Sunset Kustomer and intergrate Zendesk into our system

  1. Allow for customers to self-services
  2. Reduce time and cost of interaction
  3. Unify help center system across companies

This is so our CX can scale with us, without any problems.


Kustomer was not scale at the rate we where and wasn’t able to cover all the features we need


Knowledge Base (KB) team + Customer Experience (CX) team + a myriad other teams all the was from app to webflow.


Working with KB + CX team to understand their goals, requirement and pit-falls for Kustomer.


Started with collecting all requirements and information in one place


Design process

From UX → UI → Copy



I collected the coming UI2022 styling for reference and created components.


Figured out the customer flow and a simple structure. Took it to crit for feedback.


Once the UX was done, turned our focus to UI and copy. Working closely with CX & KB teams to fill in the lo-fi designs with some real content.


Finally reviewed it with the stakeholders and off we went building it.


This is the final outcome, and this is what we'd rolled out to all our help centres.


⚾️ Curve ball: engineering draught

All our engineers where busy working on feature releases and we didn’t have engineering support from Zendesk, as it was not accounted for in the orginal agreement. So I offered to fill that gap. Zendesk uses this lanuage called HandlebarsJS, fairly simple to be honest but I learnt that and went about implementing the design.


With sign off from all parties; I rolled out this template to our sister companies, matching it with their respective look and feel. This project was deliveried 2 month ahead of schedule and under budget as I ran it end-to-end in respect to design and development.

Launched on the 24th Jan, it went really smoothly and was recieved incredibly well across the company. Click logos to see live sites.

pollen logopollen logoCampus vacation logo


44% of conversations started on Polly (chat widget) were deflected

72.5% were marked as solved with the first reply.

First median reply time on chat down by 32%

4x more metrics to work with; higher level of insights