Promo Bot
SITUATION
As a UX specialist in backoffice product suite, a part of my responsibility is to give ideas on product evolution. I decided to focus on enhancing app efficiency and user productivity in creating promotions.
TASK
I have overtime collected user studies (such as Contextual Inquiry and User Interview) documentation that highlights pain points identified by end-users for all of our enterprise products. These documents ensure that users voices are represented during design phases of projects. From these documents I gathered insights that users value task completion time to be faster.
action
I took insights from the user study documents to prototype a new product that automates offer setup with minimal input from the user.
Various operator studies reveal that end-users frequently set up fixed promo offers on a 2 weekly basis.
Frequent offer set up is time consuming - It requires users to complete a wizard that is time consuming for routine tasks.
No way to predict offer appeal - There is no intelligence to predict successful offers for a specific market.
Offer can be customized down to a player segment - It is not manually possible to create personlized offers for each player
Resolution
Allows user to enter basic parameters into a chat bot, which will then generate complete promotional offers, eliminating the need for manual setup and saving time.
Setting up frequent offers: Multiple user studies reveal that operators setup 5-8 different but consistent offer types on a regular basis. They re-use these offers in rotation to keep players engaged. These offers have been identified as lucrative by their Business Intelligence teams. If we use these offers as templates and automate their creation without having to refill the entire form again and again, we save a lot of time. Chat Bot leverages predefined offer templates and recommendations thereby automating routine offer set-ups.
Predict offer appeal: We currently don;t have any data insights to know how an offer is performing. This is a new but extremely valuable feature that needs to be added to our products. Currently this is done through operator BI team algorithms. If we can plug those algos onto our apps or create our own engine to do that, we could learn and predict offer performance. The bot aims to use data insights to recommend best offers for specific markets, reduces manual errors while setting up offers, personalize a seasonal offer at a player level, and feedback offer response into data insights.
Offer can be customized down to a player segment: Through the use of Machine Learning we could eventually be able to monitor offer performance per player and be able to personalize an operator created offer mechanic with personalized numbers that appeal to the specific player.
result
It was well received by stakeholder and on the shelf as a future product.