Recruiters might be missing the perfect candidate that is already in their CRM or ATS. Sourcing existing candidates from your database should be easy, yet due to the limitations of using Boolean search technology, recruiters tend to focus on acquiring talent from job boards and social media sites. This means you need to allocate more time and money to sourcing candidates externally, whilst not maximising the budget you have already spent on candidate data.
In this video masterclass hosted by Louise Triance of UK Recruiter, David Mercer - Daxtra Head of Sales EMEA - explains why recruiters experience challenges that prevent them from optimising their CRM database. Let’s take a closer look at these challenges and see how AI can help solve them so you can make more out of your existing candidate data.
What’s wrong with CRM search?
Most CRMs built-in search functions use structured data to filter and return existing candidates in your CRM/ or ATS database. You can usually search by job title, skill codes, attributes, and tags.
Yet there might be a few issues with this approach:
Challenge #1: Missing information - you’re only as good as the data you have
When you're searching your database, missing information on the candidates' records means that these candidates will not surface in your search. Unpopulated database fields will simply not show up in the results. Therefore, you might be missing relevant candidates - or even the best candidate - simply because their information was not properly loaded into the CRM.
Challenge #2: Manual skill coding does not always get done
Manually coding and tagging candidates is not always done properly because consultants can’t afford to spend time perfectly skill-coding each individual candidate. Not only is this time-consuming and so cuts into the productivity of the team, but might also be inconsistent as human subjectivity and bias comes into play.
Challenge #3: Automated skill coding can be inaccurate and out of date
Even if some databases can automate the coding of new candidates, in most cases it does not code the candidate perfectly. It uses a very basic algorithm to see if a candidate has a particular skill or not, which often leads to incorrect categorisation. The algorithm also goes out of date, and so your team will still need to manually update the terms as new synonyms and aliases for skills and job titles develop.
Challenge #4: Results are not ranked
The CRM search returns a list of candidates who have the searched term in their data without there being any sort of ranking. This means recruiters will need to spend extra time reviewing the candidates and find the ones who are the best for the role.
Boolean search - better but not the best
Boolean search is a method of logic that “allows you to combine words and phrases using the words AND, OR, NOT (known as Boolean operators) to limit, broaden, or define your search”. In recruitment, this process is used to either narrow or widen your candidate search on job boards and social networks, as well as your ATS/CRM databases.
This is supposed to help recruiters narrow down the pool of relevant candidates since it is really precise when the Boolean string is coded properly. However, this can create new challenges.
Challenge #1: Recruiters are not hackers!
It is difficult to build a Boolean query! To use Boolean search, recruiters need to code a string of operative words into a query where a single missing bracket, speech mark, or operative word can really skew the results. Even the most senior recruiters need to dedicate a significant amount of time to check over their Boolean strings and make sure they are coded exactly right - it is often difficult to identify any mistakes or typos made in the Boolean string.
Using such a complex search method puts the brakes on the productivity of senior recruiters. Yet, it also increases the time it takes trainee recruiters to start making placements. The aim is to reduce the time between training to billing - helping trainee recruiters to start making placements as soon as possible. However, not only does learning to use Boolean search efficiently take a long time, it also cuts into the time available to learn valuable skills such as building relationships with candidates and clients.
So why make searching your CRM more complicated than it needs to be?
Challenge #2: Does that word exist in the CV, yes or no?
Supposing that your recruiters are expert Boolean searchers, this still does not avoid the problem of term repetition in a CV. Boolean searches are binary - all CVs that mention the searched term are included in the results while CVs that do not mention the term are excluded.
This is a problem when it comes to finding the most relevant candidates.
If you are looking for a candidate who has experience as a product manager, you will probably use the term ‘product manager’ in your Boolean query. However, this would return all CVs that simply mention the term, and so your results would include candidates who mention they ‘report to the product manager’ in their CV. This means your team needs to manually sift through the results to find the relevant candidates, instead of reaching out and engaging with candidates to start making placements.
Ultimately, a Boolean query only knows if the word exists in the CV and cannot deduce from the context whether the word is used in the same way as it is intended in the search. Just as you do not want a Google search to provide you with irrelevant content, searching your CRM should return relevant results that streamlines your teams’ productivity.
Challenge #3: Ranking quantity over quality
Similarly to using the CRM search, a Boolean search does not return a list of candidates that is intelligently ranked by suitability for the role. Even if it does rank candidates, at best your results come back ranked based on how many times the searched term has appeared in the CV.
So, when searching for a candidate with product management experience, it may rank a candidate that mentions ‘reports to product manager’ 50 times on the CV over a candidate whose job title ‘product manager’ is mentioned once. This also means that candidates using jargon to stuff their CVs will be prioritised, who are not necessarily the best and most qualified candidates for the role.
AI-powered search- get the most from your database
There are four main ways that AI-powered search like Daxtra’s Search Nexus can make searching your CRM database easy and efficient.
Using Machine Learning and Natural Language Processing, Daxtra’s AI-powered search understands the context of the searched term. You then have the option to use 4 techniques that help maximise your candidate search:
Term categorisation: Understands what the term means, not just what it says.It might be helpful to differentiate between terms that refer to the candidate's job titles, skills, and companies. For example, if you are searching for a candidate who has experience working at Microsoft, you are not necessarily looking for candidates who have Microsoft Office listed as a skill. Term categorisation lets you define what the term is - in this example, a company - and so will not return candidates just because they have Microsoft Excel listed as a skill.
Term expansion: Find relevant candidates, no matter how they describe their jobs.
Recruitment terminology is constantly evolving and candidates use many different terms to mean the same job or skill. For example, a candidate suitable for a product manager role might also have described themselves as a ‘PM/ product director/ project officer’. Our term expansion feature lets you cast a wider net with the same relevancy as it automatically includes alternative terms for the same job title.
Term proficiency: Specify the exact level of experience needed to fit the job criteria
Daxtra’s Search Nexus has a search filter that lets you include years of experience in your search, meaning you can filter out candidates that don’t meet the required experience of the job criteria. Therefore, if you are looking for a product manager with 5 years of experience, it will exclude candidates with fewer than 5 years experience from the results - making it quicker and easier to get a relevant shortlist together to send to clients.
Intelligent ranking: Surfaces the best talent intelligently, just as a human would.
An AI search returns the candidates in order of the most relevance to the search, ranking them based on the criteria you specified, taking into account how much experience that candidate has and how recent that experience is. This means that your team spends less time trawling through lists of irrelevant candidates, and more time reaching out and engaging with the best matches for the job.
A bright future for search
Boolean search is complicated to use, can return irrelevant candidates, and lacks intelligent ranking as it prioritises CVs based on the number of times your searched term appears. AI-powered search understands the context of the searched term and surfaces the most relevant candidates, meaning your team spends less time on manual non-billable tasks and more time engaging with candidates.
Find out more about how Daxtra Search Nexus can help you optimise your candidate search.