/ Customers and Buyers

What if quality of data matters more than the quantity?

Big data, artificial intelligence and machine learning open a new world of possibilities that have grabbed attention and captured imagination. The ability to reach huge markets delivering personalised messages is indeed beguiling. Targeting is based on patterns and represents an estimate of what is most likely to resonate. When volume counts, the ability to deliver in quantity is critical.

Equally there are situations where 'probably' doesn't really cut it and when the quality of the data becomes the defining factor.

As an example consider the Google traffic application. The earlier versions parsed significant quantities of data to calculate the average time to travel along certain routes at certain times. The quantity of data was huge. The later versions use a smaller quantity of better data to provide a picture of traffic real time. Less data but a far more valuable result. Of course, in the case of Google 'less' is still a toe curling volume.


In the average business small numbers matter too. Sales are made as single transactions, one at a time. Plus the outcome is binary: you make the sale or you don't. At this level quality can trump quantity.

So what does this mean for those wanting to take commercial advantage of advances in technology?

Building pipeline is a numbers game: marketeers are well advised to attract interest using tools based on quantity.

Sadly for sales, the majority of big data, AI and ML technology is poor: the parts that are good are not interesting, and parts that are interesting are not good. Winning sales is binary - win or lose: salespeople will close more deals using data and tools focused on quality.

The reason most technology for selling is weak is because the problem is being fit to the solution, rather than fitting the solution to the problem. Let me explain. At the point of purchase the customer wants to be able to rely upon what they are told. That means data needs to be not only correct, but relevant to their needs. Technologies that use data in the context of customers needs are far more likely to grow sales.


The parallel of the self driving car

To make a self driving car there are in fact a limited number or operations to control - predominantly changes in velocity (acceleration, deceleration and direction) - based on a small number of environmental factors - route, road surface, other vehicles, cyclists, traffic volumes, pedestrians, weather etc.

It doesn't matter what the other cars are, who the pedestrians are or their individual purpose or plans, what day of the week it is, or what the passengers are wearing (or what lycra the cyclists are wearing!) It doesn't matter why they need to travel.

So to make a self driving car the approach is not to teach a computer to drive, exhibit judgment and take decisions. The method is to mimic what a person driving would have done in similar circumstances. This is achieved by recording millions of hours of driving and refining the matching techniques the compare collected data to the current situation.

Thus far we have the parallel of big data and tools to power push marketing. To make a good self driving car the next step is to ask some personal questions to establish need:

  1. Where do you want to go?
  2. What time do you need to be there?
  3. Do you have a preferred route?
  4. Would you like to stop along the way?

Defining a specific need enables the quality of the data to be enriched because only relevant results are used.

People will only buy a self driving car that is going to take them where they want to go...

So in conclusion...

Quantity of data is an enabler. It will support broadcast marketing enhancing the likelihood of getting across a message that will resonate.

Quality of data enhances individual outcomes which is highly relevant to making sales. Making data specific to a customer's needs significantly increases its quality and value.

Organisations looking to increase sales through the use of AI and Machine Learning should be concerned with the mechanisms of collecting the actual customer needs and applying data to address them.