4. Fine Print
Job postings and organic
website traffic is not
something I've focused on
over the years. Some
companies rely heavily on
them for candidates. I don’t
judge if that's how you do
things. I prefer going out
and getting who we want. I
believe it’s impossible to
build a great company
another other way.
Disclosure: We're investors in
LinkedIn & Treehouse.
6. spray & pray doesn’t
work
trend is operational & data driven
to find the fish, be strategic & leverage the web
pool is small, but new ways of growing it
Page 6
9. 100% (X) Total Out reach
(Top of funnel)
Avg Response Rate:
10% to 20%
10-20%
Basic
Metrics 0.7%
Intro to Offer Rate: 10-
to
15:1
2%
0.3%
to Acceptance Rate:
2% 40-100%
Page 9
10. 1,450 Greylock Talent Team
Outreach in Q1
320
(22%) Response rate
Real
Example
435
Total Introductions
29
Intro to Offers
(15:1)
17
(59%) Acceptance Rate
Page 10
11. Top of the funnel Increase outreach
using
github, linkedin, etc.
Response
Rate AB testing outreach
Levers
Intro to
Debrief after interviews
Offer
Rate to improve conversion
Accepta
nce
Rate Pre-closing
Page 11
17. Best sourcer: high relevant, high intent
“Pull Sourcing”
Scraping
Spamming
Relevancy
Agencies (do not scale well)
Job boards
“Push sourcing”
Intent
Page 17
18. Matching algorithm
It’s a data game
Tags (relevancy) Behavior (intent)
Extracted tags from candidates &
Matching
Convert actions taken on
Whitetruffle
companies profiles and other
by companies & candidates to
sources available.
knowledge to improve matching.
Matching score
Human validation
Match is released
Page 18
26. TalentBin’s 360 view
*
Extract professional activity
Analyze & surface skills
Aggregate all contact info
* he’s happy because he never updates LinkedIn, but now, with TalentBin, you can find him!
Page 26
32. Higher education is broken
with increasingly higher costs
for students and society at
larger. Education is no longer
a one-time event but a lifelong
experience. What is taught in
universities is a mismatch with
what employers
need, especially in technical
fields.
Sebastian Thrun
Udacity Founder. Google Fellow. Stanford
University Comp. Sci. Research Professor.
Autonomous Car Inventor. 50 Smartest
People in Tech (Fortune). Education disruptor.
Page 32
34. Big breakthroughs happen
when what’s suddenly possible
meets what’s desperately
necessary. Higher education is
prohibitively expensive. But
web technologies now allow
100,000+ students to be
taught by one Stanford
professor.
Daphne Koller
Coursera Co-Founder. 3rd generation Ph.D.
Stanford University Professor. MacArthur
Fellowship winner. Education disruptor.
Page 34
37. To Recap…
trend is operational & data driven
There are new amazing tools, but there are no shortcuts
We can help create bigger and better pools of talent
Page 37
1. Who am IGreylock, Mozilla, Rypple, previous start experience2. My role at GL -> Education, Programs (university), high quality product and engineering hiresMy view of the world as shaped by startups
What is perceived as a lack of talent is really a lack of operational rigor mixed with passivity. We could still benefit from a bigger pool of talent but it’s able to be navigated.
What do I mean by pasivity: You have to got out and build the companies that you want. They don’t’ just come to you. Job Posting have always been a historically small piece of what I’ve done in the past
Spray and pray does not work Hope is not strategy, must be strategic and data driven. Must leverage all available tools. For those companies that are larger, they need to look at ways of expanding the broader talent pool available to them or startups will bleed them dry
Started with Salesforce in 1999. 1. they were able to use data to add predictability into the sales process. We all take forecasting for granted now, but it's a fairly new revelation. With the rise of data the need for quantitative people has become essential. 2. In no place is it more evident than marketing. One of the toughest searches to find right now across the portfolio. The reason is that the quantative marketer is in demand and most current marketers are mostly qualitative. Companies now drive their marketing like a sales funnel 3. Same holds true for recruiting. The best recruiting leaders inside of our portfolio are incredibly quantitative. They understand the recruiting funnel exceptionally welThe treat recruiting much like a sales pipeline or marketing leads
We consisently here the wave of decision on data. Reid Hoffman founder of linkedin mentioned that he felt that web 3.0 would all be centered around data.Data allows us to be understand were our candidate is flow is coming from, where we’re loosing our candidates in process and where to allocate our 2 most precious resources: money and time.
While my focus is essentially on external outreach. The team at Zynga breaks down this data for internal referrals and all inbound candidates from the website or job postings.Time in ProcessSimple, but companies fall down on this constantly. Measure how long people are in process, it should not be longer than 1-2 weeks. If you're taking longer than that your process is broken and needs to be optimized. Look at where people are being held up and make adjustments. Keeping momentum throughout the process is a huge advantage. It shows the individual that you know how to operate efficiently.There are other tactics with especially difficult or competitive positions such as offering large internal referral bonuses or offering candidates extra cash or equity to sign by a certain date. But I would only use these tactics sparingly, if it all. Having wide internal comp discrepancies causes it's own set of problems.
The world has changed drastically since 1999 when the phone was the most widely used sourcing tool. Everyone has a large web footprint and it’s impossible to be completely incognito if you’re a talented developer
Linkedin and Google are Table StakesLinkedIn:Along with google search, is the only essential recruiting tool. If you're not using it then you brought a knife to a gun fight.
LI recruiter has broadened their offereing to include marketing campaings and targeted advertisting at perespective candidates. They’ve added a pipeline tool to better track status and search. It’s still pretty early in it’s development so there are a few things to still figure out.
Whilelinkedin is still the main source of recruiting there are new tools and products that are helping Better determine matches between individuals and companies Predict intent as to when people are potentially look to switchBuilding better networks from developers free of recruiters spamming their inbox A lot of these companies are still very early in their development and they will continue to get better. They should only be used a part of comprehensive talent strategy. LinkedIn was not useful for the first few years of it’s existence, it was only when it acquired a critical pass of people did it real become irreplaceable.
2. WhitetruffleIs using data and machine learning tactics to match individual and companies. Their algorthm learns over time based on the candidates that companies select who to show them in the future. It also does the same thing for active candidates in the system.
Github
GitHub - They recently raised a 100M in fundingWTF is GitHub? Have a probably heard about it. Code Repository and has over 1M users and 2M respositories. It's essentially a social network for developers where they interact over code. fork, pull request and merge4. We're seeing more students using github to improve their engineering abilities. It's often something I encourage students to show their proficiency even if they haven't had much work experience.5. There’s an amazing amount of activity on top of it
GitHubTypical profile looks like this: Jeff is one of the developers at mozilla, and my model for the day, you’ll see him again later
Gild wants to demystify gitbhub by showing you:How good a developer is at each languageUndersatnd years of experience Determine influence in developer communities. Go to slides
The implicit web is a synthesis of personal information multiple sources on the web to create one single coherent picture.
Explict sources include. LinkedIn, resume and to a lesser extent facebook.
Tweeting about security, python or even the san francisco giants show’s an affinity and knowedge for a particular area.
Talent is aggregating all the pieces of information around an individual to tell one comprehensive story.
So far we’ve covered people self selecting to show they have the intent to move and we’ve looked at ways of identify people inside of networks and also being able to aggregate profiles. But is there a way to predict intent.
Entelo believes there is. Patent pending algorithm that analyzes over 70 variables to predict if a passive candidate is going to become active. This will be what really definies their company. They have search and it works well. Did a quick search on scala – new programming language
Before the pool. I want to reiterate that these companies are all still very early in their development and that companies still need to be very effective at doing out reach. None of these products will do your recruiting for you. You still have to tell compelling company stories and run an amazing process. For those of you concerned that the talent pool is still too small, Over the next few years there will be unique opportunities to create bigger talent pools. We will be able to make more developers
Education is being fundamentally changed.People are able to develop specific technical skills at an amazingly fast pace. They’re also able to take classes from some of the best professors in the world.
Udacity was started by SebatianThurn who is a google fellow and a part-time researcher. He ran google X. which developed concepts like google glasses and the self-driving car. He wants to teach people at a global scale the specific technical skills they need to know. So far the have 11 courses offered in web development, testing Advanced techniques like articficlal intelligence and crytography. My OD hate comes on when I think about this. Instead opening additional headcount why don’t we train our current people in advance of what we’re going to need.
There are no prerequisite programming courses required and they even work with partners to help distribute resumes.
For those of you who haven’t seen it there is a phenomenal Ted Talk by Daphne Koller co-founder at Coursera that’s well worth checking out. That’s really dives deep into how people learn and absorb information. The work work with the best professors around the world to take classes for free.
There’s an upcomingscala class on coursera that at last count had over 10K registered participants. Scala is relatively new language based on java that very few people have knowledge of.
Focusing on web development and design. Have worked with companies like Disney, to retrain their people and modernize their engineering team