Anyway, let’s start this talk with a couple of questions.
Um, do you think you could ever
find a job that was customized for you.
Um, for your needs, for your skills,for your personality,
a job that would actually satisfyall of your expectations?
Do you think companies could get to know their employees so well
that they knew when they were not happy
and how to gratify them?
The answer today is probably no.
We just settle for what’s out there.
But facts speak for themselves.
Only 13% of global employees are truly committed to their jobs.
Do you imagine that 13%?
Turnover rate reaches up to 25%in several industries
and yearly associated costin the US
only reaches exceeds 500 million dollars.
Truth is companies still use outdated practices,old processes
and no tools nor relevant data
to respond to today’s demanding talent market.
Today’s time practices are executed by people like you and I
who take high-stakes decisions based on their own assumptions,
what is called an Unconscious Bias.
And most of the time reacting,
rather than proactively and strategically answering to talent needs,
which leads to rushed decisions
and to one third of new hires, quitting their jobs after only six months.
After study the data of thousands of candidates,
I started to develop an algorithm
capable of simplifying and quantifying elements
that companies believe were impossible to handle
or were only for mathematicians to know.
Let me show you a graph.
The first line the orange one shows
someone’s level of authority challenge.
This is the personality trait
linked to someone’s readiness through
challenged authority and traditional values.
As you may see, the line started quite low,
but raised as time passed.
The blue line shows someone’slevel of conscientiousness,
which is the personality trait linked to success
and according to several studies
it is the most important factor when it comes to employment retention.
What can we read from both?
Well, they show the in-all data of an employee
before he flew away.
As you may see, he became more challenging,more aggressive
as he became careless.
Now where does this data come from?
From test apply to these people? No.
People can cheat on tests and you probably know that.
Em, so I figured out that the way we express ourselves
usually show who we are and our state of mind.
We use language in different ways
and those differences reflect our personality.
Studies have shown that even though our choices are nconscious and spontaneous,
they do reflect who we are.
There is a direct association between keywords
and phrases and major aspects of personality.
For example, extroverts use lots of
fun related words, like music and party.
People with lower emotional intelligence
喜欢用负面词汇 比如生气 压力大
use negative words, like angry and stressed
and narcissists love to talk about themselves
and use lots of I, me and myself.
But it is not only words that I take into consideration.
The way we communicate also plays a major role.
a poor grammar shows lower academic education.
Absence of typos show perfectionism
and emoticons can be assigned of friendliness if the document is informal
or immaturity if it’s formal.
Long emails reflect energy.
Chaotic emails are a sign of creativity.
Instant responses show impulsivity
and no responses show a lack of interest.
So I took what most of us use,
most of the day work-related emails and chats
and I used the algorithm to help me
identify keywords and trends behind employees’ disengagement.
Was it possible to predict when someone as becoming detached?
And if so, how would it impact a company’s
retention and recruitment strategy?
And ultimately, how could we help the employees?
So far, the algorithm uses datacoming from the candidates,
individual answering trends.
So how long is he taking to responsecompared to before?
Is he or she working late as she used to?
Is he working over the weekends like last year?
It also takes information from the market.
So how is the marketplace impactingthat particular role
and ultimately it uses text mining to identifykeywords and phrases,
to then link them to a personality model that uses common language descriptors,
like the ones I showed you before.
Now when you mix all of this together
and over a period of time,
the algorithm learns how that person behaved and
it’s able to predict with a very high accuracy
when someone is becoming disengaged,
based on their behavior and personality.
Now, all those sounds really cool.
But my true purpose has been, to one,
help companies develop theirinternal business
and their external talent attraction.
And two, to make the recruitment process candidate friendly
by adapting it to each individual personality.
Evolving to a data and predictive approach,
well facilitate, recruit retention and recruitment.
And will allow anyone involved in talent acquisition,
老实说 本就该是公司所有人 而不仅仅是人力资源部门来干
which honestly should be everyone in the company,not only HR.
To know who’s going to be
the best performer in the short and long run.
And to customize the whole process
to each individual role and to eachindividual candidate.
Now, all of these sounds logic
and barry of lifting, at least to me.
But all of these raises questions about accuracy and privacy,
nevertheless, applying artificial intelligenceto the recruitment process
could insure more diverse and empatheticand dynamic workforces.
And according to studies, an algorithm could increase
the accuracy of selecting job candidatesby more than 50%.
Unlike how things are done nowadays,
elements like your professional background,
your social background, your cultural background,
your skills assessments will be analyzed to know
you’re the best candidate for the role today.
Because you will be able to
address the challenges associated to the job,
but also the best candidate for the companyin the future.
Making sure that the career path
also matches your history and aspirations.
And it is not only that algorithms
are able to probably address things better according to studies
and that people who dedicate their lives to it
usually have lots of information and probablymore than any bill,
the ones we can contain an algorithm.
But the problem is that people are usually distracted
by things that are only marginally relevant.
Could apply unconscious biases and usually don’t have the systems needed
to analyze all of the information in order to predict
what could work today, but also in five years time.
By getting to know people better,
companies will also be able to address retention.
Since now they will be able to tell
who’s becoming disengaged before they leave,
they will have the opportunity to react before it’s too late.
Either to retain that person or to finda replacement on time.
Let me give you an example.
Let’s say Bob has been working for
X company for the past 3 years.
But now he’s getting bored, because he’s millennial
and he wants to make a change.
But he doesn’t have such a good relationshipwith his boss.
So he doesn’t feel like talking about it.
And honestly, he doesn’t see lots of opportunities within the company.
So what happens now?
Bob feels trapped and he starts to change.
His response times is faster and shorter.
He changes his LinkedIn profile, starts connectingwith recruiters,
starts to actively look for job applications.
He’s just different.
When he finally decides to walk away,
he just tells his boss who panics
and runs to HR and notifies about the urgent vacancy.
HR starts to look for a replacement after Bob’s gone.
It takes them about four to six months to find one,
another two months to train the new guy.
In the new meantime, Bob’s team
are asked to take more responsibilities and workload.
So they get pissed and they also get disengaged.
But once again, no one notices
and this continues on and on.
I know it sounds crazy,
but I see such things happening on a daily basis
in all sort of companies.
With this solution,
the algorithm would be able to identify the change in Bob’s behavior,
would send a notification to the boss,suggesting how to proceed,
based on who Bob is and what he wants.
Scary, brilliant, you will decide.
But truth is using an algorithm inHR related processes
could allow the possibility of creating predictive tile and pipelines.
Since now, companies would be able to tell
when someone is ready to leave before they do it.
Yeah, the system could potentially inform
the recruitment team about skillsthat they need to start hiring for
and even suggest a list of global candidates
analyzed based on success job factors,
rather than traditional job descriptionsand static resumes.
In a very short future,
things like gender,ethnicity or age will no longer matter.
If you have what the job requires,
then there should be no Unconscious Bias in your way.
It won’t matter if your boss doesn’t like you.
If you’re a mother of two
or if you believe in the Mayan gods.
If you have what it takes,then the job should be yours.
So with all of these technology, you might be thinking
why would a company need an HR team after all?
Trust me. They will.
Because this is all about human connections after all.
We’re talking about our future here,
so we’ll want to know who’s the people behind it.
诚然 正如我们所知 传统招聘将会消失
Recruitment nevertheless, as we know,it will disappear.
It will evolve into something
that no algorithm will be able to change and to take.
It will become a strategic guy for the company
and a human element for the employees.
So this means that using a statistical model
will replace hunches with data,
presumptions with models reverseand intuition with success reaches.
But the final decision will be taken by a human
and not leave it to an algorithm.
This way in a very close future,
companies will be able toretain their job talent longer,
will be able to identifing the best candidates faster
and at the same time, us as candidates
will be able to identify and work
in jobs that we really want and that are really aligned with who we are.
And that way the question that I first asked you,
do you think we could ever find a job
that was customized for us?
We’ll turn it into a yes.