With traditional data collection seemingly not satisfying marketers’ thirst for customer insights, Maxine-Laurie Marshall explores what predictive analytics can offer marketing
Predictive analytics is one of those new trends that ‘experts’ love to roll their eyes at and tell you has been around for decades, making you feel a little silly because you’ve only just noticed people talking about it.
Pessimism and scepticism come more naturally to me than their antonyms, so I too met the words ‘predictive analytics’ with an eye roll. Using data to predict future actions seems like something marketers have been doing for a little while. But with a bit of digging I can confirm predictive analytics goes deeper than my simple explanation and is, in fact, something ‘new’ for B2B marketers.
Ian Bobbett, director and co-founder of Model Citizens, explains the subtle difference predictive analytics can offer in comparison to what you’re used to: “A business user can guess which customers might like a product; insight and analysis can improve that understanding; and predictive analysis can better refine and prioritise those customers.”
Glen Westlake, CEO of Kairos Analytics, offers a comprehensive explanation: “Predictive analytics works by considering historical information to try and predict future outcomes, using a variety of statistical techniques. By looking back at historical outcomes and the facts that lead up to that outcome, predictive models can be built that can then be applied to more recent information where the outcome is unknown. These models then provide either a score of the probability of the outcome occurring or an absolute indicator that the outcome is likely to occur.”
As opposed to the informed guesswork an insight into your customers affords you, predictive analytics offers mathematical models that are used to work out the likely behaviour of your customers.
The possibilities here are massive and have the potential to make the customer centricity dream a reality. It’s understandable why the industry is buzzing around the words ‘predictive analytics’.
Is marketing ready?
For those deeming it old news, perhaps they were one of the fortunate few who had the resources to invest in it. John Hurley, director, product marketing at Radius, says: “Predictive has traditionally been reserved for an exclusive few that could spend money and time working with expensive enterprises or consulting firms that would custom build models. Now predictive is becoming democratised.”
And while vendors and agencies offering predictive analytics services begin to pop up, not every marketing department is ready for it. Hurley continues: “If you do not have customer data and historical performance data housed in a CRM or marketing automation (MA) system, then you may be too early for predictive analytics.”
Marc Keating, director of digital and innovation of Stein IAS – who earlier this year announced a partnership with Lattice Engines, giving its clients access to a suite of predictive applications – also admits that not all the agency’s clients are ready for this new technology. He says: “Not all clients are ready for predictive tools as investment in other technology needs to be in place to realise its value. Clients that have already invested in, and are seeing value from, MA, CRM and other marketing technologies are now looking to extend their marketing technology stack by identifying tools that can enhance campaign performance and conversions. Integrating predictive analytics is the logical next step for organisations embarking on their digital marketing maturity journey.”
As well as having their tech in order, marketers will obviously need to have their data in order too. This doesn’t just mean having clean data but ensuring you are gathering the right data. Laura Ramos, VP, principal analyst at Forrester, says: “B2B marketers need to collect data that helps them understand their customers.”
You have to think more strategically about your data. It’s now possible to collect tonnes of the stuff, but it’s not all going to help you. The data you collect will depend on the predictions you’re looking to make. Ramos continues: “They [marketers] also need to learn to create meaningful hypotheses – good questions about what they really want to know – that the analytics can test. Finally, they need to put this new insight to work, not just selling what they have, but creating deeper customer relationships that benefit the entire organisation.”
Actioning the insights
Ramos’ last point about being able to action what predictive analytics revealed was echoed by several industry experts. It seems obvious: put into practice whatever it is the predictive technology preaches. But beware, this is the main challenge with predictive analytics for marketers, as they have to be able to understand how to extract the data. Westlake says: “In complex and real-world business scenarios, many [predictive] models are often interlinked and feed off each other creating additional complexity. Many companies can get sucked into this world of data complexity and data science. But the real benefit comes from the ‘action taken’.
Being able to black box the modelling aspects and focus on the action is a key skill for companies to truly exploit the power of predictive analytics.”
Tony Pearce, CEO at GRABR, warns marketers they may need help with making strategic decisions: “What an organisation could struggle with, though, is understanding how to extract the data to make strategic business decisions. It is advisable to invest in a team of analysts who have experience in analysing the data and are able to share the findings with the marketing team and explain how they can work together to create a tailored marketing plan that targets their customers.”
What can you predict?
According to Forrester’s report, From big data to actionable insight – The role of predictive analytics in B2B marketing, most of the actions being taken are currently along the same vein: “Few early adopters have made progress in predicting business opportunity beyond the customer acquisition stage. While improved lead analysis and scoring shorten sales cycles and eliminate time wasted on opportunities not yet ready to close, the real value of predictive marketing occurs when marketers apply it across the entire customer life cycle.”
According to Forrester’s report, predictive analytics is currently being used to predict which deals will convert faster. In the near future it will be used to determine which prospects to nurture and how and when to reach them while also identifying expansion and complementary opportunities with existing customers. In the long-term, it’s suggested predictive analytics will be able to help B2B marketers: ‘Find market white space and predict outcomes. It will uncover customer satisfiers and value delivered to increase loyalty, while helping to prevent defection and profile a business’ best customers and learn how to create advocacy.’
It’s clear there’s a lot of potential for the future of this technology and marketers have a long way to go before they reach its full potential. Other benefits include cost savings that come with more accurate predictions. Adi Clark, head of analysis, B2B and BI, Harte Hanks, says: “Applying predictive analytics to existing data can deliver significant bottom-line benefits. If you’re conducting an acquisition programme, it can reduce the need for third-party data or telemarketing support, which brings major cost savings.” Pearce also notes increased personalisation through predictive models can improve efficiency and reduce costs. As well as financial benefits, predictive analytics seems to join up two of the biggest talking points in B2B marketing right now: big data and customer-centricity.
Even though customers and prospects spend less time interacting with your sales staff now, you can gather more information on them than a salesperson ever could have hoped to collect thanks to their digital footprint. This wealth of data (aka big data) can now be fed into predictive analytics systems and your customers’ behaviour can be predicted. Forecasting how your customers will behave allows you to tailor your communications to them, therefore putting them at the centre of your marketing strategy and offering them greater value.
Time to forget the past?
With the ability to predict the behaviour of your prospects and customers, it might be easy to assume there will be less focus on retrospective ROI and more focus on what the predictive results show. However, as Bobbett explains: “While predictive analytics can tell us what we expect to happen, that is very different to telling us what actually did happen. The better the predictive modelling, the closer our expectation will be to the actual outcome. But it remains a prediction, not an absolute.”
Using predictive analytics may make proving ROI easier as it will focus your goals and objectives. You will be able to clearly identify what it is you’re going after, so by default measuring whether you achieved it or not will be a natural step. But marketing will not escape the pressures to prove ROI for the foreseeable future. Hurley agrees: “To build accountability across the organisation – from the board to the sales leaders and finance –marketing must show past results, even if the actions taken to achieve those results were determined from predictive analytics.”
While predictive analytics has the potential to identify new areas of investment and offer a true picture of the customer and their future behaviour, it’s not a panacea to all marketing problems. You’ll still be working to prove ROI and you’ll still need to work out the right messages to send to customers. Ramos advises: “Just because you can figure out the right audiences doesn’t mean you will also gain their attention. Matching the right content, message, or offer to their audience will still challenge B2B marketers, even those with mature predictive analytics experience.”
