ETG General Paper
2024 A-Level GP · Paper 1 · Question 6

Targeted online advertising

What this question asks

This question asks whether the increasingly sophisticated targeting used by online advertising does more harm than good, for consumers and for society.

Question type: To what extent

An ETG General Paper original study guide to the 2024 A-Level GP Paper 1 essay on science & technology. Not affiliated with, or endorsed by, UCLES, Cambridge Assessment or SEAB. A study aid, not an official answer.

Read the question first
Define these terms
  • sophisticated methods: data-driven profiling, behavioural tracking and micro-targeting
  • target consumers: reach individuals with tailored messages based on their data
  • more harm than good: net balance, and harm or good measured against whom, the consumer, the firm, or society
The hidden assumption

The question assumes targeting is one thing with one balance sheet, when relevance, manipulation, privacy loss and market efficiency are different effects that fall on different parties.

The calibration axis

conditions: targeting brings more good where it is consented, transparent and commercial, and more harm where it is covert, manipulative or political, so the balance turns on the conditions of use.

Two ways to argue it

How to approach it. Weigh whether sophisticated ad targeting brings more harm than good, defining harm and good and for whom, and resisting a flat condemnation or defence.

Option A · Conditions: relevance vs manipulation

Sophisticated targeting brings more good than harm where it is consented and commercial, since relevance benefits both consumer and firm; but it brings more harm than good where it crosses from relevance into covert manipulation, exploiting psychological data and the young, so the net balance depends on which side of that line the targeting sits.

The argument, point by point
  • Targeting brings genuine good when it makes advertising relevant rather than wasteful.
    Why Matching offers to actual interests lowers search costs for consumers and wasted spend for firms, so a small business can reach the few people who want its product instead of paying to shout at everyone.
    Example Platform ad tools let niche and small enterprises reach precise audiences they could never afford to via mass media, a documented driver of small-business growth online (global pattern, as_of 2026-06).
    Then evaluate But relevance shades into surveillance: the same precision that helps the buyer requires knowing more about them than they may have agreed to share.
  • The harm becomes serious when targeting moves from predicting wants to manufacturing them through psychological profiling.
    Why Profiling emotional states and vulnerabilities lets advertisers hit people when their judgement is weakest, so persuasion crosses into manipulation that bypasses rather than informs choice.
    Example The Cambridge Analytica scandal exposed how the data of up to 87 million Facebook users was harvested to build psychographic profiles for micro-targeted political messaging (widely documented, as_of 2026-06).
    Then evaluate The hinge: targeting that informs a choice is good, targeting that exploits a weakness to override it is the core harm.
  • The harm is gravest for the young, whose data is collected and whose judgement is least formed.
    Why Children and teenagers are profiled like adults but lack the defences to recognise manipulation, so targeting reaches them through the developmental gaps that regulation exists to protect.
    Example Singapore's PDPC issued Advisory Guidelines on Children's Personal Data in March 2024, requiring clear consent and data minimisation precisely because profiling of minors is a recognised harm (PDPC, as_of 2026-06).
    Then evaluate Yet the existence of the safeguard shows the harm is contingent, not inevitable: regulated targeting of adults differs from unregulated targeting of children.
  • Targeting also corrodes the information environment when the same tools serve falsehood and division, not just products.
    Why Micro-targeting lets different messages reach different people invisibly, so accountability collapses and divisive or false content can be aimed where it will do most damage with least scrutiny.
    Example The Capitol riot of 2021 followed micro-targeted falsehoods about a stolen election spread through the same ad and amplification machinery built to sell goods (widely documented, as_of 2026-06).
    Then evaluate This is the societal harm beyond the consumer: the sophistication sold as relevance becomes a vector for manipulation of the public, not just the shopper.
Strongest counter & rebuttal

Search, maps, email and social platforms are free because targeted advertising pays for them, and billions opt in daily, a revealed preference that targeting delivers value people want. But consent given without understanding the data trade, and to terms few read, is thin consent, and the bargain looks worse once the price includes psychological profiling, manipulation of the vulnerable and a degraded information commons, so the popularity of the free service does not settle the net balance.

Measured conclusion

Sophisticated targeting brings more good than harm where it stays consented, transparent and commercial, and the free services prove its value; it brings more harm than good where it becomes covert manipulation, preys on the young, or carries falsehood into politics, so the verdict is conditional, and the regulatory line between relevance and manipulation is where the balance is decided.

What makes this Band 1: Reaches the top band by drawing the relevance-versus-manipulation line and judging cases against it, by distinguishing harm to the consumer from harm to society, and by treating consent critically rather than as a settled defence.
Option B · Domain: market efficiency vs democratic harm

The balance differs sharply by domain: in the commercial domain sophisticated targeting brings more good than harm, raising market efficiency and funding free services, but in the civic domain, where the same tools micro-target political and ideological messages, it brings more harm than good, so the honest answer is that the technology is net positive for markets and net negative for democracy.

The argument, point by point
  • In the commercial domain, targeting is largely net positive because it makes a market work better.
    Why Better matching of buyers and sellers reduces waste on both sides and lowers entry costs for small firms, so the economy gets more value from each advertising dollar.
    Example Targeted platform advertising lets small and niche businesses reach precise customers affordably, widening competition beyond firms that can buy mass media (global pattern, as_of 2026-06).
    Then evaluate But even here the good is bounded by data extraction, so 'net positive for markets' assumes the privacy cost is contained by regulation.
  • The commercial harms that do arise are real but largely manageable through existing consumer-protection tools.
    Why Privacy loss, dark patterns and pressure selling are addressable by consent rules, disclosure and data-minimisation duties, so the commercial harms have known remedies even where enforcement lags.
    Example Singapore's PDPA, the EU AI Act in force from 2024, and the PDPC's 2024 children's-data guidelines give regulators levers over commercial profiling (PDPC; EU AI Act, as_of 2026-06).
    Then evaluate The complication: these tools assume a transaction the consumer can opt out of, which is exactly what breaks down in the civic domain.
  • In the civic domain the same targeting turns net negative, because invisible, tailored political messaging undermines shared public reasoning.
    Why When voters see different, individually optimised messages no one else can scrutinise, accountability and a common factual basis erode, so the harm falls on democracy itself, not a consumer who can simply not buy.
    Example Cambridge Analytica's harvesting of up to 87 million Facebook profiles to micro-target political ads in 2016 showed the machinery turned on elections rather than products (widely documented, as_of 2026-06).
    Then evaluate Here there is no opt-out and no refund: the harmed party is the public sphere, which consumer protection was never built to defend.
  • The domains share infrastructure, which is why commercial efficiency and civic harm cannot be separated in practice.
    Why The same data, platforms and profiling models serve both shopping and politics, so improving targeting for commerce simultaneously sharpens the tools available for manipulation of the public.
    Example The amplification and ad systems that funded free platforms also carried the micro-targeted falsehoods preceding the 2021 Capitol riot (widely documented, as_of 2026-06).
    Then evaluate This is why a flat 'more harm' or 'more good' fails: the verdict flips with the domain, yet the domains run on one engine.
Strongest counter & rebuttal

Targeting that drives gambling, junk food or buy-now-pay-later debt to the vulnerable does civic-scale damage while wearing a commercial face, so the neat split between a benign market and a harmed democracy is too clean. But this strengthens rather than refutes the analysis: it shows the real fault line is not commercial versus civic but consented and contestable versus covert and unaccountable, and the political domain is simply where covert, unaccountable targeting is most concentrated and least correctable.

Measured conclusion

Sophisticated targeting is not one verdict but two: a net good in the commercial domain, where it makes markets efficient and funds the services people choose, and a net harm in the civic domain, where invisible micro-targeting corrodes the shared reasoning democracy needs, and because both run on one engine, the task is to keep the commercial gains while caging the civic harm.

What makes this Band 1: Earns the top band by splitting the verdict across commercial and civic domains, by showing the shared infrastructure that makes a single answer impossible, and by using the concession to refine the line from commercial-versus-civic to consented-versus-covert.
How the two approaches differ

Option A runs on conditions, drawing one line between relevance and manipulation and judging cases against it. Option B runs on domain, giving two opposite verdicts for the commercial and civic uses of the same technology. Both are defensible: A is the clean conditions-based 'to what extent' answer; B reaches a split verdict that reads as more analytical because it explains why a single yes or no is impossible.

Common pitfalls
FAQ
What is the strongest example for the targeted advertising essay?
Cambridge Analytica. The harvesting of up to 87 million Facebook profiles to build psychographic models for micro-targeted political messaging in 2016 shows the harm at its sharpest: covert, unaccountable and aimed at democracy rather than the shopper. It anchors the manipulation and civic-harm arguments at once.
How do I keep this from being a generic social media essay?
Stay on the mechanism the question names, data-driven profiling and micro-targeting, not screens or addiction. Judge specific effects: relevance and small-business reach as goods, manipulation and harm to minors as harms, and the PDPC's 2024 children's-data guidelines as evidence the harm is real but regulable.
Should I conclude that targeted ads do more harm than good?
Calibrate rather than declare. The defensible verdict is conditional: more good where targeting is consented, transparent and commercial, more harm where it is covert, manipulative or political. The split between commercial benefit and civic harm, running on the same data infrastructure, is the strongest closing line.
ETG General Paper

Want your essays marked like this?

These outlines are free to read. Book a trial lesson to have your own GP essays marked against the band descriptors, the way ETG teaches General Paper from the exam backwards.

Book a Trial Class   WhatsApp the team