The Next Beauty Breakthrough May Already Exist Somewhere Else

How innovation travels across categories and what AI could change about beauty’s ability to recognise it.

By Nassim Belhaq · Paris Macro Beauty Edit

The most revealing detail in a beauty launch is sometimes the product arriving beside it.

In early October, Vogue Business reported on remarkably similar eye-patch dispensers announced by Fenty, Glow Recipe and Huda Beauty within days of one another. The resemblance prompted a familiar debate about originality. It also invited a question about how companies arrive at their choices.

Several teams can follow a perfectly competent development process and reach almost the same destination. If their starting points, commercial assumptions and available solutions are similar, convergence should hardly surprise us.

This creates an awkward problem for an industry now looking to AI to accelerate innovation. Speed has considerable value. It also makes the direction of travel more consequential.

There is a way of thinking about innovation that helps explain the distinction.

In a paper published online in August, Rajan Varadarajan identifies 55 fundamental product innovation pathways: principles drawn from earlier innovations that can be applied to other products. Combining separate functions, turning a service into a product, or removing a supposedly essential feature are among them.

The proposition is useful because it separates a product’s visible novelty from the underlying operation that made it possible. An idea can be original in its application while drawing on a recognisable mechanism.

For beauty, that opens a more interesting examination of its own history.

When Clinique launched in 1968, dermatologist Norman Orentreich’s approach informed a routine organised around cleansing, exfoliation and moisturising. Seen through this lens, part of the innovation was the conversion of professional guidance into an everyday sequence. The instructions and relationship between the products gave the consumer a method to follow.

Amorepacific’s cushion offers another example. The company describes the stamp pad as an inspiration for holding a liquid formula in a sponge and releasing it through an applicator. IOPE Air Cushion launched in 2008. A commonplace mechanism supplied an analogy; formulation and packaging development made it usable in makeup.

Dyson’s Airwrap, introduced in 2018, applied the Coanda effect to attract and wrap hair through controlled airflow. Here, a known physical phenomenon became the basis of a different approach to a familiar styling task. Recognising the possibility was one achievement. Engineering a device around it was another.

Fenty Beauty’s original foundation offer illustrates a further route. Its 40 launch shades made inclusion tangible in the assortment itself. My reading is that the strategic choice concerned who the product had to serve from the outset. That commitment had consequences for shade development, inventory and the experience at the counter.

These cases suggest several ways to search: make expertise accessible, borrow a mechanism from another context, change how a task is performed, reconsider whom a category is designed for.

They also show why recognising a pattern leaves so much work still to do. Knowing that a stamp can hold liquid does not produce a cushion foundation. Deciding to serve more skin tones does not resolve the formulation or commercial demands of doing it well. The insight becomes valuable through a series of connected choices.

This is the part of innovation that the story of the brilliant idea tends to compress.

A founder’s conviction, a chemist’s observation or a designer’s instinct may begin the process. What follows involves people with different definitions of a good product. Marketing has to make the proposition understandable. Development has to make it perform. Operations has to make it consistently. Finance has to establish whether the business can support it.

Each decision can strengthen the original idea. Each can also gradually remove what made it interesting.

A system organised mainly to minimise uncertainty will tend to favour what is already familiar. That may be commercially sensible for a particular launch. It becomes limiting when the same logic governs the entire pipeline.

External manufacturers are essential to this discussion. They can provide specialised research, technologies and manufacturing capabilities a brand could never build alone. The quality of the collaboration depends partly on the problem the brand brings to them. A precise brief gives a partner something to solve; a request for whatever is trending narrows the relationship to what is already available.

The same distinction applies to AI.

An AI system can help teams work across a wider set of precedents and propose connections they might otherwise overlook. It can compare how a problem has been addressed in different categories, organise competing explanations and identify where the evidence is thin. Used this way, it extends the work of investigation.

Its usefulness depends on the questions put to it. Asking for the next successful serum already assumes the opportunity belongs in a serum. Starting with an unresolved consumer difficulty leaves room to examine the formula, the delivery, the instructions or the routine around it.

Consider, as a hypothetical example, a scalp treatment people find awkward to apply. A team could spend months strengthening the active story while the obstacle remains placement and residue. A broader examination might point towards the applicator, or reveal that the formula itself needs changing. The useful output is a reason to investigate one route before committing to it.

That demands more than a convincing explanation.

A recent study in Scientific Reports provides a useful caution. Researchers developing an AI knowledge graph for ageing and longevity reviewed 300 relations: 97.3% of the displayed evidence was attributable to its cited record, but only 65.3% of the claims about those relations were fully supported by that evidence.

The distinction between finding a source and interpreting it correctly matters. A well-referenced answer can still take a connection further than the evidence allows. For beauty development, the corresponding discipline is to distinguish a promising analogy from a demonstrated mechanism, and an ingredient finding from evidence about a finished formula.

An innovation system also needs a way to learn when its assumptions meet actual use.

In September, Prose’s scientists described an approach to testing formulation changes through the company’s direct-to-consumer model. A randomised, consumer-blinded comparison of personalised conditioner variants generated 82,447 survey responses. This is the company’s own technical account, but it offers a concrete illustration of how development can continue after products reach consumers.

What interests me is the connection between formulation decisions and feedback from use. It gives the next development decision a body of evidence the previous one did not have. That capacity to learn can accumulate, provided the organisation retains the findings and understands their limits.

For AI, this suggests a demanding role: helping knowledge move between disciplines while preserving the reasoning behind the choices. A scientific possibility has to meet a consumer need. A consumer need has to make sense for the brand. A distinctive proposition has to survive the practical conditions of making and selling it.

This is the conviction behind NAWA NOVA, which I founded to connect market intelligence, brand context, product development and business viability. The ambition is to make those relationships easier to examine while the decisions are still open.

There will always be ideas whose value is difficult to demonstrate in advance. Good systems need to leave room for them. Past successes offer routes to explore; they cannot define the outer boundary of what a brand is allowed to imagine.

The next advantage in beauty may lie in the ability to hold both positions: to learn rigorously from what has worked, and to recognise when a familiar pattern is becoming a constraint.

Creativity needs that freedom. It also needs an organisation capable of carrying a promising idea through development without losing the reason it mattered.

—

Nassim Belhaq is the founder of Paris Macro Beauty and NAWA NOVA.