Skip to main content

Rise of the customer decision engine

By 18/10/2022October 19th, 2022Blog Posts, Thought Leadership
The sun and moon appearing over earth at night.

Sometimes software products seemingly come from nowhere, then suddenly they’re everywhere. They catch the zeitgeist. Some are here today, gone tomorrow. Others remain and become part of the fabric of our business world. We wonder what we ever did without them and why no one thought to build them years ago.

Customer decision engines are the latter. Let me explain why.

What exactly is a customer decision engine?

A customer decision engine is a piece of software that analyses different sources of data to create insights into customer behaviour and needs. Based on those insights, the decision engine recommends actions (or decisions) to be taken in relation to each individual customer, or segment of customers.

Customers are the centre of our universe.

Name me a retailer who doesn’t claim to put customers at the centre of their organisation. Who would argue that meeting and exceeding customer needs isn’t their company’s north star? In 2022 being customer-centric isn’t up for discussion. It’s table stakes. 

The correlation between delivering first class customer experiences and business success is clear. We know this from being consumers ourselves. Retailers who fulfil our needs, make us feel special, and offer excellent value for money are the ones we purchase from time and again, and recommend to our friends and families. 

What makes a great customer experience?

By customer experience we’re just talking marketing, right? Wrong. Think of it from your own perspective as a customer. What contributes to your experience when browsing or buying from a retailer? For most readers, these things will feature high on the list. 

  • The product range on offer. 
  • The product range specifically curated to me. 
  • Products offered or bundled together due to their complimentary nature. 
  • The substitute products offered where my first choice is out of stock. 
  • The price of each product. 
  • Product promotions and offers. 
  • How the retailer markets to me. 
  • The physical store environment. How I feel when I walk in, navigate and check-out. 
  • The availability of products on a physical or virtual shelf. 

Now look at this from a retailer’s perspective. None of the above happen by accident (or they certainly shouldn’t). Each element of the customer’s experience starts with the retailer making a deliberate commercial decision. The result of that decision could be the product range they sell, the price they sell at or the advertising they display to customers as they browse their store or website. Each decision therefore carries significant weight. 

The importance of commercial decision making.

Making great commercial decisions is a necessity for business success. As far back as 2013 Bain reported that “our 10-year research program involving more than 1,000 companies shows a clear correlation (at a minimum 95% confidence level) between decision effectiveness and business performance.”

Despite this finding, few companies have been able to implement effective decision making, according to McKinsey – “only 20% of executives believe their organisations excel at decision making.”

The need for commercial decision-making excellence has come into much sharper focus recently due to the convergence of three significant factors, which impact each business the world over.

Economic pressures, cost price increases and the cost of living.

Global health pandemic, war and supply chain disruption.

Rapidly rising customer expectations.

Retailers who fail to make the best possible commercial decisions face a desperate fight for survival. This is abundantly clear from the devastation of traditional high street brands, many of which have become obsolete.

How to make customer-led decisions?

Fundamental to making customer-led commercial decisions is building a deep understanding of customer behaviour and needs. That understanding should then inform the decisions that impact a customer’s experience.

The most effective way to do this is through the lens of the products or content your customers buy and browse. It’s an approach championed by the likes of Amazon, Netflix, and Spotify, to name a few. The approach we’ve taken with our decision intelligence platform can broadly be summarised as follows.

Understand customer behaviour and need states based on how they shop, browse and interact with your products.

Create detailed product features at the click of a button, utilising all your data, including how your customers describe and interact with your products.

Uncover the hidden relationships between customers and products, customers and other customers, and products and other products.

Use the insight created to make customer-led commercial decisions which drive revenue and profit for your business.

The steps outlined above typically involve analysing customer transactions, web browsing behaviour and product data. Through this analysis you can uncover what your customers’ shopping habits tell you about them. For example, do they have a preference towards ethically made clothing? Do they always buy greetings cards with presents? Do they stock up on low priced essential groceries each week, but also splash out on luxury meal kits once a month? 

We use a combination of affinity analytics (to understand customer and product relationships), attributes (to describe products and customers in helpful ways) and customer need states (how shoppers make buying decisions) as the data science and AI techniques which uncover these valuable insights. Armed with these insights, you can move towards making breakthrough commercial decisions.

Enter the customer decision engine.

I have a confession to make. At the start of this article, I suggested customer decision engines were new. In fact, they’ve been around for some time. It’s more they’ve been revitalised for the modern business era.

Until recently, customer decision engines were predominantly used by companies offering their customers credit. For example, banks or vehicle finance providers. The decision engine would apply a scorecard to determine whether the providers should offer the customer credit, and if so, on what terms. That scorecard would be built based on the customer’s credit history, how long they’d been registered to vote at their current address, family associations and various other factors.

Customer decision engines are also historically associated with large software vendors who promise their products will make every single decision the business needs, but months or years later, the product still isn’t implemented due to the scale of the task. This breeds frustration and enormous wastage of time and money.

Now here’s the key difference. These legacy customer decision engines were designed to be automated. They rigidly applied a scorecard or set of rules with no room for human digression. Much to the annoyance of anyone who has had their application for credit declined or were subjected to a ‘computer says no’ customer experience.

Modern customer decision engines deliver decision intelligence. They empower humans to make the best possible customer led decisions they can. They support, not replace, humans.

They enable a set of interlinked decisions, which directly impact the customer’s experience, but don’t claim to solve every problem the business has. They provide a pragmatic solution that delivers business results rapidly.

How a customer decision engine works.

Modern customer decision engines typically ingest data, perform analytics, serve easy to consume insight and recommend a decision, which can then be sent to a downstream system. Each customer decision engine will vary to some extent. Below is a top-level illustration of HyperFinity’s architecture, with our decision engine at the centre. Please note that we work predominantly with retailers, hence the industry specific examples.

Every customer decision engine supports a different set of decisions. For our part, HyperFinity’s decision intelligence platform enables a set of interconnected, customer-led decisions.


How do we create relevant experiences through deep customer understanding?


How do we optimise our assortment across all our sales and marketing channels?

Pricing and promotion.

How do I use deep product relationships to power my pricing and promotional campaigns?

Marketing and media.

How do we create next best action for truly personalised marketing and media campaigns?


How do we optimise website experiences to increase customer conversion and loyalty?

Supply chain.

How do we become customer data-led with demand and availability forecasting?

A key component of our decision intelligence platform is the ability to create customer segments at macro, micro and individual level. The power and flexibility that brings to a customer decision engine is essential as many companies don’t have the tools needed to deliver personalised experiences to customers.

Customer decision engines tend to fit well into a composable technology stack. By that I mean selecting best of breed software applications and connecting them via APIs. The alternative would be to choose a single, monolithic vendor whose product covers all your needs. In our view this approach means you settle for some substandard features where the vendor isn’t a specialist. Plus, changing vendors is devilishly hard. In a composable architecture, swapping out individual vendors is far more straightforward.

Why invest in a customer decision engine?

Ultimately, spending money on any piece of software must generate a return. Let’s explore some of the key benefits of customer decision engines.

Increase revenue and profit.

The whole premise of a customer decision engine is to make commercial decisions which deliver exceptional customer experiences. The economic benefits of doing this are crystal clear. HBR, for example, state that “customers who had the best past experiences spend 140% more compared to those who had the poorest past experience.”

Customers who feel valued, recognised and their wants and needs catered for, will invariably spend more, become more loyal and recommend the retailer to others. Retailers, in turn, will generate more revenue and higher profits. A true win-win situation.

A customer decision engine can also play a major part in setting the optimal price of products. Traditional pricing methods tend to rely on historic data and gut feel. Using data science and AI to truly understand what motivates customers to buy, and uncovering the relationships between competing and complementary products, enables data informed pricing decisions. For example, where demand will remain even if prices are increased, and where demand will flow to competitors if discounts aren’t applied.

Armed with these insights, profit margins can be protected through intelligent pricing decisions. Given the stark increase in the cost of goods retailers are purchasing, and customer spending reducing due to the cost-of-living crisis, pricing optimisation has risen to the top of many executives’ priority list.

Reduce costs.

Let’s not forget the primary purpose of most retailers – to generate financial profit. One way to achieve this is by selling more products. The other is controlling costs.

A customer decision engine helps achieve this in several ways. For example, concentrating marketing and media spend on those customers most receptive to the products being promoted, rather than broad brush marketing with poor return on investment.

A major cost contributor for retailers is the stock they hold. Being ‘over ranged’ is common for retail businesses – but needn’t be. A customer decision engine provides insight into how customers buy and browse products, which products they have a strong affinity to, and the set of products being considered before customers make a purchase.

The insight created can help decide on the full assortment a retailer offers, how that assortment varies for each physical store location, website, or distribution channel, and how to curate the assortment to every single customer.

The results are likely to suggest a reduction in the size of product range the retailer needs to offer, changes to the amount of stock held and its location, and how it should be merchandised physically and digitally to convert browsers into buyers. All of which lead to lower costs and higher profits.

Empower employees.

If there’s one thing people hate, it’s not being able to do the job they’re employed to do to the best of their ability. Yet this is precisely what happens when decision makers don’t have access to the information they need to make the decisions they’re paid to make.

Think about your own role. No doubt it involves making significant decisions much of the time. Are you comfortable making those decisions without first understanding the factors at play and the likely consequences if you make choice A, B or C? No, neither am I.

A customer decision engine empowers employees with the insight they need to make the best possible decision, every time. In the case of HyperFinity, we’ve focused on putting data science and AI in the hands of any business user, regardless of whether they can code. We achieved this by building a highly accessible, visual, point and click user interface.

By doing this we’re mitigating the chronic shortage of data science and AI talent globally. Now anyone can access the insight they need in minutes, not have to wait weeks or months for scarce technical resource to action in your request.

How to find out more.

We firmly believe that modern customer decision engines will be rapidly adopted across industries, with retail being an early adopter, such is the strength and value of the use cases.

HyperFinity, with our decision intelligence platform, is part of a cohort of modern, agile, pragmatic technology and consulting vendors pioneering this movement.

If you’d like to discuss your companies’ use cases or share thoughts on the benefits of customer decision engines, please get in touch with me at or connect with me on LinkedIn at

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.