What impact will AI have on client onboarding?

Client onboarding is one of the biggest challenges for firms. You can pour a lot of time into trying to fix onboarding and yet it never seems to be exactly how you want it. 

There are a couple of reasons why. Onboarding is a complex problem that involves:   

• Copying or syncing data into multiple systems. This is time consuming and error-prone 

• Carrying out internal compliance tasks (ethical letter, ID checks, registering with ATO/IRD) 

• Convincing the client to send over a large amount of information (and following up, and following up) 

• Maintaining a good client experience (i.e. not asking for the same information twice) 

There are several approaches to client onboarding, each with pros and cons. None of them have a complete approach but some are better than others. 

Options involve exploring features within your practice management (some suites have invested more here than others); using a dedicated onboarding app, a very new category for accounting firms; turning to a data aggregator or integration platform to unify your data; or, if you have deep enough pockets, building your own solution using a centralised database or data lake. 

In our Agents of Change workshop on client onboarding we divide tech options into three categories: 

Baseline functionality 

• Manual processes 

• Back and forth with clients 

• Not standardised 

• No single view of clients across service lines 

Modern  

• Client collaboration  

• Auto filing docs  

• Agent Link Online (AU) and IRD (NZ) 

• ID verification  

• Workflow monitoring 

Visionary  

• Enter information once into all systems 

• Surface all information possible from existing documents, correspondence and data repositories  

• Integrations and automation  

• Single client view across service lines 

AI and client onboarding 

There is limited use of AI right now in client onboarding. Some practice management software is using generative AI to help draft and summarise emails, or using AI to read a document sent by the client. 

The most likely area of development will be through agents. We looked recently at how to use ChatGPT with task automation platform Zapier. For firms building their own onboarding process, they can use Zapier to get ChatGPT to function more as an agent and carry out a set of onboarding tasks with minimal input.  

The big advantage is that instead of bringing documents and text into an AI chatbot, the AI can appear within the workflow. The release of Custom GPTs (covered in this week’s training) will also encourage firms to invest time this area. The GPT can decide that something needs to be sent to the client, or the client needs to be set up in a particular system, and then carry out that task.  

In the future, we will likely see AI decide what we need to ask the client, where it needs to be filed and whether the client has sent the information or not. 

In the December Agents of Change workshop, we looked at:  

• the pros and cons of each approach 

• What the applications (seven reviewed!) do 

• And how firms are building their own 

We came to some conclusions about the best approach for firms of different sizes.   

If you are looking at onboarding for your firm, you can watch the full research through the +Training plan, available here.