Why do some next best conversation programs work better than others?
Who hasn’t heard of “next best conversation” strategy? Nowadays it’s a ubiquitous part of any self-respecting Customer Experience program. Yes, it’s called different things; next best product, next best offer, next best activity, next best action (and sometimes means different things to different brands), but the intention is much the same, to ensure that every customer interaction, be it inbound or outbound, makes sense to both parties. That the next best service, product, offer or activity identified and suggested by the brand meets the unique needs of the customer. That it’s decisioned by the brand’s access to data intelligence about the customer. And of course, that it’s contextual when the interaction takes place.
We see these programs working exceptionally well for some clients, but in the course of meeting with potential clients, we also see programs that just aren’t performing. They usually include one or all of 5 common pitfalls.
1. Timing is everything
The thing with regular people (remember, that’s what customers actually are), is that they do stuff all the time. They’re active. They don’t stand still. Windows of opportunity open and close on an almost daily basis. Many organisations spend significant amounts of money developing next best conversation models that essentially rank things to talk to customers about – at a point in time. The problem is these models are rarely integrated into real time environments or refreshed in real time, meaning the most recent interactions, which should form the basis of any next best conversation logic, just aren’t factored in. Identifying recent interactions and responding accordingly, be that validating existing logic or changing the conversational tack, in real or near real time, is about being relevant – not just right. Imagine if one of your customers calls in to change their address. Do your operators rely on pre-defined offers just waiting for the customer to make contact, or, based on the call reason, or information the operator collects during the call? Is the system offering up relevant next best offers, in real time, to work into the conversation?
2. People don’t live in little boxes
We hear the word omni-channel all the time, and it’s true, your customers live an omni-channel existence. They research things on your website, unsubscribe from your newsletter, open or click email offers, return products in-store, and buy products online. Next best conversations in channel isolation just don’t work. It’s about seeing the whole picture, across all channels, and using this macro context to work out what the next best conversation should actually be. Take online retargeting for example. We’ve all been retargeted after viewing hotels in New York, Giant Mountain Bikes or iPhone 6S plans (or at least I have!). Retargeting works. Of course it does! But the problem is that even after we’ve visited the store or call centre, and purchased the product, digital retargeting continues to follow us around. It’s weird, then irritating, and then downright infuriating. When backend systems don’t talk to each other, channels end up working in isolation.
3. It’s not always elementary, Sherlock
Just because someone is browsing a particular product, and has an obvious interest, doesn’t necessarily mean that it’s the next best product to be focussing follow up activities on. Very useful insights are achieved by looking into the natural adoption pathways of customers, and interestingly it’s often the not-so-obvious products that offer a sales path of least resistance. And better meets their needs from what you may have inferred from their interest elsewhere. One of the examples I’ve seen is in the automotive space where car purchase and servicing go hand in hand. You’d think. But what we saw was ancillary services like roadside breakdown offers were a smarter direction, and became the immediate opportunity for next best conversations, the adoption of which actually drove servicing!
4. Put that sales pitch away
The thing about customers (remember, they’re real people) is that persistent selling can become tedious and alienating. There are always opportunities to build brand value with customers, but this kind of messaging is rarely a part of next best conversation programs. Sometimes customers need to hear about what makes you special, why you’re different, what your brand values are and how you might both connect on a more emotional level. Remember brand resonance significantly impacts important KPI’s, like NPS and advocacy scores and therefore should always be a part of any next best conversation strategy. An example comes to mind: a colleague of mine lost his wallet whilst we were travelling overseas. He called his Australian bank in a reasonably agitated state (I was sitting next to him at the time!) At the end of the call, rather than reinforcing the bank’s brand values of customer first, or their market leading peace-of-mind credentials, they proceeded to try and sell, wait for it……a personal loan next best product! I’m quite sure the loan cross-sell was an analytically rigorous decision, just waiting for the customer to appear, but sometimes, you just have to put that sales pitch away!
5. The product is right, it’s the packaging that’s wrong
Many organisations focus on getting the right product or service defined, and the channels to deliver it as the foundation of next best conversation strategy. Whilst these are critical components, what is often overlooked is the suitability of different incentives attached to the offer. Think of this example: if you’re a general insurance provider and a customer with a vehicle only cover makes contact, the next best product conversation inevitably gravitates to a home and content cross-sell. Seems logical, but it’s often how the cross-sell is wrapped up that really counts. Do you lead with a discount offer (e.g. 25% discount on premiums for the first 3 months), a value led offer (e.g. $500 home security kit for free, or a related benefit (e.g. basic pet insurance included). A savvy marketer would say test it, and that’s correct, but all you’re going to identify is the best offer overall. The organisations we see doing next best conversation right, in an automated fashion, have decisioning logic in place that ensures the right incentive (based on the types of offers the customer has responded to in the past) is matched with the right product to form the basis of the next (best) conversation.
In conclusion, I’m sure your next best conversation program is a winner, but, just for a second, ask yourself, could it be better?