The Dark Side of Online Job Advertising: The Hidden Cost of Free

Talent Tech Labs’ mission is to raise the state of the art in recruitment technology. We do this primarily by providing actionable intelligence around technology via research and advisory services, giving clarity to stakeholders around the industry on vendor capabilities and solution appropriateness for solving actual business problems. Sometimes this mission also requires us to shine a light on practices that we think are a negative for the industry. Recently we’ve become aware of certain practices in the job advertising industry that we think are bad for both candidates and clients alike. In this blog we hope to raise awareness of these issues. 

There’s a saying that there’s no such thing as a free lunch. It’s true for marketing pitches in Hawaii, and it’s true for online job advertising. A number of online job advertising vendors — typically job aggregators such as Lensa, Talroo, Jobalyze, et al — offer their services for “free”. While enticing at first glance, the underlying business practices and economic model that support these businesses is bad for both candidates and their potential employers, and we think these businesses may be better thought of as data resellers than job boards. This article describes how these businesses work, shows how companies and candidates may be impacted, and offers some thoughts on how to better manage job advertising given the prevalence of these practices. 

Does 'free' actually mean free when it comes to #jobads? @TalentTechLabs explains the dark side of online #jobadvertising in this article: Share on X

The primary issue happens when a vendor that does not have a direct relationship with a candidate inserts itself into the application process and attempts to capture that candidate’s data with the intention of selling or retargeting that candidate before the candidate ever completes his or her application. How could this happen? Let’s look at an example to illustrate. 

A candidate, Mary, is searching online for a job. She lands on Google’s jobs search widget, but when she clicks on a job she’s interested in, instead of being directed to the employer’s career site, she is instead routed to a job aggregator that has “scraped” the job posting and put it on its site. What this means is that the vendor has taken the job post information from some publicly available source, such as the employer’s career site, and listed it on its own site, making it look like there is a direct relationship between the employer and the board, even though often there is not. 

Before Mary can apply or continue in the process, she is prompted by the vendor to give all her information — i.e. name, address, phone number, email, resume, job preferences, etc. Mary thinks that this is an agent of the employer, and so she provides her information. At this point, Mary’s data has been harvested, and it may be sold to other job sites looking for similar candidates, or other businesses altogether. 

If Mary is lucky, she will at this point be redirected to the employer’s career site where she can begin the application process. Unfortunately, the journey is often not so simple. In many cases, Mary will instead be redirected to yet another job board where the data may have been originally scraped from (where the cycle may repeat itself multiple times), or she may ultimately be directed to a “dead link”; that is the job post that was originally scraped was subsequently pulled down by the employer, so even though Mary provided a vendor all of her information, there is no job at the end of the process to even apply to. 

The original thinking behind the aggregator model was to offer a better candidate experience by making the job search easier. Indeed originally pioneered the model, and job boards loved it because it drove traffic to their postings. With newer vendors on the scene, the model has become less about the candidate experience, and more about simply collecting as much data as possible by whatever means necessary, with the goal of using that data to retarget or sell. 

Nearly all job boards today try to collect candidate information. The distinction seems to be whether data collection happens pre- or post-apply, and how that data is used once collected. Good actors attempt to collect candidate info after they’ve gotten a candidate through the initial apply and typically use the information to provide relevant job ads for employers it has relationships with. Bad actors collect the data pre-apply and sell the data to whoever is willing to pay.

Our research suggests that these sites do not have a substantial amount of direct employer relationships, and thus we’ve seen this issue arise most often with job distribution tools and with some variants of programmatic job advertising. Both of these technologies leverage large networks of job boards, some of which engage in the kinds of practices discussed above. 

With job distribution tools, it happens because many offer “free” distribution not just to the boards a client has contracts with, but also to additional partner sites. With programmatic, it can happen when clients set a budget that is too low to be bid on by “top tier” job boards, and thus is only picked up by boards primarily interested in data harvesting and selling. For example, Equest has a product called BLAST which it positions as a tool to bolster job advertising efforts via free job boards, while Appcast’s Xcelerate product at least in part leverages such boards for candidate flow.

Outside of these tools, firms generate candidate traffic either organically via SEO or more commonly by “buying clicks”, i.e. paying other job boards to send them a candidate that clicks a job ad. Thus, even if you don’t use such a job distribution tool or programmatic solution, your job posts still well may end up on these sites. 

It’s important to note that in theory these tools can drive candidate traffic to employers’ job posts, though our conversations with practitioners suggests that this has not been an effective channel. One of the world’s largest staffing agencies reported that it actively blocks such vendors and has implemented anti-scraping technology specifically to stop vendors from using its job posts (not to mention it actively monitors its Google jobs feed to ensure that no vendors it does not do business with advertise its jobs), while a number of large programmatic job ad buyers reported having to play “whack-a-mole” to stop companies from engaging in these kinds of practices. 

Some employers will opt to use these tools, but those that don’t wish to should audit any use of “free” job board distribution tools, and understand what happens in any extended job board network that posts are being distributed to. Programmatic advertising (which is approaching nearly a billion dollars in aggregate media spend, by our estimates) return on investment seems to start when media spend exceeds 100k per year; for engagements smaller than this clients should inquire as to how reliant the vendor will be on “free” boards to drive traffic. 

At the end of the day, “free” job boards have a cost: poor candidate experience and the loss of data provenance, which can lead to silly things happening such as a candidate clicking on a link to apply for Uber instead being shown jobs at Lyft prior to them ever reaching the Uber application page. While these tools may drive some traffic to open positions, employers should go into these kinds of engagements “eyes wide open” and understand what the candidate journey can look like. 

Given shrinking recruiting budgets and the need to do more with less, the promise of cheap candidates is enticing. But sometimes the hidden cost of free is more than it’s worth. 

History of the Online Job Advertising Vertical

Online job advertising is one of the largest talent acquisition technology verticals by revenue and one of the most mature, generating in the neighborhood of $15 billion dollars globally per annum. The industry’s beginnings were career sites like HotJobs and Monster, which built direct relationships with millions of candidates and which offered employers the ability to pay to post open positions that candidates could search for and apply to. Today, there are more than 10,000 such sites globally, many of which are portfolios of brands and sites owned by a single entity. 

In the mid-2000s, two new job search models emerged. One was professional networking, which combined social media and job searching, and was pioneered by LinkedIn. The other was job aggregators, pioneered by Indeed. Job aggregators scrape the web and seek to be the “single source of truth” for all jobs across the web regardless of source; they scrape career sites and other job boards to provide job seekers a unified job search experience. 

In the 2010s, two additional job search models developed. One was job distribution; these tools were initially designed to make it easy to post jobs to multiple job boards, and later evolved to redistribute jobs to a network of partner sites to increase candidate exposure/applies, a model pioneered by ZipRecruiter. The other was Programmatic Advertising, which sought to replicate how ad buying works on the internet. These tools let employers spread job advertising dollars across multiple job boards and sites, with pricing set dynamically based on supply and demand for different kinds of candidates. 


Trends Report: Mastering Remote Work Management