Marketplace Musings: Searching But Not Finding
I’ve noticed a curious shift in thinking with staffing and recruiting professionals who have expressed disappointment in their ability to search and match candidates in their database with their needs. The candidate shortage is certainly a contributing factor, as fewer applications leads to increased use of your internal database to fill roles, but I don’t think that’s the entire story here.
Why is there is such a disconnect between the tools we have available to us and recruiters’ high expectations of search results? While advancements have been made across technologies, recruiters are wondering why those advancements don’t yield more precise search and match results.
Here are the recent advancements that I believe that have led to this logical conclusion:
- Availability: In today’s world, being “available” for work can mean different things to different people, and factoring in availability accurately is a significant challenge for any company. Look at poor redeployment percentages and the difficulty in getting accurate data around actual end dates. Very few projects have been timeblocked since Y2K efforts. For instance, “I am always available for a $50K a month position innovating, otherwise, it depends,” is difficult to build automation and search terms around.
- Business Analytics advancements: Companies have made significant improvements with their BI tools and can provide improved visualization of their data and trends. Many of these companies have funded teams of data scientists who have cleaned and visualized their data in order to see what is happening across their organizations. This improved data should indirectly help search/match.
- Candidates abound, but there’s limited ability to rank them: Most companies are likely contracting to multiple services to receive candidates—both from posting job orders on job boards and advertising with similar or additional PPC providers in an attempt to reach their desired audience more quickly. Compound this for several years and you’ll have an enormous candidate database.
- Machine Learning and Artificial Intelligence: These technologies have created a gold rush effect, with everyone interested in how quickly they deploy these products for their enterprise. The issue is that quick solutions only met a relatively small need around basic qualifications, while adoption at the candidate-level remained relatively low as well. Just ask those Waymo cars being stoned how people feel about automation impacting them. To expand into the specific automation functionality recruitment leaders are seeking, it may take additional “training” for that tool or very large sets of results from which to extrapolate answers. Simply put, you are employing a chatbot and they need training just like the rest of your workforce.
However, companies often lack the tools to properly manage these large pools of individuals, and the awareness to differentiate an excellent candidate versus an average candidate. The search function works well, but delivers mixed results against mixed expectations. Additionally, the match comes after the return of the search results, so additional filtering is typically required. This can be impacted by geographic conditions (e.g. rural areas with less available talent) and many other skill gap and data quality issues as well.
Combining all of these factors seems to have led to an idea across our industry that people should be able to match candidates far better than their capabilities allow today. But the majority of search and match tools were built in technologies that originated in the year 2000. I do not wish to imply that search tools available today are not working; they are mature and very capable, they have just had the finish line moved on them
Bullhorn has many clients using TextKernel, Daxtra, SourceBreaker, and other solutions, and are seeing improved results over standard offerings. But many clients are also looking for tools that serve all their needs, across all of their touchpoints, rolled into one device that returns more specific results. This conventional wisdom seems to suggest that collectively, the advancements listed above should be organically leading to better matches when searching.
Imagine how much more specific search and match can become if candidate searches were using all the historical data you have collected, the learnings of past hunts and results of those searches, who ranked well but just wasn’t hired, whether they follow you or your employees across social media, and if they’ve visited your site four times a month for a quarter. I sure look forward to those days, and I’m confident our clients will have new needs by that time as well. Innovations can never stop and demand for what’s next is always increasing.
Bullhorn will continue to build new technologies or partner with companies that can solve these needs for the enterprise client as quickly as we can. In the meantime, follow your existing recruiting process and see where it can be improved from today’s tools, but keep a pulse on what’s available. Match is yet to be solved, but we’re getting there.