted to continue that trend: Keep users actively engaged, and have our conversations be relevant to what the user cares about. Based on the interactions that the user has with Hopper during the shopping experience, we can shape the conversation around what they care about most.

Still, to use SMS as a testing ground for conversational purposes was daunting. There were a few questions that we had to solve before we began to scale up:

  • Will this feel too invasive?
  • How do we communicate value effectively?
  • How do we make it more human?

We also needed to avoid these types of experiences:

Are you a real person or an automation?

And replace it with this:

Fantastic! Thanks. So cool that you’re a real person haha

Because our push notifications are already designed to drive users back to the app, and our SMS messaging takes place outside of the core app experience, we needed to make sure that users were not confused by the two separate interactions and understood that they both originated from the core product. The simple solution to having a cohesive experience is to be as personal as possible, leverage your existing data, and rely on the trust you’ve already established with your brand.

Here’s an example of one of our outbound messages that has all three qualities:

Hey {{first_name}}! Trevor from Hopper here — Thanks for booking your flight to {{destination city}} with us! Can I ask you a quick question about your trip?

In the five seconds that it takes to read the message itself I’ve already introduced myself, included the user’s first name, and added the personal touch of including where they’re departing to next. While the messaging is still automated, we modified what information we included in our messaging depending on where we wanted to lead the conversation.

Ultimately we found that including even one element of what the user already accomplished in the app was enough to trust that the message came from someone at Hopper.

The use of the rabbit emoji (or even some of our animated GIFs) is equally as important — it serves to reinforce our brand and establishes a connection to the app that they’re already familiar with (we have a lot of bunnies):

Using this method, we were delighted to find that customers were more than willing to communicate with us. Here are a few real-world responses that showcase their appreciation of the medium:

I appreciate the option and the text format: super convenient! Thanks again

Awesome. Thank you. This is my first trip booked through Hopper, so far I am very pleased with the experience. Kudos for the text rather than the follow-up call.

What we didn’t expect to happen were users reaching out to others about the experience they had had with Hopper over SMS. Here’s one last example from a user that we loved:

Hi, I was wanting to book a flight to Amsterdam. A flight my cousin booked. I saw the price is low rn and she mentioned that you guys reached out to her […] She said you guys texted off this number and I’d love to know more details?”

It was at this point that we realized that SMS was an effective tool for our team that we could leverage to create all kinds of meaningful experiences.

Finding the Right Fit

For each product that we pushed over SMS, Hopper’s data science team was essential for determining who we messaged as well as how we adapted our messaging to the user. Depending on the use-case for each product we tested, we factored in several trip-specific variables:

  • Do we reach out to users flying to and from a specific destination city?
  • Is this user going on vacation or is it a business trip?
  • How many people are they flying with? Are they accompanied by family or friends?
  • How long have they been actively waiting to buy these trips prior to booking?
  • How important is this trip for them? Do they have everything they need?

While these are some of the more obvious variables, there are dozens of less explicit dimensions of a trip that we can analyze to find patterns in user behaviour. By modelling our messaging on these patterns, we were able to approach each conversation with a level of personalization that played to each user’s specific trip, maintaining that degree of trust we previously established and learning what messaging resonates with which type of user.

The approach used for an outbound message is as important as determining which variables apply to them. In our testing, we discovered how much variation there was between the effective open-ended questions performed compared to a direct up-sell.

In a sample size of over a thousand users, conversational openers improved conversion by 40% and increased response rates by 200% when compared to its up-selling counterpart with the exact same value prop. It was clear that over SMS our users not only preferred and reacted more positively to a natural tone, but they were also more likely to convert overall. We want Hopper users to feel great about the trip they’re taking, and to remove the guess-work that goes into researching and purchasing additional products they may need. Running these open-ended, conversational experiments continue to make experimenting seamless and decidedly user-focussed.


Back to that notion about ideas living and dying at the hands of the written word. How do you adapt the ideas that work in a text-based environment to a fleshed out in-app experience? Why do so in the first place? We needed to determine whether we could show the user the same value while they were in control.

We sought to validate our idea in a product we knew we could offer immediately after the user booked a flight: Flight Upgrades. We were excited to see if we could have the same rate of success that its SMS counterpart had achieved without any human intervention.

Our product team stripped down the conversation flows experienced over SMS and adapted it into an automated, choice-based, user-driven interaction. We didn’t need to map out the logic to make this work as we already knew what messaging performed best, what information to display, which users to target, and at what stage in the product journey to offer it — All based on the feedback we had by talking to our users over SMS.

In our first in-app run, a pop-over appeared on screen for select users after booking with a simple question:

“Did you know you purchased a ticket with some restrictions?”

Hundreds of users who purchase restrictive fares every month don’t even know what they’re buying despite the information shown during the actual shopping experience. For those that responded “No, tell me more” we displayed a follow-up message that clarified the restrictions of each user’s individual ticket and offered them the ability to upgrade with a single tap. Suddenly, 15% of users were upgrading for $50 per passenger on average.

Initiating conversations outside of the core app experience can exist alongside the native environment, and it can drive sustained, organic product development and growth if it is constantly learning about its users through real human interactions. Helping our users is at the heart of what we do, and if our users are helping to shape that experience, then we know they’ll experience their best trips when booking with Hopper.

Hopper is hiring! We have open positions at our Montreal, Cambridge, New York and Sofia offices. Check out our careers page.


Written by


Hopper uses big data to predict when you should book your flights & hotels. We’ll instantly notify you when prices drop so you can book travel fast in the app.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade