System Design Chatbot UX Rethink (Case Study)

Challenge: To improve overall service design and user experience of an online website contact and FAQ area, and avoid robotising humanity

The Background – The Problem With Relying On Chatbots

It’s easy to think technology can solve communication channels, considering how easily we can chat to someone on social media these days, but new bells and whistles can also create the ‘rabbit in headlights’ approach of ‘ooh, what’s that, let’s try it’.

This has become the case with lots of online companies these days who rely too much on chatbots to ‘answer’ questions of users and potential customers.

Here’s the usual situation. You need to get through to someone, a human, but all channels on the website are directed towards email or chatbox FAQs. Modern websites think this is a good solution as it saves them labour costs on having call centres (who generally can only deal with the bare minimum of enquiries anyway).

The chatbox experience assumes you have all the time in the world to read through every |FAQ and find out your solution, but most likely you still need to speak to someone anyway regarding a delivery, sale or technical aspect you might be having.

It can end up posing like it’s a good user experience to be ‘directed to the right place’ but more often than not it’s just a frustrating waste of time as all you really want is a number or a direct contact to solve a problem, without having to spend ages either waiting on an answer or reading through everything there is in FAQs and still find you are getting nowhere.

You can understand why this system came to be. Technology allowed it and previous call centre and phone responses were even more inefficient.

This project needed to take a much deeper look at the service and system design elements within the company. The company (not named for requested privacy reasons) weren’t alone in their approach, as they genuinely thought they were improving their service, only to find when we looked into testing that experience that many lost sales were because potential customers were fed up of dealing with the very robotic sales tactics.

When we looked at some of the staff employed by the company most were tech engineers, and few were UX designers or customer service reps. The general consensus was that robots can deal with that and make the user experience better, and give us more resources to concentrate on improving the technology side of ‘experience’.

Researching Better Ways To Improve System Design & Human Communication

Technology may be improving, and AI might ‘read’ what humans want, but only humans are able to truly bring that human experience, so we needed to flip a switch on technology ‘aiding’ humans rather than humans becoming assistants to technology.

We did this by concentrating on two elements. What ‘humans’ said they wanted, but also what ‘humans’ would need that they might not realise they needed just yet.

You see, when the initial focus groups and user testing was done on the chatbot idea the company good a lot of positive feedback, but what they didn’t realise at the time was that their demographic of testers were generally young, tech orientated people, and that with any new toy it soon loses it’s glossy appeal quickly after (as you then see the actual functionality that lusting blindness can make you miss at first).

That’s the reason we focused on a better system design AND service design, as only focusing on system design might bring about a soulless solution, and only focusing on service design might not see what a user really needed beyond the appeal of the technology and that false experience.

We looked into what humans wanted first in terms of a humanised experience, and many wanted to feel like there was a channel to easily and quickly speak to someone, a channel where they didn’t mind giving feedback as long as it was listened to, and they felt the live representative would be there to speak to when needed.

We looked at a similar example of a well-directed user experience in Gophr, and could see that whilst the website interaction could be improved what stood out was the ease to order a delivery and be linked to a ‘human representative’ who would be able to deliver that package directly to you (or to where you want it to go).

When looking at other delivery services there were tracking options and apps, but no way to actually speak to a human for an update or a change of plan etc. It left people waiting all day at home, or within a long time slot, and sometimes these were cancelled leaving people to be waiting and simply have their time wasted, and then be responsible for having to rearrange another delivery that they now couldn’t trust.

With Gophr people didn’t mind the higher prices of same-day delivery so long as they felt it could be humanised and trusted that their package would be delivered, and if not there was a simple way to communicate directly with the driver or the company.

In terms of having a human connection Gophr provided a better solution, but the more we researched into tech chatbots and reliant communication the more we found a broken system.

Let’s take other tech-reliant situations as examples of creating a frustrating user experience because of a broken system that only put more strain on the customer and the sales staff because they rely on tech too much.

If you’ve ever been locked out of your account because of a password error (whether your fault or not) it can be very frustrating having to try and prove who you are who you say you are, and then have to go through thousands of other steps to just try and get back into your account. Or maybe you have password confusion that is taking you between two accounts because the system is getting confused. Or maybe you have to keep repeating your sales return or tech issue to different people because there’s never a dedicated person to deal with it, so technology just relays it to the next guy who asks you what the issue is again. Or maybe you’ve requested something to be done but because of a poor system in place, someone forgets to do it (often because of a miscommunication somewhere down the line of passing one middleman to the next and so forth), so then you have to keep reminding them but they’ve done the same for other customers and now there’s a backlog of work on their part, and a lot of wasted energy both ways.

Well, these are all examples of inefficient systems that use technology at their core. It’s part of a value extraction mindset that has been drilled into companies – not trying to actually help customers but to trap them in instead – and it just doesn’t work for the 21st-century world we live in today.

Broken systems tend to run on either old comfort-led ideals that don’t fit a changing world, or on new tech innovations that forget the human touch and turn humans into more robotic assistants than thinkers.

“The more we rely on technology the more we have potential stumbling blocks in communication”

A Solution To Fixing Broken System Design & Technology Reliance

We wanted to change this broken system for our client, and we knew we had to find a more humanised approach to make it work.

Knowing the company was in the tech industry there was going to be friction in suggesting less tech, but we demonstrated the feedback from our research.

The solution was to adopt a people-first approach to how their customers could go through both the online and the offline (omni-channel) process.

Online users would have a chat button in the bottom right corner, but one that led to human responders, but the aim wasn’t to ask a question and then wait for a response in a robotic way (where reps use the usually trained ‘hey (enter name), thanks for reaching out, what can we help you with today’ automated way of responding), but instead the rep could receive an issue and get to work on it straight away, without delay, and no wasted time in between – what we call ‘lost space’.

Lost space is a big problem in chatbots and live chat communication channels. It’s that time waiting for the responder to come back to you, only to say ‘hello’, then for you to say what your issue is, and then to wait an age again to find they aren’t looking into it but they are going to say something like ‘how are you today’. It’s not the efficient response people would expect from live chats.

Instead, if users knew that someone was working on their issue, to help them, they would be less stressed and more likely to get on with something else, knowing their issue is being dealt with. A lot of ‘lost space’ on telephone calls is because you are waiting for a service rep to answer only for them not to be trained in enough areas, or for the conversations to lead into ‘have you checked our website’.

There’s nothing more frustrating than checking a website that says call us, and then to call them and be told to check their website. It’s a groundhog day loop.

Therefore, we set up a ‘solution resolution’ 3-step process. If the issues weren’t dealt with by the time it found step 3 then it would be escalated, and only then. But this step process would happen behind the scenes, so the end-users wouldn’t know the efforts going on behind the scenes, or be told ‘we are still working on it’, they don’t care about that, they just want the issue sorted.

Instead, they are told we are working on it now from the start, and only contacted back when more info might be needed, or the issue is fixed, plus there’s a direct personal rep number that comes available if step 2 has been reached (most issues are actually sorted in step 1 without further need of direct communication).

The last thing people want is to be repeating issues to different people, so this approach ensures they only have to explain the issue once, at the start, and then get on with their day.

The main aim was then to ensure that the systems created would ensure this journey was possible. It means minimizing the overload of systems that just created needless loops and lost space, and instead focused on a simplified step process where each person in the process would easily know what they had to do. It would cut out as many middlemen as possible but ensure that when communication did have to occur there was a human available to send that issue into action.

Why The Future Needs To Be Chatbot Free

If this case study isn’t enough to demonstrate the failings of relying on chatbots today, then consider the future implications.

The main reason humans hate chatbots is because they generally aren’t sophisticated enough to mimic natural human responses, but if we look further it’s also because humans are trained to mimic chatbots.

Communication is meant to be about code-breaking, problem-solving, and growing in thoughts, words and knowledge. It’s not meant to be the opposite, having predefined answers to treat every problem with the same generic response.

They can end up dumbing us humans down and just lead to us becoming more frustrated at not being heard properly, and then leading us to areas of a website that we don’t want to go, plus it stops our very human ability to seek to solve problems as we become reliant on a robot that is essentially dumber than we are.

The engineering tech argument to that is that they will become more intelligent in time, but do we honestly want that?

There’s a ton of data-inputting jobs out there today that are basically aiming to train AI language bots to become more sophisticated, but this simply takes us further away from the human approach, not to mention takes jobs away from humans.

Of course, there are certain jobs that AI can do better than humans, mainly generic, repetitive ones, but communication is not, and should never be, one of them.

Bots could possibly help you with spelling errors when writing content (like Grammarly), but they don’t replace human contact.

In a future that is full of chatbots we could lose our greatest ability as humans, to create and problem-solve, as we just rely on systematic robots to think for us. No thanks.

A CASE STUDY FROM: