In a nutshell, this project was about creating a mortgage recommendation via the dashboard and discontinue doing it via Intercom.
A customer used to receive a recommendation from Habito via an email which a mortgage expert (ME) sent out using the back office sourcing tool.
This email contained automatically generated documents.
One is the ESIS and the other is the EoR.
The documents essentially
explains the mortgage product recommendation to the customer and the research the mortgage expert has done to reach this recommendation.
The customer would either reply to the email or go to the Habito intercom chat and reply over intercom asking for another recommendation if they weren't happy with the current one, or told the mortgage expert that they wanted to accept and continue with the recommendation.
1- It’s not a great customer experience.
2- It’s time-consuming for both parties.
3- We don’t have any data to know when a customer accepts a recommendation as it’s just free text in intercom. It’ll be great to have this data as the customer would be accepting the recommendation.
4- The user has to leave the Habito website.
5- We are constrained by an email to publicise our products like Habito Go and Marketplace.
Create a digital experience which allows the mortgage expert to show one or two mortgage products to the customer via the Habito website.
The customer should be able to interact with the product(s) and understand the recommendation. Finally, the customer should be able to accept a recommendation.
This is a vital part of the journey for the customer, it’s essentially when they get their biggest value from Habito; a free, personalised to them mortgage recommendation.
We should try and make this a big deal and a celebration for the customer, because it is!
1- Must show the customer either one or two mortgage recommendations.
2- Must show the customer how we’ve calculated the recommendation.
3- Must allow the customer to download their ESIS and EoR documents.
4- Must allow the customer to select/accept their recommendation.
5- Must work on mobile, tablet and desktop.
6- Must be within brand/design guidelines.
We still want to be perceived to be providing the customer with a personalised mortgage recommendation which the mortgage expert has worked on and tailored to that particular customer.
We do not want to appear to be a comparison website with a whole list of options/products.
How can the engineers build this iteratively?
Measuring success with data will be difficult as we do not have anything to compare this to currently.
We will measure success based on accepted recommendations.
We can try and compare the conversion of accepted recommendation by asking data science to look into intercom data before this goes live.
The introduction of Habito Go and Marketplace into the journey.
Every week I would get key stakeholders in the same room and show the latest designs updates. These session were normally consisted by me presenting the designs and everyone in the room would note their thoughts/suggestions into four buckets: Likes, Questions, Challenges and Ideas.
We used PingPong to run our remote testing sessions.
Testing was run by one of our research colleagues whilst the rest of the team were in a different room taking notes.
We had 6 participants testing our concepts split between 2 days.
The results were clear, giving customers two recommendations was confusing. We also learned that people would like to see a full list of rejected mortgages and why they've been rejected by the ME.
This feature has been developed very recently.
We don't have enough data yet to understand the real impact of it, however, as soon as I am informed about the success or failure of it, I will post it.
Fortunately I was surrounded by and incredibly talented team, who supported me in creating the best experience, to which I am hugely grateful.
