Research

Targeted Online Advertising

Practical guides11th March 2022

This publication forms the first of three short guides focusing on the responsible procurement and use of specific data-driven recruitment tools. It is intended to be read in conjunction with the general AI recruitment guidance published in December 2021, which provides detailed steps to determine whether a tool is fit for purpose.

This guide focuses on additional considerations specific to targeted online advertising.

The guidance has been developed jointly by the Recruitment and Employment Confederation (REC) and the Centre for Data Ethics and Innovation (CDEI). The first in the world of its kind, the CDEI leads the UK Government’s work to enable trustworthy innovation using data and AI. More information about the CDEI can be found on the CDEI web page or by contacting cdei@cdei.gov.uk.

Targeted online advertising - a short guide

Targeted online advertising is used at the sourcing stage of the recruitment process to reach the most suitable pool of candidates, and provide a bespoke service to potential applicants by showing them the job adverts most relevant to their experience and skill set. Highly targeted channels can capture greater diversity in the applicant pool than traditional, standardised approaches. For example, targeting candidates who have professional backgrounds that are not currently reflected in the organisations’ workforce allows recruiters to find candidates beyond their usual talent pools, and gain access to a wide variety of viewpoints and perspectives.

However, the use of targeted online advertising is not without risk. Targeted advertising dictates which candidates are aware of opportunities and, without proper usage, may have discriminatory or otherwise unfair consequences - in particular for protected groups.

1. Evaluation of value-add and effectiveness

Online platforms collect data about their users and can deploy advanced data analytics to make predictions about their users’ behaviour and show adverts to them on that basis.

Targeted online job advertising enables advertisers to target content to specific groups of people online, based on data held about them.
Recommendation algorithms on job boards will promote the jobs assessed to be most relevant to an individual, based on the information collected about them.
Some providers (such as Workable) also offer tools which match a job advert with candidates, meaning that recruiters can receive a list of the best-fit candidates

Once you have established whether online targeted advertising will be beneficial to your campaign, and if so, how it will fit into your wider recruitment process you need to consider the key risks and issues associated with this technology.

2. Risk of discrimination

Algorithmic discrimination Case study Targeting process Positive action Other targeting criteria

Discrimination can arise from the use of an algorithm, and multiple studies have shown that advert delivery on online platforms ‘can be skewed by gender or race due to hidden algorithmic optimisation by the platforms, even when not requested by the advertisers’¹.

Many platforms use an automated bidding process, which is a strategy that automatically sets bids for your adverts based on that advert’s likelihood to result in a click or conversion. This process has been shown to bias against women where adverts are optimised for cost; advertising to women is more expensive² and so “gender-blind” advertising can produce discriminatory effects. Although this may be unintended bias, recruiters need to be aware of the potential outcomes.

Case study

In 2021, LinkedIn introduced a Fairness Toolkit that aims to mitigate against potential bias in advert delivery, for example: tools to detect statistically significant differences in model performance across different subgroups³.

More detail can be found in the LinkedIn Fairness Toolkit⁴.

The targeting process can lead to unfairness and discrimination in a number of ways:

Explicitly targeting advertising using protected characteristics is likely to constitute direct discrimination unless the activity falls within the scope of positive action (see box below).
For example, some advertising platforms enable adverts to be targeted at men or women. Targeting a job advert at men because most previous recruits have been male would be likely to constitute direct discrimination.

Positive action

S.158 of the Equality Act 2010 enables an organisation to take action to compensate for disadvantages that it reasonably believes are faced by people who share a particular protected characteristic. In the context of sourcing potential recruits, this could include specifically targeting job adverts at groups that are under-represented within your workforce.

Targeting to improve diversity is important and should not be discouraged.

Other targeting criteria can act as proxies for protected characteristics. In some circumstances, such targeting could lead to indirect discrimination, a key test is whether the criteria being used can be objectively justified.

For example, targeting an advert based on geographic location will impact the mix of racial groups that see that advert. In many circumstances, the location of a job role will be highly relevant to whether candidates are likely to apply, so this might be justifiable. However, geographically targeting a job where the location was flexible, or for roles where candidates might be reasonably expected to relocate to take up a post, might not be justifiable or might even be counterproductive.

Where discrimination does arise - even if unintentional - both advertisers (the recruiter) and publishers (the platform) can be held legally responsible⁵.

A structured approach to assess the risk of discrimination:

➔ Consider what targeting criteria are relevant and appropriate for the role you are recruiting for.

Outline the potential side effects of using your chosen targeting criteria. (E.g. if you target a specific field of academic study, you may miss out on candidates who have learnt relevant skills in the workplace).

Adjust the targeting criteria to mitigate these side effects if necessary. (e.g. in the example above, consider other contexts where candidates might have been likely to obtain relevant skills).

Make sure you can justify the targeting criteria you have selected, and can explain how it will help deliver the right candidate for the job.

➔ Ensure targeting criteria are not explicitly biased or potential proxies for protected characteristics.

Ask the platform for aggregate statistics on who has viewed your specific targeted advert, to understand the reach.

Ask the platform about how they are monitoring whether outcomes are non-discriminatory (e.g. not disproportionately targeting social care roles at young women) across their job advertising practices.

➔ Consider the reputation and history of the service or platform:

➔ Understand whether there is unfairness:

Collecting information about the targets of your advertisements may indicate whether there is a skew in the audience.

Consider collecting demographic data for the applications you receive to help assess the impact of using different advertising criteria and platforms.⁶

3. Data protection

When placing a targeted advert on an online platform, recruiters are typically not processing personal data. Data protection responsibilities are therefore the obligation of the platform hosting the advert.

It is still advisable to seek assurance that your chosen platform’s practices comply with UK GDPR requirements, and to confirm that you will not have access to personal data through your use of the platform.

4. Transparency

At present, limited information is provided publicly by many platforms: both for the advertiser (limited information on the audience of the advert) and to recipients of the advert (why they are being targeted). This makes challenge or redress difficult, again for both advertisers and recipients.

Think about the transparency indicators used by the platform you are considering, particularly if these can help you understand which groups of people your advertisement is being shown to and so enable you to assess fairness.

5. Communication

Proactive communication and transparency around how a recruitment round is being conducted is important for inclusivity. This is particularly true when there is a data-driven tool involved in the process, as candidates may feel unclear about how decisions are being made.

Although the use of targeted online advertising may have less influence on a recruitment outcome than, say, an AI tool in the sifting or interview stages, communication is still key. Being clear about targeting criteria and how targeting is being used, to ensure a broad array of groups are aware, may be a useful step. It encourages a culture of transparency.

Practically, this could include publishing a document which outlines the justification for the use of targeted online advertising, and a cumulative evaluation of its effectiveness (not specific to a role). It is preferable that this information is available to existing employees and prospective applicants.

Consider communications around the use of targeted advertising under the positive action section of the Equality Act 2010.

References

¹Auditing for Discrimination in Algorithms Delivering Job Ads, 2021

²https://www.london.edu/think/is-online-advertising-exacerbating-the-gender-gap

³Auditing for Discrimination in Algorithms Delivering Job Ads, (2021) Ali, M., Sapiezynski, P., Bogen, M., Korolova, A., Mislove, A., and Rieke, A.Discrimination through optimization: How facebook's ad delivery can lead to biased outcomes. In Proceedings of the ACM Conference on Computer-Supported Cooperative Work and Social Computing (2019).

⁴https://github.com/linkedin/LiFT

⁵ https://www.equalityhumanrights.com/sites/default/files/ehrc_advertising_-_checklist_12_0.pdf

⁶Collecting and analysing demographic data for the purpose of testing fairness is permitted within UK data protection law, where explicit consent has been provided.

Other guides in this series

The Centre for Data Ethics and Innovation (CDEI) reviewed bias in algorithmic decision-making which highlighted and, together with the Recruitment & Employment Confederation (REC), developed this series of guides to help recruiters make informed decisions on their use of data-driven tools. This guidance aims to give recruiters who are looking to buy data-driven tools the mechanisms to effectively evaluate and responsibly deploy them, ensuring that appropriate steps have been taken to mitigate risks and maximise opportunities.

Asynchronous video assessments are interviews where applicants are asked to submit a video recording of themselves answering a set of predetermined questions, with answers subsequently undergoing analysis through automated technology (natural language processing), human oversight, or a combination of the two. This differs from a traditional interview format because there is no human interviewer in attendance.

Automated sifting tools, also known as candidate screening software, are a set of data-driven technologies that assess the applications received for a particular role. They often use natural language processing to evaluate CVs and personal statements and assign them a score. This score is then typically used by the recruiter or firm as part of the prioritisation process for invitation to interview. In most cases, the tool will score or rank the candidates using keyword search results based on criteria defined by the employer; for example, matching the keywords in the job overview or candidate specification.

Share this article

More research this way

Discover more

Research

Targeted Online Advertising

Targeted online advertising - a short guide

Other guides in this series

Data-driven tools in recruitment guidance

Asynchronous video interview assessment tool

Automated Sifting Tools

More research this way

Jobs Outlook - November 2024

UK recruitment industry status report 2023/24

Labour Market Tracker

Research bitesize