STRATEGY MEETS ARTIFICIAL INTELLIGENCE 

Marketing strategy is undergoing a revolution. Gone are the days when it was acceptable to build an entire campaign on musings of an advertisement team. Marketers today understand that they need to be much more strategic. They have to deal with intense competition. Under the growth pressure, marketers want to find out what exactly contributes to their market share. They are more accountable than ever to prove the value of marketing programs.

 

When it comes to data-driven marketing, the classical approach is about leveraging insight from data analysis to make marketing more efficient and effective. It involves analyzing, measuring, and optimizing marketing efforts so that adjustments to campaigns can be made more quickly.

Yet, the classical approach assumes that the marketing strategy in use is the one that works.

Not a single analytics tool that focuses on marketing optimization builds the foundations of a winning strategy. Divera is different. For the first time in marketing, strategy meets artificial intelligence.  Let's continue with a reminder :

 

STRATEGY IS NOT

A CUSTOMER SURVEY

"customers do not know what they want unless they experience it."

Customer experience is the basis which generates the purchase decision.  It is the heart of the any marketing strategy. However, in real life, multiple experiences can not be tested on customers to observe which works the best. It is neither practical nor feasible. Moreover, the observations of the impact on the purchase decision are elusive.

A customer may have an experience or a belief. Experiences turn into beliefs as well.

Beliefs drive the behaviour. However, asking customers how a new belief will shape their decisions is futile…They would not able to tell. They have to experience it.

OUR SOLUTION:

 VIRTUAL SELF

A virtual customer

We know that future cannot be told.

To observe future behavior,

DIVERA develops virtual consumers from REAL CONSUMERS in a computer-generated simulation

Virtual Self is an approach allowing for the first time in marketing - based on a simulation -, to observe consumer behavior directly as if we were intervening in real life with a new communication campaign or a new product and deduct the gains in terms of market share

Virtual customers are autonomous. They can take decisions.  Artificial Intelligence builds them from the data obtained from real people. 

NOT A DATA POINT

Three principles embodied "Virtual self" idea :

1. A customer is a person not a purchase behavior data for statistics. Data has to tell us “why” a person behaves in a certain way rather than what her / his behavior is.

2. We should preserve the uniqueness of the individual within a technological framework.

We should not focus on average. Outliers are not anomalies. On the contrary, each person has a unique trait which makes him or her an outlier on a different dimension. 

3. We shall not segment. Man-made segments are fake. Each person is a segment of one.

These principles became the basis of Divera’s AI technology.

Each virtual consumer is a discovery. Because each virtual consumer represents a real person, and each real person has a unique mind to discover. 

Virtual consumers are not statistics, they are not data points.

 

They are virtual selves.

BUILDING A MIND

FROM DATA

Mind built from data

To create a virtual self, AI system needs three types of data.

 

Purchase behavior : Frequently bought and / or used product / service

Brand perceptions : A person’s beliefs about the brands on product / service basis.

And the most important type of data, the key to the unconscious, biological data.

For each person that we will virtualize, we have to capture the residue of the experience in the form of belief, convert it into a numerical form and feed it to the system.  To extract belief data from the mind of the consumer, we use a special algorithm which quantifies the subconscious reactions of the consumer to certain belief statements as the probability of change of mind.  Subconscious reactions are captured in pupil dilation, heartbeat rate change or similar biological metrics. We convert all biological data into probability vectors with direction and magnitude. The output is a unique mind-model explaining why a certain individual is loyal to a certain brand and the product.

POWERED BY

NEURAL NETWORKS

neural-network.png

Use of artificial neural networks in an unconventional way helped us to solve the puzzle for each person. Each virtual consumer is powered by an unique neural network which functions as a decision engine.

To learn more about Neural Networks click here :

MARKET SIMULATION:

WHERE  VIRTUAL CONSUMERS LIVE

A virtual consumer is situated in a market simulation. Market simulation is the virtual environment where simulated consumers exist and make buy decisions. In the simulation environment, we can upload new beliefs to a virtual consumer. Change in beliefs may lead to a change in behaviour. To observe the outcome, we run the simulation with new beliefs.

WHAT IF?

To test strategy options, AI replaces existing beliefs of the simulated consumer with positive new beliefs about our choice of brand. Examples...

Speed

What if the consumer had an experience of Iphone as the fastest smartphone compared to the alternatives...

Would it be more loyal ?

Camera

What if the  competitor's consumer had an experience of Iphone as the best low-light camera...

Would it switch ?

Design

What if the  competitor's consumer had an experience of Iphone as the thinnest smartphone...

Would it switch ?

Interface

What if the  competitor's consumer had an experience of Iphone as the least cluttered interface..

Would it switch ?

COUNTING DECISIONS:

BOTTOM UP

STRATEGY

When we repeat this process for all the virtual consumers in the simulation, we re-create the market but with a different future.

Through iterations, we can build a product that will gain the maximum market share from the competitor of our choice, or discover the most impactful marketing message.

VIRTUAL CONSUMERS PREVAIL

WHERE TRADITIONAL FAILS

In highly developed, competitive markets, the source of growth is the competitors. Competition could be direct or indirect. Idea is to achieve growth through switching customers. However, to discover which marketing strategy makes the competitors' customers switch is impossible using traditional methods like focus groups, surveys or any web metrics. People cannot tell their future behaviour.

Moreover, one winning move may not be enough. Even it could be a very short-sighted strategy. One has to know how to counter the competitor as well. No one lets the rug being pulled under them. Everybody fights back. Using virtual consumers, we can find out what counters the counter-attack of the competitor and be ready for all possible competitor responses to your growth strategy. Again, testing counter-moves is not possible with traditional methods.

We have to keep in mind that competitors want your customers too. They will have their switch strategies to lure your core customers. Right from the start, in our market simulation DIVERA develops protective strategies against possible competitor attempts. Optimizing a marketing strategy to find the balance between switch and loyalty is not achievable using well-known methods like focus groups.

The last but not the least, you have to win with the available resources at hand. if the company needs to invest more than what they have to implement the strategy, we may gain market share at the expense of profitability. DIVERA always propose an optimal strategy to balance profitability and market share gain. Such fine-tuning of the winning strategy is not possible through methods of the past like surveys or polls.