By Mel Day, CEO
A couple of years ago, I got chatting to someone in charge of customer feedback at one of the big airlines. I asked how it all worked, and was a bit disappointed to hear that people who take their precious time to fill out a survey after getting one of those annoying post-purchase emails may as well just not bother.
Rather than a human reading all the feedback and there be a chance of change actually happening, all the responses go into a big spreadsheet. A bot then tallies every word of every entry – happy, angry and so on. The bot then gives a rough current RAG status to the company based on the percentages of pre-determined positive and negative words.
I am probably paraphrasing but hopefully you get the gist – I for one will never ever be tempted to fill out a survey from a big company ever again!
Well, strap yourselves in, because as I predicted a couple of years ago it’s now happening to recruitment – and it’s not all good.
While there are record level job openings in both the UK and in the US, lots of people still have to apply to sometimes hundreds of jobs, even in sought-after fields like technology marketing. Conversely, many companies are complaining that they can’t find the right talent.
The answer is bots, and their collective noun AI.
Here’s a good example of where bots are going wrong in the recruitment process.
As detailed in The Guardian’s US edition (I’m so global), a jobseeking Data Analyst recently received £8000 compensation from Bloomberg, precipitated by a ruling from the Information Commissioner’s Office (ICO) – the body in charge of GDPR.
The crux of the case was that he applied for a job with Bloomberg, was qualified for said job, but was rejected based on his performance in an online assessment asking him to spot trends in certain shapes and patterns – which had nothing to do with the requirements of the job.
His application email, CV uploading, CV scanning, assessment email, assessment results scanning and rejection email was all performed by a bot built by a company called Plum. The applicant did not meet a high enough overall score out of 100 or scored too low in one particular category that made up a section of the 100.
He instructed a lawyer, went public and got the payout.
This type of thing might not be as rare as you think. One contributor to the article thinks that up to 75% of all applications in the US are handled by bots – although I find this extremely doubtful for the UK and would put the figure at more like 10 – 25% for us.
Regardless of my thoughts, the fact that Linkedin, Indeed and the Monster job board admit to using AI in at least one part of their candidate screening processes tells you exactly which way the wind is blowing.
And to an extent, you can understand why companies use AI – for reasons of efficiency and cost saving. This is despite the fact that 88% of business leaders know that using AI rejects qualified candidates. Clearly to them that small negative is outweighed by all the benefits.
To be fair, here are just some of them:
- Stops time being wasted on applications from unqualified people (the bane of my life)
- Democratises the recruitment process by giving everyone a fair chance, judging every candidate the same way
- Some human recruiters read only a fraction of the applications, stopping once they get enough suitable applications. AI reads them all
- Applicants from diverse backgrounds who are often overlooked fair better when AI focuses on skills rather than educational achievements
- AI clients have increased their diversity and retention rates (according to Plum)
And now the drawbacks. Take a deep breath, they are severe:
- Some bots have discriminated against women, such as Amazon in 2018
- Longer job descriptions deter female applicants, as women tend to apply only when they fulfill most of the requirements (men have been proven to exaggerate their accomplishments – who knew!)
- Some bots have “learned” to favour people called Thomas and Jared, and those that mention the words Lacrosse and Church
- Any gap in employment regardless of reason can result in an automatic rejection as part of a “knock-out” filter
- Job descriptions with too many criteria and skills reject many qualified applicants who may be missing just a couple of skills from the list
- Bots have rejected qualified candidates because they scored low in one important category, even when they got a near perfect score in all the other important categories – preferring job applicants who receive mediocre scores across the board
Thankfully, there is hope. The UK government is planning a new regulation of algorithmic decision making. In the US, a recent local law requires employers to inform job seekers how their application materials are screened by AI upon request.
Until then, here’s what you need to do to ensure you aren’t being jobBOTomised / BOT down / AIvicted (I intend to trademark one of these terms once I nail it!):
- UPDATE YOUR CV. You will need to spend as much time as it takes to make sure that the keyword for EVERY skill and experience is in there
- Remove any gaps in employment
- Replicate keywords from each job you apply for into the CV that you submit for it
- Add super obvious things like Microsoft Office
- Do not use images or special characters such as ampersands
- Go to church, play lacrosse and change your name to Thomas or Jared*
To discuss in confidence how Marketing Moves can support your organisation’s growth, please contact Melvin by any of the means below.
* I prefer Thomas