Managing tensions between the Mobile teams and the Conversational AI teams

I was reflecting on the post I wrote about HSBC's hiring of their Senior Chatbot Manager in their Conversational Banking team.

The job listing highlighted the reporting structure thus:

This global WPB role will be a key member of the Global Conversational Banking team, reporting to the Global Head of Conversational Banking, in Group Customer Channels. The Conversational Banking team is globally accountable for our Chat, Messaging and IVR channels that deliver a fantastic customer experience through a combination of automated technology and agent support. 

All good.

Until you look closely.

Deliberately separating Conversational interfaces from, for example, Mobile and Online, is both an opportunity and a deep, deep challenge.

The operative word is separating.

In some organisations your team might be 'separated' but it's just a function of line management, with team members constantly collaborating across teams and disciplines.

In other organisations, human-made firewalls spring up quickly.

I've seen this recently in several high-profile financial organisations. For the longest time, Mobile has been the darling of many organisations - especially in banking, for example.

Teams and individuals have grown used to the constant affection and support from senior executives. Bonuses, adulation, and continual mentions in the quarterly and annual reports. Victory laps by the CEO, highlighting the huge amount of traffic flowing through the mobile apps. It's all been soooo good.

Until the new kid on the block arrives: Conversational AI.

In some organisations, they've managed this really carefully and really effectively.

In others, it's been an afterthought.

In one very high-profile bank, I saw a huge amount of issues forming the moment they deployed their Conversational AI capabilities - because customers started adopting it. They started using it. They really embraced it.

Within 2-3 months, the chatbot was receiving a huge volume of enquiries.

Very quickly it became clear customers wanted way more than the generic capabilities offered by version 1. They wanted the chatbot to start orchestrating stuff. They were expecting things to be 'done' by the chatbot.

But the chatbot didn't have access to the APIs it needed.

That's because the APIs all belonged to the Mobile team.

The Mobile team felt they'd spent years amassing all these APIs (with a lot of effort) and didn't think it was fair that someone else got access to them easily.

So they dragged their feet.

They made excuses.

They cited technical challenges.

Executives got frustrated and the Mobile team leaders read the tea leaves and decided to respond, slowly.

The Mobile team drip-fed the most basic APIs to the Chatbot team, always making sure to prioritise Mobile first.

Often you'll find the Chatbot can be enormously capable if it's got access to the right APIs. This can sometimes mean it needs better quality or more in-depth APIs than even the Mobile channel currently has. In one example I saw, when the Chatbot team asked for more capability, the requests were rebuffed and denied because they weren't on the Mobile team's roadmap. Their budget is for Mobile, not for Conversational AI. So by quietly denying, delaying or ignoring the Chatbot team's needs, they were doing the right thing. For their team.

Crazy. Isn't it?

It sounds crazy from the outside. But I'm sure we all recognise these behaviours and the concerns driving them (e.g. bonus, compensation, status, job security).

At another organisation I came across recently, the Conversational AI team was placed under the Head of Mobile. Why? Politics. The Head of Mobile successfully lobbied for it to be their responsibility since the Conversational AI was deployed in the mobile app. It also meant the Head of Mobile could argue for a pay increase given they had extra responsibility, more FTE and so on.

What can you do about these issues? How can you avoid these sorts of behaviours from happening? What can you do if you are experiencing issues like this?

The right plan is so nuanced because every organisation is different.

Here are some ideas that you might consider:

• Foster Cross-Team Collaboration: Encourage regular meetings and joint projects between the Conversational AI, Mobile, and other relevant teams to break down silos and promote a unified customer experience strategy.
• Align Incentives: Ensure that team performance metrics and bonuses are tied to overall customer satisfaction and efficiency (or other generalised 'team-wide' metrics) rather than just channel-specific metrics. This can help reduce competition between teams.
• Create a Unified API Strategy: Develop a company-wide (or department/division-wide) API strategy that ensures all customer-facing channels have appropriate access to necessary APIs. This may, for example, involve creating a dedicated API management team. Avoid encouraging or enabling the hoarding of APIs.
• Executive Sponsorship: Secure strong executive sponsorship for the Conversational AI initiative, ensuring it has the necessary resources and authority to integrate effectively with other channels. This might come from you, or it might need to come from another (set of) leader(s).
• Organisational Structure: Consider placing the Conversational AI team at the same hierarchical level as other channel teams (e.g., Mobile, Online) to ensure equal footing and decision-making power. Give some consideration to the compensation question too. Often there's a super-star Head of Mobile who suddenly has to compete with a newly arrived super-star Head of Conversational AI.
• Cross-Functional Governance: Establish a cross-functional governance committee to oversee channel strategy, ensuring that all channels (including Conversational AI) are developed in harmony. I know that an efficiency fairy dies every time someone mentions the phrase 'establish a committee' but I think there's some value in this step - or integrating it into an existing committee workflow.
• Skill Development: Invest in training and skill development for team members across all channels to foster understanding of Conversational AI capabilities and potential. Don't be surprised if most of your colleagues outside of the Conversational AI team assume it's going to be a useless service. Chatbots of yesteryear were really, really bad. So it's important to make sure you're enlightening everyone and showing just how effective next-generation systems can be.
• Customer-Centric Approach: Regularly gather and act on customer feedback about their channel preferences and experiences to guide development priorities. No surprise, the Mobile team will constantly tell you that the customers they speak to love the channel. The Call Centre team will tell you exactly the same. So a common feedback mechanism is important.
• Transparent Roadmapping: Create and share transparent roadmaps for all channels, encouraging teams to identify opportunities for synergy and collaboration. Try and avoid roadmap approaches.
• Cultural Change Management: Implement a change management program to shift organisational culture towards embracing innovation and cross-channel cooperation. Sometimes you need a lot of active involvement. It's rarely about the technology – that's the easy bit. The hard bit is the people.
• Regular Audits: Conduct regular audits of API access and usage across channels to identify and address any bottlenecks or inequities. Give some thought to asking a third party to come in and have a look (that's something I can help with – or your McKinsey partner would be delighted to assist too!)
• Pilot Projects: In some situations, it might be helpful to build awareness, credibility and buy-in of all teams by starting small with Conversational AI (although this can create problems if not managed correctly). One technique is to start small and develop a cross-functional pilot project to demonstrate the value of collaboration between Conversational AI and other channels. For example, it can be particularly effective if the whole team comes together to 'create' the first instance of the Conversational AI offering so that everyone feels a common ownership and responsibility.

What have I missed? What are your experiences?

Please comment below or send me a note with your feedback.