Predictive Hire
- Former user (Deleted)
- Janica Daryanani (Unlicensed)
Summary
IT STARTED WITH AN OBSTACLE
Before founding PredictiveHire, Paul and Jonathan ran their own respective companies and were frustrated by the difficulty in identifying quality candidates who would perform well in their unique business environments. After trying multiple assessment tools, they felt they could further improve performance and reduce staff turnover. They searched for a product that could scientifically predict the success of new hires.
There wasn’t one. So, PredictiveHire was born.
WHO ARE WE?
PredictiveHire is a team of data scientists, organisational psychologists and recruitment experts who’ve joined together on a mission to change the way recruitment is done, by leveraging the power of artificial intelligence. We believe all people decisions should be based on data and analytics – not gut feeling. With the help of technology, we want to level the candidate playing field and help reduce unconscious biases in the hiring process.
Offices in the US, UK and Australia.
The Technology
YOUR DATA
All of our predictive models are built on data from your organisation and are centred on the business issues and corresponding key performance indicators (KPIs) you wish to address. Once you have identified the KPIs that matter most to your business, we gather relevant performance data and couple it with data collected via our Questionnaire Experience (QX) deployed to your existing workforce.
This ensures that later, when a candidate completes the QX, the resulting performance prediction is relevant to your business – not the world – and will enable greater organisational efficiency and better results over time.
OUR SCIENCE
We use artificial intelligence (AI) technologies to build our models.
The AI consists of a neural network that is used to identify what drives performance in your organisation based on the data we collect from you.
We build our algorithms to achieve very accurate predictions from the start, and the model improves over time through machine learning.
Additionally, we use validated behavioural science when developing our algorithms. This ensures maximum predictive accuracy while prioritising the candidate experience.
Technology Demo Notes
6.20.2017 Call Notes
Kick off for internal pilot
From Wilson: Erin McGaughey, Cynthia Cancio, Marisol Hughes, Andrea Bowen
From vendor: Jonathan Nicholson, Steven John, Gary
Our Data + Their Science = predictive AI model
They’ll deploy questions to our workforce from those who are our superstars, our middle “steady” performers, those who aren’t where we’d want them to be..
Then will ask someone here (whoever it will be), to identify who is who in terms of buckets
Create an algorithm based on results and our indications of responders
Ask applicants the same questions to assess cultural fit based on our model
Their process will remove reliance on resumes, outside of required certifications, education, etc.
Look for the person to identify who the people are that belong in which bucket
We have goals, top performers, awards, etc.
What data in colleagues do they need? Way to identify individual - can be a name, email address, candidate ID - they would prefer ID for anonymity when someone is answering questions, but they can use any of those things. Then would need to be able to connect the responses to the appraisal scheme
They want everyone in the recruitment consultant role, and then bucket them into the 3 tiers; then typically get 80-85% completion rate
From all of those colleagues who respond, would want us to bucket them
Shouldn't take most people to spend more than 10 minutes to complete questions
They suggest a week to ask everyone to get it completed, then allocate a 2nd week to ‘prod’ anyone who hasn’t completed it to get it done. And that ends the employee time spent on the process
Would want to consider any other time requests the business is asking of them at that time (surveys, etc.)
Go live:
USA / Canada
Rest of world will be separate
Training: where it belongs, how to interpret in to process
They requested style guide; Marisol said will get from marketing and other from past for style They’l then send us a few mock ups for us to choose from, with comments/feedback before deployed into live environment
They’d like to have it before being sent out; before 10th of July if that’s the date we’re asking our team to do the questions.. So ideally, want it by July 3rd or sooner.
Who to follow up with?
Related articles