top of page

Omni-channel Series #3: What does a perfect Omni-channel model look like?

So, in our earlier articles in this series we looked at why you should adopt an omni-channel customer experience and 5 strong use cases for omni-channel analytics. Now let’s take a look at what an effective omni-channel operating model should and shouldn’t look like.

Let’s start with what an omni-channel experience SHOULDN’T look like, most likely commonplace in the majority of organisations. We call it the hourglass shape and hopefully you can see why (see fig below).

What we see is a fat self-serve at the bottom which is good as it’s the cheapest mode. Chatbots are being used to field a lot this section off. But then live chat is not being used as strategically as it could be as the next level up. All too often customers end up coming back to use Voice as a channel, often through channel failure or switching which is expensive and bad for customer experience. In this model everything can be seen as trying to deflect from voice and First-Call resolution (FCR), as opposed to being a joined up strategy.

Even though Voice is the most complex and the most expensive channel, typically 70% of traffic is exactly that. Ideally Voice should deal mostly with the most complex queries.

Now let’s look at the utopia, how your omni-channel experience should look like.

– You would have the biggest number of transactions going through self service – Then information requests or simple questions would be handled via a chatbot so no need to have an expensive agent conversation – and its accurate and optimised! – You would offer a live chat provision to help with operational contact management (so concurrent customer conversations) where there is some complexity so requires an assisted channel but is handled efficiently – And finally, traditional telephony agent conversations. These would be customers seeking guidance on what product to buy or maybe highly emotional conversations requiring a human interaction

In this utopia state, every customer knows what channel to use for what transaction, every interaction is handled correctly first time, there is no failure and the world is lovely.

OK, so we now know the utopian vision that we should strive for but the reality is perhaps something like this, which we call more of a target state.

There will be an element of failure, represented in red, but this would be significantly reduced in a more joined up approach.

The start point for many organisations is essentially delivering the different channel options. And we know, a lot of organizations are still yet to go on the chatbot journey, and some don’t have live chat.

The aim is a more joined up approach, understanding what is failing within each channel, and having the sort of the insights and the analytics to help improve and reduce that non first contact resolution. This is a great step forward and what we should be striving to achieve.



bottom of page