Leveraging AI and Machine Learning for Better Matching with Indeed Smart Sourcing - Raj Mukherjee, Indeed.com

[00:00:00] you're being so kind, right? You know, you know this bit where we would every

[00:00:06] recruiter, every recruiter in the world has done this bit at one point. Usually it's

[00:00:11] early in their career. They go to indeed, they'll go see a job that okay that

[00:00:16] nursing practitioner they want to go cut copy. Go back in the work. Yes, yes,

[00:00:25] again, no worse. They'll use a job description from 10 years ago that's on there on

[00:00:31] there. You know, that was for a job that previously they had to fill. They'll use

[00:00:35] that job description and make you won't update it at all. They'll just post it. I mean,

[00:00:40] again, young and once career, you make a lot of mistakes. But I like I love the

[00:00:45] fact that you walk them through. All right. I want to talk to you for a moment about

[00:00:49] retaining and developing your workforce. It's hard recruiting is hard retaining top

[00:00:54] employees is hard. Then you've got onboarding payroll benefits time and labor

[00:00:59] management. You need to take care of your workforce and you can only do this

[00:01:03] successfully if you commit to transforming your employee experience. This is where

[00:01:09] I saw comes in. They empower you to be successful. We've seen it with a number

[00:01:14] of companies that we've worked with, and this is why we partner with them here at

[00:01:18] Work Defined. We trust them and you should too. Check them out at iSolvedHCM.com

[00:01:25] Hey, this is William Tick-Up and Ryan Larry. You are listening and watching to the

[00:01:33] Use Case podcast. We have Raj from Indeed on. We're talking about a new product that they

[00:01:38] have called Smart Sourcing. And he's going to tell us all about it. So let's just jump

[00:01:43] in and Raj. Hello, how are you doing? Doing great, William. Great to be here. Doing

[00:01:49] great, Brian. Great to be here as well. Yeah, good. So we've talked through the years.

[00:01:54] We've talked about all kinds of new and innovative things that Indeed does and has

[00:01:58] done in the past. So tell us about Smart Sourcing. What do we have here? Yeah. So let me

[00:02:05] start by setting a bit of context. I think that might be helpful. You know, I've been

[00:02:11] adding Indeed for the last seven and a half years. And one of the things that I have seen

[00:02:15] consistently is the challenge employers face. Job seekers also face lots of challenges

[00:02:20] as well as target employer challenges. Right. Employers that challenges face on average

[00:02:25] and William, you probably know this in and out like it takes on average more than 50 plus

[00:02:29] days to hire for a role. And what's even more staggering in a recent survey we found

[00:02:36] for sourcing on a role like when you're reaching out to actively to candidates in that scenario.

[00:02:43] An average recruiter or a hiring manager is going to spend about 13 hours per role. So

[00:02:49] I imagine if I just imagined that, like, you know, last time I was talking to Brian,

[00:02:55] we were talking about hey, I don't have a bridge. People have probably five to 10 roles

[00:02:58] but they're worth one. It's got big 10 roles. Yeah, right. Three weeks, three weeks, you're

[00:03:03] just sourcing to even get a list of quality candidates. So it's and it's very manual. And I have

[00:03:09] seen it not change much. So people going in, typing in complicated search queries and scrolling through

[00:03:15] lists and this page is candidates. Right. Look, looking at someone's resume for less than 20 seconds

[00:03:21] to determine whether that person is a good fit. And oftentimes getting very biased in their outcomes.

[00:03:28] And then good luck when you reach out because the person may not be interested.

[00:03:32] And so that's sort of being more of a data, the Earth, the data is old. You're spending more

[00:03:38] wasted time reaching out to an old email address or an old phone number or something like that.

[00:03:43] Very astute algorithmic version. Yes. And oftentimes not hearing back. And the reality is

[00:03:50] we haven't really reimagined sourcing for a long time. And so circle 2021

[00:03:56] we launched this product called matched candidates. And I'm giving you a little bit of history,

[00:04:00] so you kind of connect the dots with what exactly we're launching today. But with matched candidates,

[00:04:05] what we did was you give us a job. We will parse the job description. We will ask you the right

[00:04:11] structured questions. For example, hey, do you really need three years of experience in nursing

[00:04:16] for a nursing role? All those kind of things. And then once we have identified those,

[00:04:21] we will automatically deliver the best most active candidates. Very important, the best qualified

[00:04:29] and the most active one candidates. And we found that the results were staggeringly good.

[00:04:36] When employers would or as would reach out when a recruiter would reach out through matched candidates,

[00:04:42] the job seeker on the other end would be 17 times more likely to respond than if they found

[00:04:47] the job on their own. Like literally 17 times more likely. So Roger, and I know you're

[00:04:53] we're going to go through the evolution here but how are you determining the activity level

[00:04:58] of these potential candidates? They can see usage in indeed. That's my guess. So they can see

[00:05:04] how often they were in or last time they were in and how many searches blah blah blah. That's my guess.

[00:05:09] Of course, Roger. Correct me. Sorry. I'm so sorry.

[00:05:18] I genuinely love it when I have folks like William who actually understand not just the

[00:05:24] external usage but why it behaves the way it behaves. So yes, very true. And we capture every

[00:05:31] single interaction with indeed and whether it's a search and if you let's say you get into a

[00:05:38] search result and we will show you a job in the view job itself, we'll ask you questions.

[00:05:44] Hey Ryan, do you have five years of recruiting experience?

[00:05:48] Hey Ryan, do you live in this location? Like all these questions will confirm. They think

[00:05:51] that they're inferring and those become part of our activity so we know that the person is active.

[00:05:57] So you don't you don't you don't have a little person or a person sitting in a car outside.

[00:06:05] No, but you're enriching the data. I mean, the cool thing is you're doing two things. You're

[00:06:10] enriching by validating, confirming, et cetera. You're enriching that data and it's also giving

[00:06:16] them a better experience. I mean, we're talking about the warrior. But really they can't it.

[00:06:21] These are better results for them. So if they say, oh no, I'm not in Boston anymore of move to

[00:06:27] Philadelphia. It's like, okay, well you can change that and then I'll start now surfacing things that

[00:06:33] are more Philadelphia based. That is totally that is exactly right. That is exactly right.

[00:06:41] I think the part that is really important here and actually this is a good segue into that comment.

[00:06:50] We do not believe a resume is either a perfect reflection of someone's background nor is it the only

[00:06:57] authoritative document that you should use. Right. So what you just opened up this window as box

[00:07:02] that we just open up around activity, all of these activities, preferences, qualifications that

[00:07:08] exist about a job seeker plus their resume, plus assessments that they take on our platform,

[00:07:13] all of that together becomes a 360 view of a job seeker. And that's what we are using to do the matching.

[00:07:20] I don't know if it's for me what it gets is it gets the activity also can get you and you're matching on

[00:07:29] quality. And you're matching on activity, somebody set up an indeed profile 10 years ago that they've

[00:07:36] never logged in since. Okay, that's that I mean, it might be a great match so it might fulfill one

[00:07:43] criterion. But if they're not active, then you know that might waste time it might again it might

[00:07:51] waste the the source of the reverse time and then trying to contact them. But I want to go back to

[00:07:56] really quickly too when you parse the job description. You're also I think you can probably

[00:08:02] using more AI machine learning now to parse both the all the experience and all of those types

[00:08:08] of things and doing the match between those two right. 100% and it's it's not keyword matching I

[00:08:15] get this question all the time. And I actually have I have people who keep stealing me like should I

[00:08:21] put more keywords in my resume or more keywords in my job this is no no it's far more intelligent than that

[00:08:28] today. Well, it's like the continuous skills that you have so if you know how to like you know just

[00:08:37] take take a they take a take a software development language. If you know this language,

[00:08:42] we already know that you probably have a pretty good understanding of five languages that are

[00:08:48] continuous did you list that on your resume or your profile? No but we we can kind of put the dots

[00:08:55] together and an AI can completely do that they can put all those dots together for us. Yeah and

[00:09:02] I mean if you take take it one step further these these these these are all things that are

[00:09:05] happening right now. Let's say you're a engineer working for company X and we have enough

[00:09:11] profiles from company X. We know the kind of systems that that company X uses so so we can start to

[00:09:19] infer things that yeah so all these like we are starting to build really a skills database

[00:09:24] and in that skills database we are connecting the dots and essentially what would be called

[00:09:31] in other other parts of the world knowledge graph but I mean that's a mouthful like very simply

[00:09:36] what I say is connecting the dots we just connecting the dots effectively. With the job description do

[00:09:42] we do we do we need to do a better job or do we have a way to make that more enriched so that we

[00:09:49] understand if they ask for what they're really saying is they need B and C. So multiple questions to

[00:09:57] that question let me start with a lot of people ask me this question around how is AI getting used

[00:10:02] in job description right so what is happening. So if you come to indeed you post a job we will guide

[00:10:09] you through a very prescriptive step by step on what you need to share to make your job description

[00:10:15] awesome. Oh cool okay literally like every step for example I am posting and continuing the

[00:10:22] missing examples so I'm posting a nursing job they will say hey nursing jobs you need a

[00:10:27] nurse certification you need to have that other requirement and be kind of very much like.

[00:10:33] Exactly so those kind of things we're adding the job description itself is one of those

[00:10:39] you know writers blocked like people come in and they see a blank page. You're being so kind

[00:10:46] right you know you know this bit where we would every recruiter every recruiter in the world has

[00:10:53] done this bit at one point usually it's already in their career they go to indeed they'll go see a

[00:10:58] job that that okay that that nursing practitioner they want to go cut copy. So go back into work yes

[00:11:07] yes you did no worse they'll use a job description from 10 years ago that's on there on there

[00:11:15] you know that was for a job that previously they had to fill they'll use that job description

[00:11:20] and make it won't update it at all they'll just post it I mean again young and once career you

[00:11:26] make a lot of mistakes but I like I love the fact that you walk them through to make sure that to

[00:11:32] give them the best opportunity of getting matched. Yeah so Raj are you are you writing are you providing

[00:11:39] the writing the copy behind the job description or just prompting them to to include certain things

[00:11:46] now we actually are also essentially writing or auditing the job description yeah very

[00:11:56] easy yeah well because you know this you know what people search for so you can actually again

[00:12:03] understand how people are searching how they use you know the search functionality and indeed

[00:12:09] you can then tailor that to them and say hey listen no you don't have to say hey listen

[00:12:14] hey I just provides the sentence paragraph and say did you go in and edit it I'm sure but why if

[00:12:21] indeed providing the copy that's going to give you best kind of research results why why change it

[00:12:30] 100% and there is always the human side of things yeah we want to yeah there's always and

[00:12:37] and you know these technologies they're very new right I always say this like you can use

[00:12:44] genitive AI is particularly great at summarization so if you give it a lot of content can summarize very

[00:12:50] easily but it's not particularly great at generating stuff that is 100% accurate okay there's always

[00:12:57] going to be some we call it hallucination lying whatever you want to call it yeah there's going

[00:13:02] to be some made up stuff and the reality is you'll have it and so you have to make the human aware

[00:13:09] that hey you know review the job description so there isn't something there that you are offended by

[00:13:13] right I love it because I think all hallucinations because it's like like fair I guess I guess

[00:13:22] that's a good word but they probably could have used something else because I think when people see

[00:13:25] that they're like I don't understand the solutionation stuff it's like okay I mean there's a more

[00:13:34] colloquial term you're essentially lying but I don't think that's true for a mission

[00:13:38] yeah a machine doesn't really lie I mean it's no no no it doesn't have feelings I say that's

[00:13:46] the way it's all the time like the machine doesn't have feelings it doesn't it's not like us

[00:13:52] it'll be interesting when it does have feelings that will be a different conversation but

[00:13:57] that will be a different conversation and we are not quite there with artificial intelligence but yes

[00:14:03] I think but getting back to the stream of consciousness that we are talking about so the idea here

[00:14:09] is very simple make the job description awesome because that's the foundation of matching

[00:14:13] make the profile awesome because that's also the other foundation of matching

[00:14:16] you will both have them the both sides and then I'll give you a sense of scale like we use

[00:14:21] 140 million qualifications and preferences every day to match right so if you if you go back

[00:14:28] into now connecting it back to smart sourcing because you asked me this question and I kind of

[00:14:32] went you to a journey now get back get back into smart sourcing very quickly smart sourcing

[00:14:38] is sourcing reimagine taking all of the stuff that I just described number one we have the most

[00:14:45] active pool of candidates we have nearly 300 million resumes on our platform and many of them are

[00:14:51] active almost the size of the United States wow so having that that's one second all this

[00:14:59] recommendation power this matching power that I just described how do you bring that to bear

[00:15:04] and essentially make it really easy whenever you have a job to see the list of candidates that are

[00:15:10] going to be active and qualified for your job and third make use of generative AI to reduce the burden

[00:15:17] so I know Ryan you when you and I last shared it you're talking about the time you would spend in

[00:15:23] reviewing resumes trying to find out if this is a good match what we have built is this concept

[00:15:29] of candidate highlights where we summarize I mentioned that generative AI is good at summarizing

[00:15:34] it summarizes the list of skills this person has and what I would call the opportunities this person

[00:15:42] has vis-a-vis this role right what might be an opportunity person has four years of experience you

[00:15:47] are asking five years of experience yeah so it's bottom line so I have a review of the resume so

[00:15:52] that's another one and then a very easy way for you to reach out to that person using our smart

[00:15:58] message right so combine all of this that is what smart messaging is it's it's really like finding

[00:16:04] candidates who are great finding them very easily reaching out to them very easily and making

[00:16:10] a connection that's fast yeah I think you know yeah so Rajas is interesting um and and I know a lot of

[00:16:18] platforms offer a lot of different things but I think historically recruiters have had the

[00:16:25] the it's real challenge have had the challenge of not necessarily just finding the people but making

[00:16:32] the imprensive of well Ryan only has four years of experience but they don't understand that Ryan

[00:16:38] also used so used to work guess my third person for you William talking in the other person thank you

[00:16:43] so much right I also worked at company a during this time frame using this skill set and during

[00:16:51] this time frame I potentially worked on these three or four projects that were ongoing at this

[00:16:57] organization at that time which gives me potentially the experience that I don't have in my resume

[00:17:02] you have that information and I think that's important for recruiters to understand that

[00:17:08] yeah this isn't this isn't replacing and not that this ever would this is part of the recruiting

[00:17:14] projects yeah well where these tools have gotten so strong right has been the ability for me to say

[00:17:22] well here's 10 people and here's 30 things I don't know anything about these because I don't

[00:17:28] it's not listed on their profile and I'm not smart enough to understand these 10 different dots

[00:17:35] that we're connecting across these these list of experience or the list of experience and I think

[00:17:40] that's a genius idea that is really I mean again idea here is what we are doing is

[00:17:51] outside pretty cutting edge in terms of matching and the ability for us to bring this kind of

[00:17:56] tool to the market but it's also what Ryan's referring to which is the market is disjointed

[00:18:02] and the company has like large companies have on average 16 plus systems and they don't all

[00:18:08] stop to each other right so you want a streamline the end-to-end user experience that allows you

[00:18:14] to go not just fine candidates but reach out to them very effectively then once you have reached out

[00:18:20] very effectively move them into your application tracking system ensure that you have the analytics

[00:18:25] that tell you what I can exactly happen with hiring right all of these things need to be packaged

[00:18:29] and that's what we're doing you know what I love well there's so many things that I love about

[00:18:33] this one is it's the the data that you're sitting on sitting on is better than it's it's it's

[00:18:41] bigger than I'll say that way it's bigger than any data set than any other one company has

[00:18:48] so there's that because you're using I'm probably some machine learning as well it's learning

[00:18:54] from itself so as it goes on it just gets smarter much smarter than any humans you've you've

[00:19:02] you've ticked the two boxes of speed and quality that every recruiter struggles with

[00:19:08] and so I can see in the I could just see kind of some of the next steps especially first folks

[00:19:15] that have other things that they need you know like background checks and reference checks or

[00:19:21] a behavioral assessment personality assessment all of those types of things for that position

[00:19:26] it's a match and in it's a match go do as a candidate go do these four things

[00:19:32] that we'll see if you're still matched and further kind of making them qualified or

[00:19:38] disqualifying great one of the one way or another then then it's even a better match it's hell let's

[00:19:46] schedule an interview yep yeah absolutely yeah please go ahead oh no no so Roger as

[00:19:55] Williams talked about matching so you know question that comes in my mind is

[00:20:00] throughout the history of matching as this was introduced into recruiting a lot of recruiters

[00:20:06] didn't put a lot of trust into it right it's you think we mentioned it earlier where it's no

[00:20:10] longer about keywords and historically it's always been well we're just matching keywords and you

[00:20:17] have x amount of keywords and here's the match and here's ranks at 85% and then the recruiter goes in

[00:20:24] and they didn't have a great experience with that maybe walk us through what is different now and

[00:20:29] you know and I know that a lot of the listening audience is going to understand this already but

[00:20:33] there's going to be the subset that does it doesn't and that they're still stuck in the keyword matching

[00:20:39] realm of things maybe walk us through a little bit at the difference of what you are all doing now

[00:20:44] versus just keyword matching that we are used to for the last 10 years yeah yeah I mean very very

[00:20:51] very important question actually this one this is the heart of matching and the historically what

[00:20:57] used to happen is what you described people would go in and even today you can go in and you can

[00:21:02] say I'm looking for a customer service person so I'll type in customer service representative

[00:21:09] on my keyword and I'll type in the location where I'm in Austin Texas so I'll just say Austin

[00:21:15] Texas and you'll get a list of resumes back based on those resumes you will evaluate them you

[00:21:23] will see which ones you feel are a good fit you might say I want five years of experience today

[00:21:27] as a filter you'll say greater than five years of experience it's a great the filter resumes that

[00:21:32] have that you might say I really want people who have been active in the last 30 days so you can

[00:21:39] filter against that and you can continue to have some filters we are people who have worked in

[00:21:44] let's pretend that I'm fighting for indeed I might say I want people who are working in a tech

[00:21:49] company so I'm going to put that so whatever it is I create those filters manually go through all

[00:21:54] these resumes and come back this historically used to be done using keyword search now when

[00:22:01] you move away from keyword search the way you started was you build machines intelligence in

[00:22:07] understanding much more about the words than just that I'm matching to strings for what do I mean by

[00:22:14] that let's say the word customer service that is an occupation customer service is an occupation so

[00:22:23] if you just took customer and service use those two strings and started matching that gives you

[00:22:28] one result but if you basically said customer service well I won't understand customer service or

[00:22:33] salesperson could have also had customer service a retail person could have also had customer

[00:22:38] service and then you start to map that as an occupation under skill and you start to connect the dots

[00:22:44] with n number of skills and then you certainly start to return results that may not even have the word

[00:22:51] customer service directly in the resume right you're starting to infer so that's intelligence we do

[00:22:57] it we just do it intuitively because we as humans we have been able to pattern and

[00:23:03] and understand that like if somebody comes and tells you I work in maces as a retail associate

[00:23:08] maybe like yeah you work in customer service you're mapping that mentally a machine doesn't know that

[00:23:14] so you have to train the machine so that's where what we have is occupation based taxonomy

[00:23:19] and then we use that as a foundation to create the connections with skills

[00:23:24] and then once you have occupations and skills you can start to infer and connect the dots

[00:23:29] going back to that original thing that I described in a much more meaningful way than you can do

[00:23:33] today and that's what we've been working on and we've been particularly good with English

[00:23:39] markets us being a good example right and we're evolving in other markets outside of us as well

[00:23:45] but this is the foundation and then on top of that you have machine learning algorithms

[00:23:52] how much does it render the results sorry very fast no very fast actually to your point I mean so

[00:24:03] let me sort of break down the ingredients and then you'll understand the speed concert

[00:24:08] so you have at the very back end data that we have collected about job seekers and employers

[00:24:14] then you have this taxonomy layer which is understanding this data then you have a search engine

[00:24:20] that really sits on top of it and then you have a user experience that's sitting on top of it

[00:24:24] like very crudely I mean there's a lot of system there's a bunch of other stuff in there

[00:24:29] a bunch of other stuff but think of it that way so the search engine is built with latency mind

[00:24:35] so our goal is always to return the results as fast as possible but return the appropriately good results

[00:24:41] there's no point in returning a bunch of results that are not useful so we're always trying to

[00:24:45] balance the quality and quantity with the speed that we need to return to the result set and it's very

[00:24:51] snappy like you I encourage anyone in listening to this call because it's actually free to sign up

[00:24:57] so just go and sign up for smart sourcing and once you do that and we can at the end I can work

[00:25:04] with William to show you the URL of where to go but very simply just go and sign up if there's a

[00:25:10] free trial try it out this doesn't take it's generally it's free so just try it out and see for yourself

[00:25:18] I love it I love it Ryan do you have something to do?

[00:25:23] so one of the things that we're seeing is and this is kind of an old problem but we're seeing

[00:25:32] kind of exacerbated now is that candidates that moment of attention we probably called a couple

[00:25:38] different things but they're enamored with something and then they're on to something else you know

[00:25:44] it's like it's like looking at TikTok right so you're looking at TikTok and I'm saying you

[00:25:48] look let's go scroll go the next one go the next one go the next one so would you ever

[00:25:56] like so that renders the results okay smart smart source it renders the results I'm a recruiter

[00:26:03] bam I've got five people that this is the sound this is this is perfect I need to reach out to them

[00:26:09] but they don't for the kind of a historical problem of recruiters and hiring managers thinking

[00:26:16] that there's time to then reach out to those people do you see a necessity at one point to try

[00:26:23] to change that behavior by I'll say death clock that's not the right phrase or but you know I'm say like

[00:26:32] you can render the five candidates and say these five are the best five for what you're looking for

[00:26:39] but if they don't get in touch with them right away you know I mean like right away

[00:26:45] seconds and minutes right away then there's there's a chance of it's okay three weeks from now two

[00:26:50] weeks from now they go back to those candidates those candidates are already on to something else

[00:26:55] totally 100 percent this is do you know what I'm saying like I call it death clock which is a horrible

[00:27:02] but it's like I want them to expire I want to get I want to give the I want to give the recruiters milk

[00:27:07] and say yeah expires in three days there you go I mean what we're talking about is and I don't want

[00:27:15] to use this term necessarily was I'm just really really helping us on connect the dots from

[00:27:24] a right how should we think about perishable at any time yeah anything special right so perfect

[00:27:31] so let's take an example that we all do because we all travel and we book flights we book hotels

[00:27:39] very non-stop I love it I love it and if you if you go and if if you look at whatever your favorite

[00:27:46] side is like two of my favorite sites for booking is actually booking.com and Airbnb

[00:27:53] and so you go in there for hotel booking you're going there and you basically understand that

[00:27:59] if we don't go these things are gone and they tell you how many people are looking at it they give

[00:28:04] you sense of the number of yeah so that's the same story here like very same story here so our goal

[00:28:13] ultimately is to nudge employers like hiring managers and recruiters to understand that people

[00:28:20] are going to move on they have other opportunities I guess let's look at the macro like let's look

[00:28:30] today's job report that came out the labor market is still tight very tight yeah so if you are

[00:28:38] going to sit and not reach out to these candidates the valuable candidates that we surfaced up

[00:28:43] they're going to go somewhere else somebody else will reach out to them that's right yeah I love both

[00:28:49] of those examples both the perishable talking about perishables thinking about perishable candidates are

[00:28:54] perishable I think that could change that could really help to change the behaviors of the

[00:29:01] recruiters thinking that they're always going to be available are you actually then presenting them

[00:29:07] with there's this candidate's applied to x number of jobs and has x number of conversations are

[00:29:15] you presenting them with that information so we are going to so some of that we are starting to but

[00:29:22] yes the idea here is very simple we want to help employers understand that this candidate is in

[00:29:28] demand and whatever the demands signals might be and that's really important that's really important

[00:29:34] how you could do I move because you're sitting on the data it's how many people have looked at their

[00:29:38] profile you know again how many people they're interacting with currently like like again

[00:29:46] you can treat that perishable in so many different ways but it's giving them that data

[00:29:51] to then change the behavior yeah so it does yeah that's perfect good for work upwork

[00:29:59] upwork does this with their with their gigs as as your plot you see x number proposals have been

[00:30:06] submitted and you need to make that decision as a freelancer do I want to compete with this right

[00:30:14] am I only going to go for ones that have zero to five proposals or 20 to 30 like how do you

[00:30:19] know and you got you have to make that decision but you use your credits wisely yeah yeah right so the

[00:30:27] recruiters will have to use their time wisely and say well Roger applied for the opportunity

[00:30:33] but he's also got 17 conversations I really like Roger maybe maybe I don't like him so much

[00:30:40] let's move on to the next right I mean it I love the idea but it gives him the sense of hey they have

[00:30:47] to move quickly yeah I thought when you were first going into the flights deal I book everything

[00:30:53] through American Airlines because they're in DFW their headquartered here yeah but when you when

[00:31:00] you book you know that itinerary is only good for 24 hours yeah like I mean you know that's it

[00:31:07] and so let me save a flight okay I like that let me save a flight I get 24 hours in my

[00:31:13] decision because after that he's gone that deal is like 10 minutes and you're done

[00:31:20] but it's like that deal was gone yeah it is if we can re rearrange that thinking for the corporate

[00:31:28] side that hey these they're they're gone after whatever period of time or whatever I think it's great

[00:31:35] I think that again whenever you launch that we'll have to do a different podcast about that because

[00:31:39] I think just the behavioral change because we used to think of I still think a lot of

[00:31:46] high managers thinking this way where there's a candidate driven market and employer driven market

[00:31:52] and that's usually tied to recessions and things like that there's no such thing anymore it's

[00:31:58] all a candidate driven market so it is it is the you know Sponda don't get that you don't get

[00:32:06] that talent you might get at somebody else right with it with a lower batch or whatever but

[00:32:12] you'll get that person demographic shift is very clear in this country it's clear in every

[00:32:19] Western economy and if we if we we're actually very important if you know recruit which is the

[00:32:26] overall mother company yeah they're based out of Japan and so because of our tight partnership

[00:32:32] with them we know a lot about what's going on in the Japanese market right and you don't

[00:32:37] have to look that part like the Japanese market it's a shrinking labor pool people don't change jobs

[00:32:42] as often in Japan there's also an aging population it's an aging population right so and in US

[00:32:50] and I'm not going to be able to talk about immigration or anything I don't know much about it

[00:32:55] like to be honest on what's going on but I can certainly tell you that our overall population is

[00:33:02] that's just happening and so if you have shrinking population our aging population we're not

[00:33:08] going to be working and then you don't have an inflow of people coming into the country you're

[00:33:12] eventually going to have a smaller labor pool that's just going to happen which drive

[00:33:18] what's great for candidates because it drives prices up but it sells sell sell sell so again

[00:33:24] you might because of the shrinking pool you might as an employer you might have to deal with

[00:33:30] not the best candidates because you know you just don't have as many of them yeah that's very

[00:33:37] important because the best quality candidates may be taken out of the market by others so you

[00:33:42] the I mean really what we are talking about is we want active candidates candidates who are

[00:33:47] serious about looking for a new opportunity there's no point in reaching out to many many passive

[00:33:52] candidates which is a recruiters brain like I talk to our internal recruiters external recruiters

[00:33:57] everybody come here I reach out to so many people and they never respond back or they respond back

[00:34:02] negatively so how do you reach out to people who will respond positively so that's one part of

[00:34:07] matching the second part of matching is how do you reach out to the right people the people who have

[00:34:12] the appropriate qualifications and the people who are likely to be the ones who are going to take

[00:34:18] your opportunity seriously so I think that's the second part and then what channels and communications

[00:34:24] we use to really have that fast connection so now you know I'm bringing it all back to this whole

[00:34:30] notion of you need they're called the highest reach and I think Ryan you were talking about is like

[00:34:36] the highest reach of your seekers we want to talk about the best matching and we want to talk

[00:34:41] about fast connections if you have these three and that's really what differentiates smart sourcing

[00:34:46] you have that highest reach of active candidates you have this awesome matching engine that's allowing

[00:34:52] you to get recommendation based no longer search which means smarter and smarter and smarter over time

[00:34:59] yeah exactly you'll get smarter and smarter and smarter over time and your fast connections have

[00:35:04] any on the platform you know what you haven't you haven't talked about yet Raj is how this actually

[00:35:12] helps you know it's two-sided marketplace right it but this this particular I want to say

[00:35:19] feature but this particular product it helps both sides the kind of experience and the recruiter

[00:35:26] experience so both of them get better so the candidates get better matched to better judge the

[00:35:33] jobs that kind of fit them so it serves up things for them they get connected to people that desire

[00:35:40] their skillset et cetera and for recruiters they don't have to waste time which is frustrating but

[00:35:47] also they get to interact with candidates that are better fits so I can actually care about the

[00:35:53] opportunity yeah yeah it's like both sides I can see their happiness scores right both of them go

[00:36:00] up because it's like indeed is solving this problem I don't know how they're solving it but

[00:36:05] indeed solving this problem where I'm now connected with people that want me desire my skills

[00:36:12] and recruiters are now not wasting that time and getting frustrated and talking to people that have

[00:36:18] the skills like again not everyone over one of them is going to be the person that you pick for

[00:36:22] the job but their happiness is going to go up on both sides 100% and this is the heart of what we

[00:36:32] are saying make hiring simpler faster and more human bring the human back in hiring right we should

[00:36:38] technology should help us not make our lives more complex and so at the job seeker side you

[00:36:44] talked about that I'll share a couple of stats that might maybe new to you all but also would be

[00:36:49] something that you may already know so number one more than half of job seekers more than 50%

[00:36:57] of job seekers say they're super frustrated with slow process they're complaining about it

[00:37:02] what was the percentage more than 50 oh yeah that's that's more like 90 I got you yeah

[00:37:13] I would like who isn't frustrated by the by the slowness of the process I love it all

[00:37:21] when you're making ribs that's one thing but no this is this is so you know get it they're frustrated

[00:37:27] their frustration comes from I did hear from them or if I heard from them it took forever and oh by

[00:37:34] the way it wasn't a great match I think they're getting back to customer service example earlier you

[00:37:40] know they want customer service in the context of this exactly I've got great customer experience but

[00:37:46] it's in the context of this yes okay sorry dinner no no and and 40% of them this is also another

[00:37:55] step that just recently learned 40% of them have been contacted by a company for a job

[00:38:00] there was just not right for them this is this is the other so job seekers are frustrated

[00:38:07] and and we are always job seekers first and yes we have built smart sourcing as a tool for employers

[00:38:13] and we think it'll really change the game and sourcing but it is also really putting the job

[00:38:19] secret at the heart of it and our when we talk about matching and I don't remember one of you

[00:38:26] were talking about this concept of what exactly we do in matching the preferences part right where

[00:38:33] your job seekers is I really want to be reached out for this location or I do not want to receive

[00:38:40] a job reach out where the salary is less than $50,000 right whatever it is those things are being

[00:38:47] honored and that's a key thing like we're not reaching out to job seekers really really yeah it could

[00:38:54] be a perfect match but if the candidate wants to do remote right it's a perfect match but you have

[00:39:02] you have a policy where it's three days in the office or whatever the bit is yes it's got a hybrid

[00:39:06] model it could be a perfect match but it's not a perfect match if it doesn't meet the right

[00:39:11] meet or exceed the criteria of the candidate exactly and employers want to know that too they

[00:39:17] too they're coming back to the wastage of time right once that I forgot to share about smart

[00:39:22] sourcing because I think this is actually going to be powerful for the users to understand with smart

[00:39:27] sourcing we have been able to save in the hiring process more than six hours per road right right

[00:39:34] for me basically for a week you're saving my power and six hours yeah if you're carrying heavy

[00:39:42] load which most recruiters are right now because they're doing more with less so they're carrying

[00:39:47] a lot of roles that adds up you know what I'd love to see and I don't know how we do the math

[00:39:52] but I'd love to see kind of at the bottom of the website how many hours of time that we're saving people

[00:39:59] yeah for every search for every search there's hours that are saved I would love to see this

[00:40:04] let's use it's like this odometer clock yeah I'd just love to see that because again that helps everybody

[00:40:11] uh I'm going to take a note of more I'm actually writing it down that's a really cool idea it's funny

[00:40:16] and funny about that will you what at the gym they got the water fountain where you fill the water bottle

[00:40:23] and it tells you how many bottle plastic bottles yeah and I gotta tell you I want to know day I go

[00:40:30] and get water I look at it I'm like it's gonna hit 500,000 next week yeah and I watch it take up

[00:40:37] and up it's I want to know I want to do it I want to know how many hours it saves

[00:40:42] and as a recruiter so if the if the cell is hard for a corporate recruiter or an employer

[00:40:50] this is real simple it's better matches faster yeah we're not gonna waste and we're saving you time

[00:40:55] which is actually saving money get right down to it right time is money so we're saving you those things

[00:41:02] and also I think that frustration level which is you know a different type of math where we're like

[00:41:09] hey listen you can actually love your job which I don't think recruiters fully understand it you

[00:41:14] can love their job because everything's been so painful for so long yes yes so like taking this out

[00:41:20] like okay it serves up candidates that are actually going to really like the call and be able to

[00:41:27] interact etc. Roger I can't say enough this is just this is just wonderful absolutely wonderful thank you

[00:41:33] thank you and we have a lot of distance to still cover oh oh yeah it's never done

[00:41:38] much come on you launch this and you're already on to the next thing I mean it's just it's not like

[00:41:45] you get time off that's that's but it's also how y'all continue to be the the innovators in the space

[00:41:52] I mean you could have had any point said you know what pause were good you know let's just kind

[00:41:56] of being a holding pattern for you know a couple years you didn't so I mean that's why

[00:42:02] that's why you're you've remained not just relevant but you've remained the place that people have

[00:42:09] to go to to get innovation so yeah I want to do your team yeah yeah I actually want to maybe just

[00:42:18] tweak a little word that you said yeah have to go to want to go well I didn't want people

[00:42:24] I really like it will do yeah that's a great point because it gets back to that frustration stuff the

[00:42:28] happiness factor then then it isn't a have to it's a one two I want to go to indeed yes you know

[00:42:35] I could go somewhere else but I want to go to indeed yeah you're absolutely right good call good

[00:42:39] call Roger thank you so much for coming on the podcast and I'll hop busy or thank you Ryan thank you

[00:42:45] thank you for carving out time for us our audience