I am data-driven and
so are you
Director of Talent, LevelUp
Jen Paxton
@Married_2_IT
Where I’m coming from
Talk outline
1. Why I became data driven
2. Process for evaluating performance
3. Tools, decisions and action
4. Carpe Diem!
Why I became data driven
Why I became data driven
June 2013-July
2015
Company Fiksu
Growth 150 to 300
What I wish
I knew
What’s our best
source of
applicants?
Emotional
state
Why I became data driven
2013-2015 2016- May 2017
Company Fiksu TrueMotion
Growth 150 to 300 20 to 40
What I wish
I knew
What’s our best
source of
applicants?
How long does it
actually take to
hire someone?
Emotional
state
Why I became data driven
2013-2015 2016- May 2017 2017
Company Fiksu TrueMotion LevelUp
Growth 150 to 300 20 to 40 150 to 200
What I wish
I knew
What’s our best
source of
applicants?
How long does it
actually take to
hire someone?
How are we going
to scale faster?
Emotional
state
Designing a process for
evaluating performance
Process for evaluating performance
Create a checklist
○ What are the most important metrics?
○ What time period?
○ Who do I share these metrics with?
○ How frequent do I pull these metrics?
What are the most important metrics?
Candidate
Experience
● Career Site Traffic
● Source per hire
● Recruiting funnel
Quality of hire
● Interview Feedback
● Referrals
● Job Description
effectiveness
Speed/Efficiency
● Time to hire
● Cost per hire
● Offer acceptance
rate
Recruiting Funnel
Tips
● Establish what a “healthy
pipeline” looks like
● Find bottlenecks in process
● Do a Post Mortem after the
role is filled or after 30 days
Phone Screens
Offers
Applicants
Onite Interviews
Hires
Referral Metrics
● Best source for recruiting
● How do we get them?
● Ensuring quality referrals
Of hires are from
referrals
40%
Measuring effectiveness of Job Descriptions
A/B Test
Job Titles
● Sales Executive
vs Sales Manager
● Rails Developer vs
Ruby on Rails
Developer
Personalize
Content
● Tailor content to
speak to your
ideal candidate
● Use words that
evoke your
company culture
Use Unbiased
Language
● Leverage Textio
(Gender decoder)
● Limited number of
must have
qualifications.
Metrics for hiring managers and Execs
Time to hire Cost per hire Source Per Hire
Use the right technology
& tools to inform
decisions
And drive actionable results!
Putting Data into practice
Define
“must
haves”
Standardize
collection
process
Determine
frequency
Identify
audience &
their needs
Tailor
reports for
audience
Carpe Diem
“Don’t put off tomorrow what you can do today”
~Benjamin Franklin
To hear from more speakers at our
Summit, watch our live stream
recording here:
https://www.lever.co/resources/tale
nt-innovation-summit/main-event
Jen Paxton
Director of Talent, LevelUp
@Married_2_IT

I Am Data-driven and So Are You

  • 1.
    I am data-drivenand so are you Director of Talent, LevelUp Jen Paxton @Married_2_IT
  • 2.
  • 3.
    Talk outline 1. WhyI became data driven 2. Process for evaluating performance 3. Tools, decisions and action 4. Carpe Diem!
  • 4.
    Why I becamedata driven
  • 5.
    Why I becamedata driven June 2013-July 2015 Company Fiksu Growth 150 to 300 What I wish I knew What’s our best source of applicants? Emotional state
  • 6.
    Why I becamedata driven 2013-2015 2016- May 2017 Company Fiksu TrueMotion Growth 150 to 300 20 to 40 What I wish I knew What’s our best source of applicants? How long does it actually take to hire someone? Emotional state
  • 7.
    Why I becamedata driven 2013-2015 2016- May 2017 2017 Company Fiksu TrueMotion LevelUp Growth 150 to 300 20 to 40 150 to 200 What I wish I knew What’s our best source of applicants? How long does it actually take to hire someone? How are we going to scale faster? Emotional state
  • 8.
    Designing a processfor evaluating performance
  • 9.
    Process for evaluatingperformance Create a checklist ○ What are the most important metrics? ○ What time period? ○ Who do I share these metrics with? ○ How frequent do I pull these metrics?
  • 10.
    What are themost important metrics? Candidate Experience ● Career Site Traffic ● Source per hire ● Recruiting funnel Quality of hire ● Interview Feedback ● Referrals ● Job Description effectiveness Speed/Efficiency ● Time to hire ● Cost per hire ● Offer acceptance rate
  • 11.
    Recruiting Funnel Tips ● Establishwhat a “healthy pipeline” looks like ● Find bottlenecks in process ● Do a Post Mortem after the role is filled or after 30 days Phone Screens Offers Applicants Onite Interviews Hires
  • 12.
    Referral Metrics ● Bestsource for recruiting ● How do we get them? ● Ensuring quality referrals Of hires are from referrals 40%
  • 13.
    Measuring effectiveness ofJob Descriptions A/B Test Job Titles ● Sales Executive vs Sales Manager ● Rails Developer vs Ruby on Rails Developer Personalize Content ● Tailor content to speak to your ideal candidate ● Use words that evoke your company culture Use Unbiased Language ● Leverage Textio (Gender decoder) ● Limited number of must have qualifications.
  • 14.
    Metrics for hiringmanagers and Execs Time to hire Cost per hire Source Per Hire
  • 15.
    Use the righttechnology & tools to inform decisions And drive actionable results!
  • 16.
    Putting Data intopractice Define “must haves” Standardize collection process Determine frequency Identify audience & their needs Tailor reports for audience
  • 17.
    Carpe Diem “Don’t putoff tomorrow what you can do today” ~Benjamin Franklin
  • 18.
    To hear frommore speakers at our Summit, watch our live stream recording here: https://www.lever.co/resources/tale nt-innovation-summit/main-event Jen Paxton Director of Talent, LevelUp @Married_2_IT

Editor's Notes

  • #2 Thank you all for joining me today, I am excited to talk about being data driven.
  • #3 A little bit about me, I have a masters in Opera vocal performance. I didn’t go to school for HR or Data Science, like many people I know I fell into recruiting to make money. I started off as an agency recruiter and then made my way into a corporate role. I’m in no way a data expert and I can only talk about what has worked for me and what I wish I had known or done in my past roles. I do hope that my presentation will help you become more comfortable using data to inform your recruiting decisions. Today I want to talk about why I became data driven.
  • #4 I’ve been at a few startups now and have learned a ton about how to use data to my advantage as a recruiter. I’ll talk about my process for evaluating performance, the tools that have informed my decisions and the action that I have taken to improve my recruiting process using data.
  • #5 I want to start this off by highlighting three startups I have worked at, the things I wish I had known
  • #6 I was at Fiksu an advertising technology startup for about two years. I started as employee 150 and grew the team to 300 within those two years. For the first year it was only me, I was recruiting like crazy, I felt rushed and sometimes felt like a chicken with my head cut off. I was hiring so fast that I felt like I didn’t have time to slow down and look at the data. Something that I wish I had known was where our candidates were coming from. Tools I was using: Jobvite, Linkedin Recruiter.
  • #7  I then joined a Series A startup called TrueMotion, they have a safe driving app. I started with them when they were in an incubator at the Harvard Launch lab when they were only 20 employees and grew them to 40 employees within the first year. In my first few months at TrueMotion I was overly confident that I could fill their engineering roles faster since it was so easy for me to do at Fiksu. I over promised what I could do within my first three months and it was eye opening to me how few applications we received and how long it took to fill a role when your company is unknown or has little to no brand recognition. I wish I would have tracked my pipeline metrics more closely, I think this would have helped me to not over promise. Tools I was using: Greenhouse and Linkedin recruiter lite.
  • #8 Currently I’m at LevelUp a mobile payment network targeted at quick serve and fast casual restaurants. LevelUp has been around since 2008. Now I’ve learned to take a more humble and thoughtful approach to recruiting and leverage data analysis to inform my decision. We are hiring like crazy now, I’ve been there for about 4 months and we have hired over 40 people. However, I feel more zen now and I’ve gotten in the habit of asking better questions that are more strategic and forward thinking versus tactical and reactive. Currently tools we use for tracking are Lever, Linkedin Recruiter and StackOverFlow Now that I talked about what I wish I had known, I want to talk about how I designed a process to help me use data to my advantage
  • #9  Incorporating in a process for evaluating performance is key to helping you run a data driven recruiting function. You need to be able to look at the data up close and then zoom out to see the bigger picture. Doing Side-by-side as well as month-over-month comparison coupled with testing different theories can show you what’s driving performance improvements over time.
  • #10 My first step to creating a process for evaluating performance is to create a checklist. I Think about what metrics are important to me and to my stakeholders (hiring manager and my COO). Next I decided what is the time period that I want to pull this data from, currently I am looking at our 2017 hiring trends, next year I will do a comparison of our 2017 and 2018 hiring trends. There are so many different things to measure and so many people you can share the data with that you need to actively think about what data you want to share and who you share those metrics with. The last thing I do is set up monthly reports within lever. Currently i’m tracking applicant flow and time to hire monthly.
  • #11 Here are some metrics that I look at. If you are interested in the candidate experience, I found that focusing on career site traffic and source per hire can be great metrics. I want to make sure that we are focusing on the job boards that will bring us the most qualified candidates. Knowing source per hire helps me spend my talent budget in the right places and helps me get rid of job boards that do not perform. Additionally for candidate experience, I like to look at all the stages of our recruiting funnel, the last thing I want to do is lose a candidate because we are slow with our recruiting process. To help assess Quality of Hire, I let interview feedback be my guide, Lever does a great job of collating the interviewer's feedback. Other things I’m focusing on are the quality and amount of referrals and how effective our job descriptions are. If I want to focus on speed or efficiency I track time to hire and cost per hire, like source per hire, this will help you to put your money in the most cost effective sources and not waste your time on sources that do not produce hires. We also take into account our Offer Acceptance rate, which is around 80% when we are thinking about speed. Let’s do a pulse check, what are some metrics that you are tracking that are not on this slide here?
  • #12 Candidates are becoming increasingly picky so your interview process needs to be streamlined to keep up. I communicate early and often with hiring managers and the interview team to ensure that there is nothing in our process to slow down hiring. Here are some tips I use to streamline my recruiting process. Establish Healthy Pipeline: When I'm hiring 1 person... think about funnel metrics.... What is the ratio from onsite to hire. Knowing how many people we need to screen or bring onsite helps me set sourcing goals. For example, the rest of this year our stretch goal is to hire 10 new Ruby on Rails developers. I know that we need to interact with at least 80 prospects to get 1 hire (given the candidate response rate I know through LInkedin analytics, the prospects that will actually convert to applicants and the applicants that will actually make it through our recruiting process). Knowing all of this, plus my non technical recruiting responsibilities--- this is more than what one person can do so this justified me to bring on a tech recruiter and use agency recruiters strategically for these hires. If I wasn’t tracking this data, I might not know that I was under water until it was too late. Bottlenecks: I do weekly pipeline checks with my recruiter Pat to see if there are any bottlenecks with our process. We look at the entire recruiting funnel starting with applicants: How are you getting applications? We use GA to track how many Careersite pageviews convert to applications. (Marketing is your Ally here and can help you use GA and make sense of the data) We also look at how qualified our applicants are. Currently tracking for this is pretty manual, we track how many phone screens convert to onsites and how many onsites convert to offers. After you have done the analysis identify bottlenecks, focus on the stage that needs the most attention, doing this can guide your efforts and keep you focused. (and you won’t feel like a chicken with your head cut off) Post Mortems: After a role has been filled or open for 30 days (you can choose to time to fill by target- I like 30 days) reflect on what worked with the interview process and what has not. Are there things that you have learned that we should do for the next round? These Post Mortems for me have been invaluable, they help me to streamline other roles and when I fill the same role again in the future I have new data that may help me fill the role faster.
  • #13 We all know that referrals are one of or the best source for recruiting. Since my days at Fiksu where I didn’t take the time to understand where candidates were coming from, I have gotten more sophisticated in tracking referral data, I can now track our percentage of referrals over time. In Q1 of this year we had 31% of our hires from referrals whereas in this last quarter since I started our referrals have gone up to 40%. Knowing this data confirms that our referral bonus and our continued efforts to talk about referrals with employees seems to be working.
  • #14 Show of hands, who’s doing are A/B Testing their Job Descriptions ? Give me some examples of what you are doing. We are also A/B testing job titles on StackOverFlow to see if we can drive more inbound engineering candidates that way. Personalize Content We have found that personalizing the content in our job descriptions has gotten us more qualified candidate. I know that I am more likely to get a response from m prospecting emails if my message is tailored. It’s the same way with job descriptions. Write the job description like you are writing a message to your ideal candidate that you want to convince to come work with you. If you have not found the right candidate within 30 days, talk with the hiring manager about their target candidate and update the job description. If the role changes, update the job description to reflect that change. Additionally, use words that reflect your company culture, if your company is casual try using more casual language. Is anyone focused on using unbiased language? Use Unbiased Language I use Textio, a text decoder that detects which words are more masculine coded or feminine coded to help our job description use unbiased language. It’s not foolproof as there are some words that it will not pickup on, but it is a good start. Additionally we found that having too many bullet points for required skills has turned women away. Limiting the required skills to a few bullets has increased our female applications. On the engineering side, for our Ruby on Rails role we only have three required skills: 2+ years professional software development experience, Team player with excellent written and spoken English communication and a Bachelor's degree in Computer Science or ample real-world experience.
  • #15 Metrics used today in communicating with hiring managers and members of our exec team: Time to hire: One of the main metrics I get asked is 'how fast are we hiring people?’ When we're deciding whether we want to bring someone in to meet the team, I will talk with the hiring managers about their pipeline and time to hire. Do they have realistic candidate expectations or should they find someone who is junior but coachable? Knowing time to hire also helps you to know if something needs to be changed with our recruiting process. (Enterprise Sales Exec role open for almost a year, I came in and looked at their process for hiring for sales and changed the process around a bit. The length of the role being open allowed me to convince the HM to broaden her scope of what she was looking for and we were able to make an offer in my first three months to an Enterprise Sales Exec. Cost per hire and Source per hire If you want to add on a certain software that you know will help increase applications/hire, Knowing CPH/CPS can help inform your decision and support you when presenting to execs. (for example we had a free trial with BiB last year, using the source per hire report in lever I was able to see that we made 3 hires from BiB. I compared the amount of hires we got from BiB to other sources we use like The Muse and I could clearly see that BiB was a better source. This provided me with the justification and data to go to execs and suggest a paid membership with BiB. Sometimes keeping cost down means needing to source ourselves and have strategic usage of agencies. Not using staffing agencies keeps our cost per hire down but this may also shift our time to hire way up. It’s a balance that is dependent on your company's need.
  • #16 There are technologies out there that will help you be quicker but there are also ones out that that will be a total time suck. It’s a noisy space and you really just need a couple of fundamentals, we use Lever to get hiring/funnel metrics, GA, Linkedin Analytics and Excel. All of this is really dependent upon your individual team’s needs so you need to make your own assessments of what works for you. When you have selected the right tool for the job then you should work on putting data into practice.
  • #17 Like buying a how you have your must-haves and nice-to-haves. Is it a good use of your time to see every point of contact between the recruiter and the candidate? Probably not. Make sure to focus on the data that will help you make more informed decisions, cut costs and improve efficiency. When you have the basics down you can start to identify areas for improvement, make adjustments to your recruiting funnel and create a framework for continuous improvement. Once you know what data is valuable to you, create a process to collect that data. You want to make sure that your team (and you) have a standard process for inputting the data. Avoid duplicates, try to keep manual entry to a minimum and practice clean data input. Garbage in garbage out. Designate one person on your team to own the process and performer the analysis. This doesn’t mean that the whole team isn’t being data driven and thinking in a data mindset way, it just keeps the process simple. Depending on the size of your company and resources available, you should determine the frequency of benchmarking, I found that it should happen anywhere from once a month to once a quarter to be truly effective and allow my team to course correct and optimize throughout the year. There are so many different things to measure, sit down with your team, your hiring manager and your exec team to find out what the most valuable metrics are to them. helps me to build tailored reports for each audience. A hiring manager may be curious of source per hire whereas an exec may want to concentrate on cost per hire and not get bogged down in the details of source per hire. I also edit my reports to give execs a 50,000 foot view of sources while with hiring managers I many get into the nifty gritty of each source.
  • #18 Being a data driven recruiter or driving a data driven recruiting strategy should start today. It is about thinking strategically and getting things done. When there is a hiring effort I have to focus on things that are high value and least effort. You can start off small and scale are your organization grows.