Measuring, Reducing & Optimising Push Notification Churn

I had an energising chat last week with a marketing leader at a top company in the consumer mobile web3 space and she made an insightful observation that caught my attention:

Marketing leaders over-index on performance marketing because the tooling and workflow lend themselves to rapid action-result loops. They can try things, see results, and iterate very quickly.

That's a naturally compelling feedback loop for humans in any endeavour of life, and especially for ambitious and energetic business leaders.

Then she said more:

The result of this appealing tool-workflow-feedback dynamic is that marketing leaders become increasingly 'short-term greedy'. They want results and they want them now, and the short-term/immediate results of performance marketing (e.g. app downloads) give them that gratification.

We know bad things happen when we over-optimise for short-term results. But normative, philosophical arguments against this kind of thing have never been very satisfactory for my practical orientation. We need our tools to do a better job of measuring long-term, incremental effects, while also connecting those long-term effects to the short-term feedback loops that influence more of our day-to-day decisions.

Push notifications are a good example.?

You can aggressively increase your push notifications and your bottom-line clicks are probably going to increase (a short-term result). Sure, you might see a decline in the rate of clicks, but the Rate is far more abstract than the clicks themselves. If I'm driving more business, does it really matter that my rate is going down??CRM and marketing leaders have specifically asked me this and they often answer the question themselves in the negative.

The problem is the unseen long-term effect: Some sequence of non-clicks converts to an Unsubscribe. Some portion of your customers make themselves unreachable. Unreachable customers are measurably less valuable to your business than reachable customers.

CRM tooling doesn't make it easy to SEE (i.e. measure) that connection. That's partly because it's a non-deterministic process. A click is an action ON the notification - the cause and effect are deterministically connected. Unsubscribing is not.?

Many apps don't effectively track unsubscribes, not at a user-level. But even if they do, an unsubscribe happens AFTER a notification is received (which itself is fuzzy - you know when you sent it, but you don't know when it was SEEN). And there's no direct mechanism that connects the two events. Usually people don't get irritated enough by 1 notification to make the effort to unsubscribe. It's a cumulative effect. And the rate at which notifications accumulate enough to cause a user to unsubscribe is statistical?and unobserved.?

Aampe has been drilling into this problem and we've seen some exciting breakthroughs in the last few weeks.?

Aampe's infrastructure will automatically model the rate at which notifications are leading to unsubscribes, and then continuously optimise notifications to reduce that rate and the overall percentage of users unsubscribing. We won't just give you the data in the form of analysis or "insight" - our infrastructure will ensure that your rate of sending will ACT to minimise your unsubscribes.?

And it doesn't stop there. While you want to reduce unsubscribes, your main goal remains driving bottom of the funnel transactions. As long as your transactions are going up, even if you know more precisely that you're making customers unreachable - you might still accept the trade-off and increase your notifications.

So instead of just reducing unsubscribes, our infrastructure will connect the 2 goals to precisely optimise the trade-off. You see the curve generate continuously - you move your lever up and down to decide what's most important for you - getting someone to buy something THIS WEEK - or getting them to stay around to do a transaction next month, and the month after.?

If that isn't an awesomely compelling capability -- what is!!??

To learn more, reach out to James Laurain or me (or anyone at Aampe ) and we'll be happy to help you identify how you can start connecting long-term retention metrics with your short-term marketing efforts.