Why marketers should be using propensity modelling
- Posted: 19th October 2016
Half of marketing and media executives surveyed in a recent report believe predictive analytics and propensity modelling to be the most helpful technologies for extracting more value from their customer data.
Essentially, predictive analytics uses big data to calculate future probabilities and trends. The business applications are far-reaching and predictive analytics is being used to great effect in sectors from healthcare to farming to finance to drive insight from data. Marketers can use the technology to track how customers engage with their brand – what are their habits, actions and patterns? Using these insights, marketers can use methods such as propensity modelling to establish how likely a customer is to make a purchase, complete an online form or click on an email.
In the past, it would have required extensive analytical skills to build relevant data-sets and pull together intelligent insights. However, modern analytics technologies such as Adobe Analytics and Apteco FastStats can perform a far swifter analysis of customer preference, buying behaviour and profile information – and you don’t need to be a data scientist to work these tools.
That said, marketers should be conscious of the fact that predictive analytics provides a binary analysis and avoid treating the customer as just another number. Propensity modelling can help track what a customer will buy and when they’ll buy it – but identifying the ‘why’ takes a bit more effort.
For example, if someone buys a large amount of plants, they will probably need some fertiliser too – and it’s more likely they’ll want it in spring, not winter. It’s a simple correlation to make, but if your customer data doesn’t present such obvious findings, your marketing team will be none the wiser. In fact, data alone can never give us the magic answer to why a customer behaves a certain way.
Looking for answers
To dig out the meaning behind a customer’s choice or action, you need to blend propensity modelling with everything else you know about the customer. Algorithms and human intelligence need to work together to ensure that the data is clean and that different customers are correctly segmented.
Too often, one business will have predictive models in place for each of its products. Let’s say a bank has separate models for loans, credit cards and mortgages: with these individual models in place, the system may overlook the fact that one customer is using all three products. Product teams then end up fighting over customers instead of considering their overall financial needs.
A complete history of your customers’ every interaction with your brand is essential for effective propensity modelling, enabling you to understand your customers’ behaviour and communicate with them more effectively. A watertight Single Customer View (SCV) will help deliver far more targeted communications.
In addition to keeping your data super clean and regularly health-checked for relevance, you need to make sure that everyone in the company has access to the data. The quality of the customer experience your brand delivers depends on a unified front; marketing, sales, front-desk personnel, finance teams and so on all need to sing from the same hymn sheet – and to do that they need to work from the same information.
One sure-fire way to delight customers is to make sure that your relationship adds immediate value to their lives. Tools like Adobe Analytics and Apteco FastStats can help: they includes a predictive element which helps marketers anticipate the actions a site visitor or customer will take based on their previous behaviour. What’s more, it ties the results into customer profiles based on demographics, providing a far clearer view of segmented targets. This means that marketing efforts can be more than just proactive, they can be precisely personalised to address individual customer needs. When brands show that they really know their customers and provide them with beneficial information and offers, they can increase sales and brand loyalty.
Propensity modelling can help analyse the relationship between an individual and your business, but it’s up to you to act on the information it generates. When the right data is used correctly, these tools can help you pleasantly surprise your customers, increase their loyalty – and see greater profit.