Ok, we’ve already talked about personalizing several different aspects of your message to engage your app users:
- Timing: Get your message to your user on days and at times when they’re likely to act.
- Frequency: Send as many messages as a user wants and needs — don’t send too many, but also don’t send too few.
- Copy: Write copy that speaks to particular value propositions or user purchase motivations, and find the users who see the world in those terms.
- Topic: Learn which specific products and product categories your users are interested in, so you can send them products and recommendations that are most relevant to them.
But, there’s one aspect we haven’t covered yet: Tone
Most apps rely on boilerplate fake-urgency when trying to draw their customers in (e.g. “Buy now!” “20% off TODAY ONLY!!!”), but this type of messaging burns users out — especially when they’re getting multiple messages with this same tone from multiple apps every day.
Your users are all different — Some want to be entertained or made to laugh, whereas others might find humor or joking annoying when it comes from an app.
Some people respond better to positive motivation — helping them discover all the good things they can get or accomplish — whereas others respond better to negative motivation, understanding the undesirable outcomes that your app could help them avoid.
At Aampe, we’ve developed a way to match users to their tone preferences
When you use Aampe, we use Machine Learning to score every message you write in terms of 28 different tones:
Using ML has two benefits:
- First, you don’t have to hand-label all your messages.
- Second, most sentences that are even remotely interesting are combinations of multiple tones and emotions, so, byy scoring messages instead of labeling them, we can pull out subtle tone signals that our learning algorithms can use to get people the messages that speak to them.
To illustrate what this looks like in practice, we scored all of the messages in Aampe’s Inspiration Library — our collection of actual push notifications sent by hundreds of apps across multiple industries around the world (You can check out Aampe’s inspiration library for yourself in our composer here.)
Here’s what the analysis shows (Bigger, darker squares means the tone featured more prominently in the industry’s push notifications):
And what can we learn from this analysis? Lots of interesting things!
Here are a few observations:
- Approval is a major part of practically all industries, but less so for the Music & Audio and Travel & Local industries (However, those two industries write notifications that span the entire range of emotion much more than any other industries.).
- Caring ("take a break!", "We'll take care fo you!", various mentions of safety, etc.) and Confusion ("Oh, no! You look tired.", "Just 1? How about 2?", and our personal favorite - "Don't be a fool!") were more prevalent in Travel & Local apps.
- Social apps were more likely than apps from other industries to express Amusement or Anger (wow...they're really playing to stereotype), as well as Disapproval, but also: Realization ("This is a lot harder than I thought it would be", "math lesson to be learned here?", etc.)
- Productivity apps were much more likely to express Optimism (so the social apps aren't the only ones who play to stereotype), but also Amusement and Annoyance.
- Aside from the music and travel apps, Finance, Productivity, Lifestyle, and Education apps showed the greatest breadth of tone in the messages.
- Notifications from Entertainment apps were the most likely to express Love.
- News and Sports apps were the most likely to express Disappointment.
- Food and Drink apps were the most likely to express Joy, but they were trailed closely by Productivity, Travel, and Music apps.
But perhaps the biggest finding isn’t which tones the apps in each category are addressing — the real opportunities lie with the tones that aren’t being considered.
For example, surprisingly few health and fitness app notifications focus on Excitement (“Get out there today and crush your last record!”), Desire (“Get the body you’ve always wanted!”), or Pride (“Feel great in your own skin!”)!
These are all valid messages that will resonate with different target users, but they’re currently completely underserved.
So, if you use Aampe, what does this mean for you?
Well, for one thing, it means you automatically get messages with the right tone to each user who wants that tone (You’re welcome!), but, in addition to that, you also get data that highlights where you have the opportunity to build out your tone coverage in your messaging.
For example, here’s a coverage report for one of our customers:
This customer is clearly doing very well at expressing Approval and Admiration, but as we saw in the analysis of our Inspiration Library, pretty much everyone expresses approval and admiration too. That’s table stakes, not a differentiator.
This customer is also doing reasonably well expressing Annoyance and Disapproval (every user has annoyances — successful apps actively empathize and show the user what they can do about it). To a lesser extent, this customer is expressing Curiosity and Optimism, but they could probably do more in those areas.
But look at the tones at the bottom of the plot: Embarrassment, Grief, Nervousness, Relief, and Remorse are drastically under-represented.
This customer sells home goods — How can someone’s home furnishings be a source of embarrassment or nervousness? How could an app that sells home goods show that they understand that, and offer the user a chance to fix it?
How many of us have felt buyer’s remorse, and how would we feel about an app that proactively gives us a way to limit the chances of experiencing that remorse? In what ways can a good find be a source of relief?
There are a lot of possibilities here.
By scoring messages in terms of their tone profile and surfacing this tone profile to you, Aampe helps you find new ways to appeal to each user and learns to speak to your customers in a way that addresses their goals, desires, and motivations..
It’s an all-around win.