Digital Disruption 2017 just wrapped up. If any session struck your fancy but you just couldn't make it, look no further than this recap.
Table of Contents:
- Maverick Rules for Leaders: Unleashing Human Potential and Building an Adaptability Advantage: Polly LaBarre, Author: Mavericks at Work
- HireVue Innovation: Past, Present, and Future: Loren Larsen, CTO, HireVue
- Uncharted Waters: The Journey to an Automated Recruiting Process: David Curry, Commonwealth Bank of Australia; Nathan Mondragon, HireVue
- In Tech We Trust: How Trust-Based Hiring Landed IBM the Top 4%: Sherri Blaszkiewicz, IBM
- HireVue Assessments: How Enhancing Talent Decisions with Video Intelligence Drives the Bottom Line: Josh Liff, HireVue; Allie Wehling, HireVue
- Trade in Poor Performance for Innovative Training Strategies & Exceptional Time to Fill: John Perez, Sonic Automotive
- Leading a Switch: Dan Heath, Co-Author: Switch
Maverick Rules for Leaders: Unleashing Human Potential and Building an Adaptability Advantage
Polly LaBarre, Author: Mavericks at Work
Polly LaBarre kicked off our second day of Digital Disruption 2017 with a bang. She tackled the changing nature of the workforce: namely, how we make human endeavor more productive, fulfilling, and creative.
According to Gallup, only 13% of the workforce is fully engaged - and this number hasn’t budged in years. Bureaucracy, a system designed to organize the non-routine through structure, clearly requires a change.
Where in decades past the primary business challenge lay in removing waste and increasing inefficiency, today the challenge for every organization is to become more adaptable, innovative and inspiring. While the pursuit of productivity requires people to emulate machines, the pursuit of innovation requires people to emulate people - embracing the traits that make us uniquely human.
Efficiency principles just don’t fit with today’s business challenges. According to LaBarre, we need new principles, principles that preserve efficiency without sacrificing innovation and creativity. These are some of the new principles LaBarre identified:
She then outlined three core strategies for embedding these “human” principles in your organization:
- Lead Without Authority. Leaders should assume that they have no power or control over anything. They should take their leadership position as a starting point and inspire their employees to do things.
- Embrace Experimentation. Expect people to start and try new things based on evidence in the field. Amazon runs thousands of “web labs” to test new ideas: for every AWS there are 10 Fire Phones.
- Change How You Change. 70% of organizations that attempt big change fail on a grand scale. Focus on change platforms (not programs) that enable internal activists to act as agents of change.
HireVue Innovation: Past, Present, & Future
Loren Larsen, CTO, HireVue
Loren Larsen started his presentation with a brief history of HireVue. Founded in 2005, HireVue envisioned using emerging webcam technology just as phone screens were taking hold in the recruiting space.
Now, in 2017, video interviews are in the same position as phone screens in 2005: they’re starting to become the go-to way to evaluate candidates.
Video screening has numerous advantages, largely because human intuition is prone to flaws. Larsen provided the example of a study where patients were given two different pills for their pain: a pill costing $2.50 and a pill costing $0.01. Patients across the board rated the more expensive pill as more effective than the cheap one - despite the fact that they were both placebo.
We see the same sort of thing when making talent decisions. Does salary history really correlate to job performance? Not really - but we assume that well-paid people are high performing.
Now with HireVue’s incorporation of pre-hire assessments with artificial intelligence (AI), those decisions can be augmented with an unbiased approach.
Larsen closed with an overview of new features coming for HireVue, namely the incorporation of third party assessments into the platform. Once this feature launches, HireVue customers will be able to view candidates’ assessment performance within a single, central hub.
Uncharted Waters: The Journey to an Automated Recruiting Process
David Curry, Commonwealth Bank of Australia; Nathan Mondragon, HireVue
Nathan Mondragon started the session with a “deep dive” into HireVue Assessments.
“Time to hire is only half the solution,” Mondragon explained. “If we got you bad hires quickly, we didn’t do our job.”
HireVue Assessments use AI to extract data from a video interview. There are three sources of data it pulls from:
- Video data (the nonverbal communication a candidate uses).
- Audio data (how the candidate said what they said).
- Textual data (what the candidate said).
Overall, the HireVue algorithm pulls over one million data points from a video interview, evaluating it frame by frame. Around 25,000 of these data points are found to be relevant to performance metrics - HireVue Assessments are evaluate candidates based on these.
Mondragon concluded with a few ROIs experienced by customers who have rolled out Assessments:
- 70% more likely to be promoted.
- 31% increase in tenure (with an associated $2.8m in savings).
- 30% increased upsell for sales roles.
David Curry followed with an overview of Commonwealth Bank of Australia’s Project FLARE: an ambitious undertaking that seeks to automate the entirety of the recruiting process.
He started with this statistic: 81% of CBA candidates are customers. It is critical for Commonwealth Bank to provide an experience that keeps them that way.
He proceeded to outline a few key points regarding CBA’s recruiting:
- CBA receives 300k applications each year.
- Nowadays, front-line employees are required to have more direct and in-depth knowledge.
- CBA has performed 40,000 HireVue interviews over the past two years.
The above points made CBA the perfect candidate for HireVue Assessments. Now, Commonwealth Bank is looking to create a better experience for candidates with an all-online, mobile experience that shows candidates a realistic preview of the job and provides real-time feedback throughout the process.
So far, things are looking good. They’ve cut their 10 day application process, which consisted of multiple rounds of assessments, down to 75 minutes. And recruiters love the insight provided by the HireVue Assessment.
When comparing the decisions a recruiter would make to those made by the HireVue Assessment, CBA recruiters say that “the algorithm’s basically right every time.” In other words, the assessment augments the recruiter’s ability to identify the best talent - it does not replace it.
Q: Have you had any experience with the recruiting team on how the AI’s working and what their thoughts are?
A: The teams that are now operating through this process are seeing the benefit and seeing the value it is creating for the business - while there was some sweating initially, they see the value in terms of the opportunities that are created through the new process.
Q: What roadblocks did you hit?
A: We’ve never made big grand promises, we set the plan out as a path we’re going to explore. This really helped when selling our vision - we made data-driven decisions to support every choice we made.
In Tech We Trust: How Trust-Based Hiring Landed IBM the Top 4%
Sherri Blaszkiewicz, IBM
Trust-based hiring is trusting that candidates with the skills (not the experience or educational pedigree) will succeed. In this session, IBM presented how this focus on skills streamlined their hiring process for new graduates and landed them the top 4% of their applicant pool.
Prior to streamlining their hiring process, if a college student wanted to apply to IBM, they would need to apply to 150 different job requisitions. Now, IBM has only 5 job postings - and candidates that apply are considered for a huge range of openings in the company.
IBM’s Trust-Based Hiring Process
- Apply. With only 5 job postings for 150 different job openings, IBM removed the hassle from applying. They do not use resume or GPA screening.
- Assess. This involves an on demand assessment consisting of coding and a video response explaining their thought process. Candidates are auto-scored: top performers (42%) move forward automatically, average scorers (27%) continue to a manual screen of their code, bottom scorers screened out (31%).
- Interview (Three Steps).
- Live technical whiteboarding with a technical expert evaluating their code in real time.
- Live “fit” interview with a hiring manager. This is used to find out what candidates want in terms of location and job opportunities.
- In-person, 2-day “finish line” event. During this event candidates visit IBM’s Cognitive Expo and see what it’s like to work at IBM.
Who Was Involved?
To gain buy-in, IBM engaged with technical leaders at all levels of the organization. Dubbed “Cognitos”, these technical leaders helped put together the new hiring process. “Cognitos” consisted of:
- Scouts (early professionals). These visited campuses to teach grads about available opportunities, and also performed manual screens.
- Gurus (experienced professionals). These were hiring managers in charge of leading the live whiteboarding session.
- Guides (senior leaders). These were senior leaders who performed the exploratory “fit” interviews
- All in all, IBM processed 15,000 applications, 10,000 on demands coding assessments, 2,000 live interviews, had 700 finish line attendees, and made 660 offers.
- 83% offer acceptance rate for finish line attendees.
- Throughput is 30% higher than the previous process, so IBM is getting the most qualified candidates into the pipeline early on.
Advice from IBM
- Put yourself in the candidates’ shoes. Were they getting the right message and did they understand the process?
- Pilot on a subset of your organization. IBM piloted HireVue coding assessments on their Watson department.
- Enlist technical teams for coding questions and whiteboarding.
- Earn trust from hiring managers by educating them on the benefits of the screening process.
- Don’t underestimate problems of scale - for example: interview scheduling and demand tracking.
Q: Have you seen a difference in retention rate and performance?
A: Yes. This is a fairly new process, so we’re just starting to collect that data. So far, hires that took HireVue Coding Assessments were 20% more likely to get the highest performance rating in our internal performance assessment than those that did not.
Q: Any recommendations for scaling hiring from new grads to more experienced hires?
A: This trust-based hiring process is really well suited to early professional hires. These candidates don’t have a whole lot of specialized skills at this point - we’re screening for that basic expertise and potential. Experienced professional hiring is a little different - generally hiring teams are looking for a specific, narrow skill-set.
Q: Any candidate feedback that was positive or constructive?
A: We’re not going to lie - it’s a rigorous assessment. Some of these folks will be building breakthrough technology, and some people do think it is hard. But if you look at the percentages, 42% of those taking the first coding challenge moved on to the next step automatically.
HireVue Assessments: How Enhancing Talent Decisions with Video Intelligence Drives the Bottom Line
Josh Liff, HireVue; Allie Wehling, HireVue
The average time to fill is 63 days, and the best candidates are off the market in 10. In other words, your hiring process might take longer than your top candidates are on the market.
This means we need to get quicker at hiring, without sacrificing quality of hire. Unfortunately, increasing quality of hire normally means adding steps to the hiring process: asking candidates more questions and putting them through more hurdles.
Allie Wehling then presented the Trifecta of HireVue Assessment metrics:
- Time to fill - it goes down.
- Quality of hire - it goes up.
- Quality of experience - it goes up.
Candidates are most excited to work for you at the time they apply. Assessing them within that “application window” is critical to quickly bringing the best candidates on board.
HireVue Assessments are completely bespoke: they are built for each specific role’s specific KPIs. With deep learning, HireVue looks at three things:
- Voice to text transcription: what the candidates say.
- Audio file: how candidates say what they say.
- Video: looking at the microexpressions made by candidates as they say what they say.
All in all, there are 4000x more features (data points) available than a 300 question assessment.
Machine Learning in Action
With machine learning, HireVue looks for the stable characteristics that differentiate high performers from lower - ignoring irrelevant characteristics like twitches or attractiveness.
- 50% reduction in safety violations: by filtering out the bottom third of candidates with questions that focus on the interpretation of mundane events.
- 76% increase in tenure: by filtering out bottom third of scorers who were twice as likely to leave.
- 58% termed employee avoidance: by filtering out the bottom third of scorers.
- 30% upsell increase: by hiring the top third and filtering out the bottom third.
Q: How important are the assessment questions to the output provided by the algorithm?
A: It’s kind of a continuum. We’ve found that even with general past behavior questions there’s still some predictive value - it’s really case-by-case, it all depends on what you want to get from the job. Even little tweaks to questions can have positive effects on the predictive value.
Q: What about if an acquisition happens? Do you need a new validation study?
A: We’d definitely do a bit of exploring. We might perform a mini-validation study to see if the KPIs and other indicators are still the same. If the job significantly changes, you’ll probably need a new study.
Trade in Poor Performance for Innovative Training Strategies & Exceptional Time to Fill
John Perez, Sonic Automotive
Sonic Automotive operates over 100 dealerships nationwide. As one of the largest automotive dealers in America, they deal with challenges endemic to many car dealerships. In this session, John Perez covered the challenges Sonic faced, how they confronted those challenges, and the results they’ve seen so far.
- Changing the way a traditional car dealership does things is tough. Most have had the same sales process since the 1940s.
- There was no set hiring process: all hiring was up to individual hiring managers.
- Car dealerships have a reputation for attracting low character people.
- Hiring managers tend to trust experience more than science, even when they’re experiencing 60% turnover.
- 35% turnover in management, year over year.
- Hiring managers spend 70% of time selling cars or service and 25% of their time on reports to the manufacturer. That left only 5% of the week for hiring.
Implementing a new process required training and driving engagement with new technology.
Blended Training & Engagement
You can’t reach your people with one channel, you need multiple channels. These are the learning channels Sonic utilized when implementing HireVue:
- Instructor Led Training. This is training that is self-based and built around key learning objectives. The training is capped with a skills test, and solidified with “job aids” placed around the office and demonstrations.
- eLearning. This learning uses HireVue Coach to train hiring managers on the HireVue platform. It also lets them know how easy HireVue makes the process for candidates.
- Getting everyone involved in rollout & adoption. Those who really enjoyed using the product acted as internal evangelists for it.
Engagement from Top to Bottom
You can’t get anywhere without senior leaders believing in what you’re doing - so Perez first showed HireVue to his operations executives and the CFO.
The EVP of Operations was so excited, he recorded an intro video which is now presented before every HireVue interview. Store managers are also empowered to record themselves asking interview questions for manager candidates while store personnel do the same for entry level candidates.
Sonic Automotive’s Results
- Cut time to fill for sales requisitions from 75 days to 50 days.
- Assuming an unfilled sales role results in $500 profit lost per day, the decreased time to hire reflects $10,000,000 in gained profit potential.
Q: How did this impact your NPS?
A: We sit at an 80 NPS, despite being a car company.
Q: What do you feel excited about with piloting HireVue Assessments?
A: We have a lot of money tied up in our assessments. If we have both our old assessment and HireVue, we give our hiring managers the opportunity to choose the one they like - ideally this will help them look at the video interview. The more I can give hiring managers to speed up the process, the better.
Leading a Switch
Dan Heath, Co-Author: Switch
If it’s true that people have a natural resistance to change, we certainly don’t show it. We embrace new tech, new fashion, and other new trends with ease. But for things like diets, making a change is incredibly difficult. It’s pretty clear that there are some changes that are easy - and others that are hard.
Dan Heath presented on how we can make the “difficult” change we all hear about like the “easy” change we embrace without thinking.
According to psychologists, there are two different systems in the brain responsible for this difference: the rational system and the emotional system. Some change is difficult because we can’t even agree in our own brains on what we should be doing - and it’s not a fair fight. The emotional system is far more powerful when it comes to decision-making, but the rational part of the brain believes it is still in charge.
But the emotional system isn’t all bad. The curiosity and desire the emotional system creates is responsible for innovation.
With this in mind, Heath provided a three part framework for reconciling these two parts of the brain and direct change:
- Directing the Rational System. Get rid of ambiguity. When there is no clarity, people create excuses for not embracing “difficult” change. After change begins, highlight the “bright spots” that come about as its result. HR leaders are best positioned to be these bright spot champions - don’t just celebrate success, study it.
- Motivating the Emotional System. Change begins when the emotional system is ready. Don’t try to motivate people with information - the emotional system doesn’t care. Rather than sharing a solution proposal, you need to dramatize the problem. Then expose stakeholders to the future: don’t just talk about it.
- Shape the Path. Make the right behavior the path of least resistance. Take into account situational influences when building a path for change - don’t blame people’s character when they resist it.
Change may not come quickly or easily. But when it happens, it tends to follow a similar path. Reconciling the rational and emotional systems is key to building a path for long-term change.
“To really implement change, we need to give ourselves permission to fail,” Heath concluded. "But the evidence is clear that great change happens all the time. By following the same path as those with success, you put yourself in the position to be the author of change, rather than the recipient of it."