- Understanding the impact of duplicate data on your Bullhorn CRM.
- Step-by-step methods to identify and merge duplicate records.
- Leveraging automation and AI-driven solutions to streamline duplicate management.
- Maintaining data integrity and enhancing user experience.
- Real-world success stories and case studies from Bullhorn CRM users.
Hello everyone, thanks for joining us today.
We are going to be talking about streamlining your Bullhorn CRM and mastering duplicate data management within Bullhorn. My name is Heather Pearson with AICG and today we have Mike Jelen from AICG who will be hosting this event.
Thank you, Heather, and thank you everybody for taking time out of your day today or if you’re watching this recording in the future. Thank you for watching. So just to talk a bit more about this AICG and really from a founding perspective, we as an organization have been in the data analytics space for about 20 years and really helping, the technology is interesting, but its always helping customers solve their business challenges or take advantage of business opportunities and all things along the way here, but really all about data and the analytics.
We’ll be showcasing a product from one of our partners called data lakehouse .io that helps facilitate specifically the Bullhorn customers out there, the users of Bullhorn and more importantly the business owners, the leadership or operations manager within those When it comes to Bullhorn, we see some common challenges across our customer base here.
Data quality is a big part of that consistency from a pain point perspective. And it’s really the data qualities you have completeness. People not entering in all of the information that is needed. And then also duplicates, especially faster growing organizations that have been established for a while, we have lots and lots of contacts or candidates or companies in there.
It’s not uncommon that you get duplicates in there. Perhaps somebody forgot the look. Somebody thought they were entering in the appropriate name of a candidate. And now over time, all those little kinds of nuances of those little mistakes add up and it can become a bit of a challenge within Bullhorn.
When you’re trying to run your business, make decisions based upon certain candidates, find the right candidates for your opportunities, your full service desk. But it really just falls into an overall sales process of how should people be entering in data from a completeness and a quality perspective that follows the sales process that you have defined for your organization.
And understand that Bullhorn isn’t the only tool or application that’s part of that sales process, but it is a key component to that process. And you got to make sure, especially, you know, newer people to the organization that they’re following the process, so that really things are complete.
Today, traditionally, if things are incomplete or errors are found, you know, it’s manual human intervention that you happen to come across duplicates or you’re being told that there is duplicates in here is the situation.
But to find all that information programmatically, or on a report within Bullhorn is very, very limiting. We also have just even organizations that they have particular KPIs as to how they want to measure the effectiveness of their agents also is lacking within Bullhorn as well, because that might just be something that Bullhorn never thought of you have your own unique KPIs that you would like to add.
But you also like to have a one stop shop or a dashboard to be able to see that data updated in in real time. Some of the other pain Any points that’s alright, you have data in Bullhorn today, you have data in other applications, perhaps you’re downloading that data manually and trying to match that data together in order to come up with your own pseudo dashboard within Excel.
A lot of organizations we find end up spending multiple hours, one or more days a week in order to download, massage, manipulate that data and get it into the right format only to have that data be as good as it was the day that they downloaded it to go and update that again, they might wait to the following week or they wait until the end of the month and do it again.
And it’s not very, very timely opportunity cost of that individual or individuals that are involved in that process is high because they could obviously be out selling, placing candidates or doing higher value activities.
Now later in organizations that you’re adding agents, you’re adding recruiters to your mix. And now you’re trying to onboard them, but at the same time, you’ll deal with some of these manual processes behind the scenes.
And oh, by the way, you’re always adding new tools or thinking about adding tools along the way to make everyone’s job easier. But with those tools comes the need to have a process and an ability to manage that process along the way from a data accuracy, data completeness perspective.
But you don’t have a huge budget. And part of that budget is you may or may not have the luxury of having an IT team. Or perhaps that IT team is pulled in different directions that the thing that you want them to work on, at least in regards to the Bullhorn and the data and the process, is not high enough up on their priority list.
So now you are left with trying to figure out, well, how am I going to solve this or work with the limitations that I have? And so you can actually have a, should I build something myself? Should I buy something?
If I should buy something, what’s even out there? And that’s where our partner, data .lacoust .io, fills that need or that gap at the end of the day to be able to address the duplicate data. So let’s talk a bit more about duplicate data.
What do we really mean by there? So set the stage. You’ve got 10 ,000 plus candidates, contacts, companies. You might have double that amount. You might have 10x that amount or more at the end of the day.
So naturally, things compound each other, the more records that you have. And really just even the number of those candidates and contacts that you’re communicating with on a regular basis, whether you’re a high volume shop or perhaps your lower volume, higher dollar shop, really comes down to your individual situation.
At the end of the day, you have no shortage of wasted time and effort in dealing with these duplicates. Great, I identified a duplicate. I want to merge them. Well, now what? When am I going to merge?
Am I merging? What’s the new primary record and what fields am I taking from one record and putting into another? Do I need to go and talk to somebody to figure out, hey, which of these duplicates should be merged?
What is the right one? At the same time, if somebody’s doing that, who’s in on revenue from your organization or at least activities that lead to revenue generating results within your organization, but it also is this inaccurate reporting.
How many new candidates do we have? It’s one thing if you have a staff that’s self -motivated, but at times you may have some contest that you implement as well. Let’s say, all right, let’s, you know, everybody put in for the next 30 days, how many candidates can, that new candidates can be fined and add to the application.
Well, if you’re going to have a carrot, some sort of reward, that could be a financial reward, could be extra days of vacation, could be, you know, a gift certificate somewhere. But in order to say, how do I know that they’re entered in 75 new candidates and Mike entered in 50, to be able to say, were those really net new or was there some sort of variation of a name or a background that they actually already existed within Bullhorn?
And they shouldn’t perhaps get as much credit for that. You know, it’s kind of a playful example of that inaccurate reporting. So it’s one thing to identify those duplicates. Now that you have them identified, you want that human intervention part for somebody.
You know, typically it’s one person organization that’s doing the merging or has the capabilities to do the merging. And they’re your quarterback. They’re going to hopefully find a lot of these and then go and talk to individuals that may have created the records to say, Hey, is this really the same person or Tom, you created this record and four months later, Sally, you created this other records.
Which one is accurate or not? But really having that control because that control is very important. So that control is what is the remaining record that’s going to continue in Bullhorn. And then what fields are merged?
Okay, if you have some differences in the fields, say, what would you values in which record become, you know, the new primary value that you keep? And then there might be some fields in there that you don’t really care about.
I’m not interested. Sure, there might be some data on the person’s favorite color, for example. All right, that’s not really meaningful. Let’s just ignore it. And the ignoring of not just fields, but just even records as well, you might say, like, these really are not duplicates, I don’t want to continue.
to be asked and that asking could be from folks on the team, hey, this is a duplicate, can you merge it? Or for the application or for Bullhorn to tell me or some other application to tell me that, hey, this is a duplicate.
Like, no, I wanna control my destiny as to what I really have to target in regards to removing these duplicates. Cause once you go through and you have a good, stable point that says, all right, I’ve made decisions on a number of these different duplicates, I’ve taken action on them.
Some of that action might be ignore, some will be merge. I don’t wanna continue to be asked over and over about these same ignored records. So we look at a simple example here. You think of, all right, my name, born as Michael.
Most people call me Mike or just refer to the last name because Mike is such a common name. And depending upon how that information gets entered into Bullhorn and when that information gets entered in, in this particular example, there’s Mike and Michael.
And you look at, okay, the last name, all right, that’s the same. First name, all right, Mike and Michael are interchangeable. But over time, perhaps, you know, I was Mike when I was younger and I want to be Michael or vice versa, and you have a different cell phone name.
And one of those records has my resume or CV, but then the other one does not. And even though they were updated two years apart. It’s not as simple to say, All right, just take the most recent record and that’s what you’re going to go with.
In this particular case, if you went with the most recent record, you could make an argument that, All right, everything that’s recent is, is what I want. But what about the resume? Are you saying you don’t want to have the resume in this particular case, even though you have it on the old record, just, you know, throw that away or ignore it and never, never ask, ask again about this.
And so those are the types of human decisions that an operations person or someone that really owns the Bullhorn data needs to think through it but need the tool or the process that you’re leveraging needs to have that flexibility to be able to account for these situations and really that human thought process of, well, not every record is going to be merged the same way I want the flexibility to be able to go on a case by case basis.
So at the end of the day, to say, Well, all right, even the record here that says, Well, this is the final merge record that on the surface, it says, Okay, great. But to be able to make the choice that’s like, No, Mike is really what they should go by to be able to kind of pick these individuals.
and go on a field by field basis to be able to say what gets merged. So it’s the process of handling duplicates that we see some of that complexity out there that organizations need that flexibility in order to meet their needs.
So now we’ve identified and talked through about the duplicate pain point that we see within Bullhorns. How am I going to solve this problem? Then it goes back to that buy versus build that we’re talking about earlier.
Am I going to buy a tool out there that handles this for me or am I going to try to build something on my own? Well, if I’m going to build something on my own, you have to select the right tools. You have to understand how those tools fit together and you have to have the skill sets in order to stitch them together.
Okay, that can be a barrier for a lot of organizations. Say, well, now I need to go from something that’s already. created that I’m going to buy. I just want to buy a tool to help me or subscribe to a tool that’s going to help me with these duplicates.
There’s still some research that you have to do. There’s not a lot of tools out there that handle and address this particular situation, but at the same time, you want to make sure that whatever tool that is out there, that it integrates seamlessly and flawlessly within Bullhorn and it’s not going to disrupt your business and it’s going to give you the flexibility that you or your operations person needs in order to handle these duplicates here.
So that is simple when you start getting into the weeds. So we look at a particular use case where we leverage Data Lakehouse to help with the duplicates within Bullhorn. Part of it was the duplicates.
There also was just the notion of… that downloading of Bullhorn data, downloading of cell phone data, downloading of LinkedIn messaging data, downloading of name another source of data and really combining all of that onto a dashboard so that there are KPIs and metrics.
So now you have something that’s more valuable to the wider sales audience and sales leadership audience within an organization, but at the same time, all right, I still need that flexibility to address those duplicates.
So getting more out of your solution. And really at the end of the day, when you combine those together with the dashboard, that pulls from those different data sources and also gives the customer the ability to work with the duplicates, that’s when you see a real change, meaningful change within the organization, top -line growth, because people are spending more time on sales -related activities and the data is cleaner.
We’ve also created this culture of leveraging data in order to drive an openness, but also drives a bit of a competitive spirit. When you think of all sales agents, they love to compete, they wanna be one up each other, whether that’s within a week, within a day, within a month, within a year.
Hey, there is that natural competitiveness that you want and you want to instill within your organization and using data to help instill that makes a huge difference. But also gives you that time of adding more agents and knowing that you have an ability to measure their progress as they ramp up.
But you also have, get them into that competitive vibe of, hey, Tom, you were here for three months and your productivity and you following the process is twice as good as someone has been here for one year.
Like, well, why is that? There’s the cheeky. kind of hope fun of each other, but there’s also from a manager perspective, they’ll be able to say, well, what’s going on here? Why is the new person able to follow a process so much better than someone that’s been here for a while?
But that allows you to scale and to grow and to get more throughput, more productivity on the resources that you have. And then as you add that new resources, that allows those folks to be productive quicker.
So if we go and think about, well, how did we do this? We as AICG, when you think of, how do you approach that type of scenario? It’s really all about turning the data into a larger asset. The data by itself is an asset, whether it’s in Bullhorn, some other application, the data there, it has value to it.
We’re just trying to increase that value to you, but we need to do that in an easy to access, but then also every day that I go in, or whenever I go in, I just need to see the most recent, most updated data.
Maybe I don’t go in every day and that’s okay. However, when I do, I don’t want to know, oh, this is actually from, this data is from last week or last month. Can’t have that. So it needs to be updated in real time, but it also needs to be a friendly user experience.
This is not an IT project that, oh boy, this thing works, but it’s really hard to work with. You want this to be easy, whether it’s an individual sales agent that goes in and is able to look at their dashboard and interact with that dashboard to find out their individual productivity and to see how they stack up against their peers.
So all of those different variables need to come into consideration, and that’s where DiddleyCouse came in. It was that perfect fit for the situation here. And when we think of that perfect fit, we had data, it in multiple sources, as the customer did, where we had the data in Bullhorn.
We had some data in Verizon. They were company -issued cell phones were able to at least get call log of what’s the phone number, when, how long, and what’s the text messages were as well. Now, the phone numbers onto themselves don’t really help.
It’s the combining that information with the data that’s in your Bullhorn to be able to say, oh, all right, now I know who you actually spoke with. And are they a candidate? Or are they a contact? And for what opportunities and what sort of revenues attached to that?
So now you’re able to just to unlock some of that information. And so we tap into those different data sources. And then we go and look at this data on a dashboard to be able to say, all right, let me look at my team as a whole.
put yourself as an owner operator, perhaps you’re a leader within your organization, and to say, oh, all right, I can see how my team did last week, for example. And to say, okay, the number of calls, how long were those calls is one thing to leave a bunch of voicemails.
Interesting. I know that’s part of the business, that there’s naturally the back and forth of trying to track down somebody, a potential candidate, or someone at a company, to talk about an opportunity or just to catch up.
And then just on a daily basis to be able to understand productivity for those individuals. So if you look at last week, we can see, oh, all right, on the 15th, that we have some activities that were occurring, that it’s varying activity.
We have somebody with this L -WIN here, at 47 calls, and it lasted 126 minutes. Okay, there’s some productive calls in there, they’re not just a bunch of voicemails. I didn’t send any text messages, which that’s kind of weird.
And then they logged calls. So when you think of logging calls, this is now where you’re tying that call data into your Bullhorn CRM to be able to say, are they following the process? Just because somebody made a call, if you only have Bullhorn and your cell phone data, there needs to be that manual process of making sure that that was logged.
Now, if you leverage a tool such as Ring Central, that gives you a little bit more control over that process. But because we are referring and diving into the Bullhorn, we know there’s needs to be a manual process that L -WIN would need to go in and mark as such, put a note with an S.
action that says, all right, here’s who I spoke to. There’s what we talked about, but this is really the results of that conversation. So if we say, well, let’s really zero in a little bit more on Elwin here, and we’re going to just select him, and it’s gonna be all about Elwin here.
And so the data is going to match your refresh here once it gets on running. Well, then we also see, well, Elwin had two days of phone calls and communications. That’s kind of weird. What was Elwin on vacation?
Was Elwin at a conference? Was Elwin doing something that he should not have been doing? So now you can use this as a management tool to be able to coach and understand what’s Elwin’s week look like?
Well, he had two days that he was on the phone, at least with his company phone, if this is a company issued cell phone. It’s also not big on text messages. Well, OK, interesting. So we can go down and just say, all right, well, let’s go look at what’s the call by call, or if you did have text messaging as well, what’s text by text, and how long were these conversations?
So when you look at the notes just around who they were talking to when, we have two minutes, three minutes. When you think of two or three minutes, all right, one minute being the least amount, anything less than a couple of minutes typically indicates that you just left a really long voicemail longer than five minutes.
OK, this was a potentially a meaningful contact, a meaningful connection. We think about the number of minutes that Elwynn spoke to somebody. So we’re starting to look at, wow, there’s a lot of voicemails here.
It sure made a lot of calls. This is like, OK, we got a half hour, almost half hour, 20 minute call, half hour, boy, there’s a lot of voicemails here. So now you can have a conversation with Elwynn to say, what’s your prospecting, or what are you going after?
Were you leaving a lot of voicemails and you’re having some productive calls on the couple of days that you worked? And you can, again, use this from a coaching perspective versus, hey, this is just a stick that I’m going to tell you, here’s what you did, but why didn’t you do more?
Like, no, let’s understand what you’re doing. Is Elwynn new? Is Elwynn going through something on the home front? Does Elwynn need to be put on a performance improvement plan to get their act together to be more productive in the workspace?
So that’s just one aspect of this dashboard and how you can use data. But we also have, from an auditing perspective, to be able to come in and say, all right, let’s just focus on Bullhorn. Let’s have to better understand what data is in there, how good is the data and other opportunities for us to clean up the data or improve the data.
So when you think of, all right, I’ve got contacts that have, you’re looking at phone numbers and how many times that happens. There might be a legitimate situation. You think of a direct company number, all right?
But then also, there’s just fine occurrence that happened five times. If somebody does not follow the process, is there an issue with those five records? Maybe they’re garbage records. They came in from who knows where.
They just need to be cleaned out. Well, then even look at, let’s look at duplicates for data quality from a different angle. Let’s look at it from you have contacts. and people that are entered, but they don’t have a phone, nor do they have an email.
That’s not very helpful to have a contact that there’s actually no way to contact them. So why? Is it a data entry issue? Is it a training issue? Or again, is there just some garbage that came in? Maybe there’s some mass file upload or some process that didn’t go well back in the day that needs to be addressed.
So just really giving you that laser focused ability and even with the phone number, we see the five duplicates here. Like, all right, it looks like it’s the same person, but they’re client contact or candidate, which is an uncommon that you get either somebody that starts out as a client contact, they quit, they become the same independent contractor and you’re working with them or vice versa, but they’re at the same company the whole time.
Like, okay, that’s a little bit suspicious with the data here. And you’re naturally gonna have some other situations where it’s not quite as obvious as to, you know, what’s the key that they are duplicates or not?
So when we get into, we now wanna take this to step level or step further, to be able to say, what sort of intelligence can we apply to determine what might be a duplicate or what might not be duplicate records?
So to apply a formula or a methodology for that. So this is where we start scoring the data. And this is something within the data lake house that we apply a score and typically that score looks at each individual record and says, okay, is there another record out here that has X% of the same information?
It could be the same name, could be the same LinkedIn URL. If some sort of data, data element, a combination of data elements that say, yeah, these are likely or highly likely that these are duplicates here.
So now you’re leveraging technology in order to bubble some of these up to the surface. And so the higher the the rating or the higher the likelihood that they are duplicates, then the more likely they are, if you get 100% match, well, that means there’s two exactly the same records in Bullhorn, they may just have a different Bullhorn idea.
And that’s the only difference at the end of the day, like, okay, for whatever reason, this is 100% match, I don’t need to take some action. And on the other end, you know, if something is a 10% match, like I might have been a focus on though, it’s not worth my time.
It’s that churn in the middle, you get that 100% especially you get down to that 60 70 50% chance that there’s a match. Alright, a human really needs to spend some time initially to clear those out. And once they’re cleared out, then you’re only dealing with that new data.
that gets updated or added into Bullburn. So the ability to essentially go in and say, all right, I’ve got these four records here. These are ones that I want to merge, you as the operations person or the leader within your organization, you wanna be able to essentially click a button that says, all right, give me my options as to how I want to merge these.
Which one is going to remain when I say the option? Which fields am I going to be able to say, keep this one, keep this one, I’m on these different records. So that when you get done, you have the one record that remains within Bullhorn.
And now you have cleaner data that you have more confidence in. But at the same time, that doesn’t mean somebody tomorrow goes in and enters a record that is strikingly similar to the record you just merged today.
Oh, great, the Lakehouse is gonna find that new record and bubble and put it back in front of you. That says, hey, I’ve got a 90% confidence that this record is a duplicate. Need human intervention to take care of this.
So you click in buttons, you’re able to do and what data lake house does daily house go back into Bullhorn and make the appropriate adjustments to that remaining record in Bullhorn. So if that’s a particular contact, that that contact will, you’ll have one record of that individual contact in Bullhorn and no longer the four individual records within Bullhorn.
It has all the data that you want to keep. And definitely understand as part of what data should I keep or which data is the record that I really want. you it’s going to come down to perhaps the human intervention needs to go talk to the people that originally created the record to say, hey, what’s the deal with this record?
And let me talk you through the situation and make a determination or perhaps you have to go back out on LinkedIn and and say, all right, really, where is this person? You look at the resume to be able to say, what is the right piece of information?
However, to be able to get to that point, you leveraging data lake house in order to do so. Got it coming back to here. So when you think of how this would look from a pictorial perspective behind the scenes, you have your Bullhorn source, you have data lake house in the middle there.
And really it’s an operations or manager that is. determining what records are duplicates, going through that merge process. But then you also have, you may have agents that log in, they’re just looking at the dashboard.
And all they look at is the dashboard with the metrics and the KPI. But the experience of whether you’re a manager looking at metrics and manager doing the merge process, it’s the same place, it’s the same underlying data that’s being leveraged.
And so that end user experience is much greater because of that. So we’re talking a lot about Bullhorn here. Folks on the call, you may have Bullhorn and other sources. You may have other sources out together that aren’t even Bullhorn.
It doesn’t matter at the end of the day from the perspective of you have your data that is specific to your staffing and recruiting agency. And the challenges that you have, really the tool is just the vehicle to carry those challenges forward.
It’s really all about getting that data into a data cloud so that you can have that consistent experience from centralized data and centralized dashboards and the vehicle that allows you to handle those duplicated records.
So go ahead and open up for questions. Let’s go for you to type them in. You won’t be able to have a verbal question, but you can have a written question. Alright, so let’s see. Got some questions coming in here.
I’ll answer them as they come in. The only order is just the order they came in. Alright, so the first question is, do you handle LinkedIn messaging? I’m presuming it’s each individual agent has their own individual personal login to LinkedIn, that they’re doing their reach outs to yes, data lake house can handle that to connect to where each individual would be able to log in to data lake house and tell data lake house and really the big great security that we have behind the scenes, how to connect to their LinkedIn account to be able to identify what are those messages that are being sent, who are they being sent to so you can you can see call log information, texting information, things you’ve done in Bullhorn and things you’ve done within LinkedIn messaging and all that’s that’s tied together.
Another question is it relates to the the phone. What if we don’t have company issued cell phones, you don’t have one overall master plan cell phone plan that you Hey, for all of your employees’ cell phones, perhaps you do just individual monthly reimbursement for like perhaps as a flat fee.
Like, all right, we’ll reimburse $250 bucks for everybody’s monthly cell phone. In that particular situation, you know, so one, we do handle the, let’s say the company issued cell phone plan. And then the company reimbursed individual cell phone plan, we handle that as well, much like the LinkedIn, each individual will have their own cell phone.
But you do have to have a little bit of a conversation around, okay, is somebody’s personal cell phone? Are they willing or are you interested in bringing in their cell phone data if they’re calling Ticketmaster to book a show at a concert to remove that noise?
Those are things and techniques that within Data Lakehouse can handle to say, all right, if a number doesn’t come up in Bullhorn, for example, then don’t show it within the dashboard because we’ll assume that that’s a personal phone call.
So you can do it, just have to kind of think if there’s not a technical limitation, it’s more of a, you know, out of principle from a business perspective. All right, so if anyone else has any other questions, feel free to reach out to us on social media, our website, and or even just respond to where you signed up here today or how you’re viewing this, you know, if you’re online in a future date where perhaps you’re watching a replay of this, go to aicg .com, look us up, out there, connect with us on social media, we’re happy to engage in any conversations that you have.
Thank you everybody for attending today. We appreciate your time and we look forward to chatting with you very soon. Have a good day!