Getting customer text messages right

Getting customer text messages right is important. The phone may be the most intimate form of interaction between you and them, and it’s easy to overstep the mark and poison the relationship. Let’s look at the good and bad of text messages.

The good

It’s worth getting customer text messages right because they can play a significant role in behavioural influence. For example:

• Text messaging students has been shown to reduce “summer melt” in the US where high school students lose all motivation for attending college during their long holiday;
• The ATO generated $800 million in tax payments by sending 560,000 pre-emptive SMS reminders; and
• Individuals who received text message reminders in Chile (PDF) saved substantially more than individuals who did not.

Getting text messages right can come down subtleties. A few years ago, for example, I wrote about the contrast in how my osteopath and hairdresser confirmed appointments with me.

Notice the osteopath doesn’t ask me to contact them unless I have a change to make, whereas my hairdresser requests I acknowledge the appointment. There are two advantages of this: first, the action of typing ‘yes’ helps the customer encode the appointment in their brain; second, it means the business knows the customer has received the text. The osteopath, on the other hand, doesn’t know whether or not their customer has received the message, increasing the risk of no-shows.

Here’s another example of a business getting its text right:

Budget Direct had been trying to contact me by phone, to no avail. Their next step was to text me to gain my attention, but what if I mistook it for spam? Including both my policy number and description of my car allayed any concern I had about it being a scam (which I had thought from the first line), while increasing my concern that something had gone wrong with my policy (i.e. it must be important). It’s not a perfect text — for instance addressing customers by name and providing a consultant’s name (if one is assigned) would make it better — but it worked.

The bad

Unfortunately most messages I receive are plain annoying, either because they are spruiking stuff that I have no interest in or because they have poor usability.

Example four comes from Nissan who sent the text a month after the problem had been fixed by their customer and failed to provide any path for resolution anyway. As this and example five from CPA Australia illustrate, customers hate dead-ends, so if you are asking them to take an action, give them a way to do so.  

Another example of a usability fail comes from Australia Post, which kindly texted to remind me of a parcel delivery. The problem is the text was more for them than me. What do I mean? They wanted me to text back whether I would be home or not — a benefit for their delivery process. Because they neglected to provide any delivery window smaller than a whole day, I couldn’t say for certain that I would be home.  

If I select “2= Someone will be home” then that would kind of commit me to hanging around, so I would better not to choose that even if I probably would be home.

Does that mean I should choose “3= Take to Post Office if not home” as the only other option? Well, no, because they’ve created a logical fallacy. While they’re trying to say that if I’m not going to be home I should choose three, the way it’s written suggests if I might be home I should choose two AND three, i.e. I will probably be home, but if I’m not, I want you to take it to the Post Office. It’s not clear whether if I choose only three they will try to deliver to my home first and take it to the Post Office only if I am not home. Confused? I was.

What was the result of this text message? I ignored it. I didn’t select either option because they didn’t make sense. A fail for both delivery optimisation and customer service.

To get this right, AusPost should have:

• Provided a delivery window. It’s much easier to say whether I’ll be home between 2pm and 4pm than a whole day;
• Retained option two but restated as “Someone will probably be home” to reduce the obligation on the customer to commit with certainty; and
• Revised option three as “Take to Post Office instead”, dropping the confusing “if not home” part.

Our final example is the horror of an unsolicited customer survey from a business. Telling me I am about to receive four questions is not the way to engage your customer. If you must survey them (something I don’t recommend), then at least ask them to agree to taking part first.

Texting guidelines

Texts can work for both you and your customer, but you need to get them right. As we’ve seen, crafting the right customer text message at the right time is not easy and, unfortunately, quite rare.

• Timing: A reminder the day before an appointment is good, a day after is not. Make sure you think about the best time to contact your customer to maximise their likelihood of taking the action you want.
• Call to Action: Provide a path to resolution for your customer — don’t leave them hanging. For appointments, ask them to confirm with “yes” rather than assuming they’ll see the message and contact you if they need to change it.
• Path of least resistance: Make it easy for them to take action. If you are texting, let them text you back or if you have to take them to a website, make sure it is mobile responsive and try to pre-populate their details. A text opt-out (e.g. text “stop”) is preferred by customers to a web form.
• For them, not you: Remember to balance the message so it’s perceived as value for them, not you. Texts should make your customer’s life better or easier, otherwise you will annoy them and taint the relationship. Remember the AusPost example above — it could have been a service for my benefit, but instead seemed like it was for them.

NOW READ: Text message scams are on the rise and it’s costing Australians millions of dollars