What you need to know:

• Real time decisioning is arming banks, insurance and telco businesses with the potential to drive strong lifetime customer value, but it's also compelling culture change and driving brands into new service areas.
• The costs are significant, but the reward is outsized returns according to industry analysts.
• But selling to CMOs may be misguided, per one. 
• Traditional marketing clouds compete across some of the capabilities but can only go so far (and are expected to try and buy their way in to the game in future).
• For now companies like Pega (especially in Australia) and SAS (globally) are winning business, with Teradata and Medallia-owned Thunderhead also feature (although Medallia's acquisition of Thunderhead could hurt it as key staff were sent packing).
• Salesforce, initially through its Evergage acquisition, and more recently with AI and Data Cloud packages, is building capability, but still has a way to go, per analysts.

When NAB opted to bring in Pegasystems almost a year ago it added to the swelling list of Australian banks and financial institutions that are betting large on the ability of AI-fuelled real time decisioning to drive improvements in customer lifetime value (CLV).

It is also a recognition that while traditional marketing clouds can deliver solid performance improvements, there are limits to their capabilities.

As NAB CMO Suzana Ristevski last month told Mi3: “We wanted to create ‘one way, same way’ in our ability to do marketing communications. That sounds simple, but for a bank, it's actually quite complex, because we are piped into a number of different platforms. So it's been an incredibly complex process,” she says. But NAB is moving through the gears, though there are still a couple of pieces of the stack to swap out. "We haven't quite got off a couple of systems that we want to get off,” Ristevski admitted.

NAB joins the likes of Commbank, Suncorp and ANZ Plus, Bupa, Optus and TPG in choosing to invest in the kinds of capabilities that address some of the unique challenges for enterprises in their sector. AGL and Toyota are also believed to have implementations.

Westpac also apparently considered the approach earlier this year, but has stayed the course for now on a more traditional path, according to two separate project insiders. An RFP was issued in January for journey orchestration and decisioning. However,  this was eventually descoped to only focus on orchestration, with Adobe via its Journey Optimizer product beating out Salesforce (likely a combination of Journey Builder and Evergage which is now called Salesforce Marketing Cloud Personalisation)  and Pega's Customer Decision Hub. 

The local shift to decisioning reflects overseas experiences, where companies like Dutch headquartered Rabobank, and US multinational Wells Fargo have likewise forged a path now proving lucrative.

It's the customer, stupid

The shift also reflects an unmistakable reality – customer experience and a philosophy of customer obsession is central to building lasting and profitable relationships. Traditional campaign-based martech set-ups can only carry a banking brand so far. The problem is, big ships take time to turn.

Indeed, in its most recent Australia Banking Customer Experience Index Rankings, released late 2022, Forrester Research revealed that Australia's Big Four – which dominate the local retail banking sector – have seen their mid-tier competitors overtake them in great CX, and not just marginally.

Per Forrester: "The average CX Index score for the five Australian banks (including all of the Big Four) we measured in both 2021 and 2022 declined precipitously, from 69.2 to 60.5 on our 100-point scale."

That has direct impact on the P&L: "Customer experience leaders grow revenue faster than CX laggards, cut costs, reduce risk, and can charge more for their products," stated Forrester.

Of the eight banks it studied, the Big Four occupied the lowest four spots on the index.

That's also translating into market share losses in key segments such as home loans. As KPMG noted in an analysis of the Big Four's FY2022 results late last year, "Increased competition both amongst the Majors and from challenger banks and non-bank lenders, particularly impacting home lending margins. The combined Majors’ home lending market share has decreased from 78.6 per cent at the start of 2019 to 75.3 per cent."

A different study – The Australia New Zealand Banking Report by CX vendor Verint – revealed an almost comical disloyalty within the next generation of banking consumers. Raised in a digital era, largely locked out of the housing market, and now struggling with a high inflation, low wage growth economy, they risk leaving marketers chasing housefly-like lifetime value.

Per the Verint study, "Younger consumers are less loyal: Over 40 per cent of people under 45 years old would switch banks if it took no effort on their part, showing that low-friction banking presents its own challenges."

But there are opportunities to earn loyalty: "Our data shows that many younger consumers need tools to help manage subscriptions or track expenses could be vital to avoid customer churn."

Which also presents acquisition opportunity. According to the report: "An easy-to-use mobile app tops the list of factors that drive customer satisfaction and influence a customer choosing a new bank. On top of this, more than 60 per cent of new accounts are opened using a website or app, and three-quarters of transactions happen using digital channels."

Combine slick, user friendly apps with genuine utility and the winners emerge: "Nearly 50 per cent of people aged under 44 need assistance when it comes to cutting costs, creating a budget, tracking expenses and tracking subscriptions which is significantly higher than for consumers over 45."

Decision time

According to Aarron Spinley, the former Australian head of Thunderhead, one of four key global decisioning players, along with Pega, SAS, and Teradata according to Forrester, "If you are a big brand, that's got complex consumer channels – banking, telco, utilities, large retailer – you really can't live without decisioning."

Now an independent analyst and free of corporate constraint, Spinley recently authored a whitepaper highly critical of the approach marketers take to customer experience. He told Mi3: "There has been a marked increase in CX investment from major brands all around the world as it relates to the technology, yet net outcomes for customers are worsening. Research shows consumers are incredibly annoyed by pushy sales interventions, comically bad chatbots, soul destroying call centre wait times, and an inability to get things done quickly and easily”

The decisions brands need help making are not simply what to flog next, he suggested.

"The root cause is that we are training digital teams to try to trace every interaction that they have with the customer as a sales interaction or as part of the sales process. There's no virtually no interest in the actual intention of a customer or what they're trying to get done. It's always manifests as sales hacks, sales messaging, and spammy chatbots."

Mi3 has identified a number of major decisioning implementations across banking, finance and insurance that seem to align to Spinley's thinking of where the focus should be. 

Across some of the implementations;

• Commbank: CBA moved early on decisioning, with seven years experience on the technology. Its customer engagement engine makes 53 million calls a day, and allows the bank to run thousands of experiments simultaneously. It is also helping it to expand the boundaries of its services. The bank's Benefits Finder service has put an average of $710 into the pockets of customers utilising the app, facilitating payment of more than a billion in grants into communities while its Smart Weather model means the bank can contact customers ahead of damaging weather events enabling them to better prepare and making it easier to provide remedial support in a crisis.
• BUPA: In a sector that often provides notoriously bad customer experiences according to a recent Qualtrics study, BUPA is driving a fast path to value from its decisioning investment. Lisa Dickson, Head of Marketing Automation told Mi3 it has already delivered strong incremental revenue growth, cost savings of circa $1m, a 50 per cent uplift in efficiencies and a 30 per cent improvement in customer responsiveness. It is also notching strong NPS results for new initiatives.
• NAB: It is still early days – the contract was only signed last September – but the decisioning engine implementation is one of the final pieces in a three year, $45m, overhaul of marketing technology, that has seen the bank switch almost its entire legacy tech stack, according to CMO Suzana Ristevski. The grail is personalisation at scale – and NAB has already driven 10x messaging growth since the overhaul commenced, moving from 50m messages to circa 500m.

The common thread across all three organisations (as well as at Suncorp, Optus and TPG) is the technology layer – in this case powered by Pega. Internationally SAS Institute is also well regarded with Thunderhead and Teradata making up the bulk of the rest of the market.

Clouds gather

Many other businesses have attempted to replicate decisioning capabilities using traditional marketing cloud stacks, but there are limitations, according to Forrester Research's VP and Principal Analyst, Rusty Warner. Forrester calls the category Real Time Interaction Management, or RTIM – and says marketers shouldn't be the ones making the decisions.

"Pega and SAS are selling to IT and data science people and in financial services, I think that's the right approach," per Warner. He's critical of those who say CMOs should own the AI strategy.

“I think that’s silly for any vertical, it would be especially silly for a bank or any other regulated industry. [Pega and SAS] are able to sell into those people who understand the data management strategy of the firm and the analytics and how that works. They also understand deeply the regulated environment. Both Pega and SAS have a deep understanding of credit, fraud, risk and anti-money laundering and Know Your Customer legislation, the way things differ around the world," says Warner.

"You just don't find that with the marketing vendors, as hard as they try. I just don't think that they have that same level of understanding when it comes to the governance and security, and everything that goes into these decisions in the banking world.”

Model customers

Both companies also understand the importance of models to these organisations, Warner says.

Banks in particular run masses of models simultaneously. “They end up with scores that are arbitrated through their business rules logic to come up with the final decision for the customer," per Warner.

“With Pega, they'll do what marketing people would do, which is to understand the customer's propensity to accept an offer. But they don't stop there. They obviously run the profit model for the bank to determine if there is long-term profitability. They look at a customer lifetime value (CLV) calculations, but looking at it truly in that forward-looking profit contribution way, whereas most martech tools think of CLV as cumulative revenue.”

He says both Pega and SAS are designed around their customers' focus on profitability. “They understand that kind of logic, and they don't stop there. So you've got the customer propensity on one side, and you've got the profit potential on the other side, but then they factor in the customer context, and they factor in the business context."

He added: "Someone like Andrew McMullan (chief data and analytics officer) at Commbank would say from the business perspective, we would make money if we offered the customer a loan, but we know that customer should be paying less for their gas and electricity because we see the direct debits. And we know that there's either government assistance for them because of their income or there are just better deals out there. There are things we can offer up to the customer from a business perspective, that don't necessarily make us immediate profit, but they are better for long term, customer lifetime value calculations, because we have that retention.”

That's pretty much a summary of what McMullan told Mi3 back in 2020.

Marketing-customer merge

It’s not just the level of enterprise investment required that means banks and telcos are more likely to invest in RTIM. Their business models hinge on long-term relationships with customers who use their products and services, but they don’t offer the vast number of products that categories such as retail,  travel, hospitality, media, and entertainment do, he says.

In many of those verticals, marketing’s primary objective is to drive revenue uplift by getting people to transact online, says Warner.

"Cross-channel marketing hubs, customer data platforms, personalisation engines, can do that. But for banking/telecoms (and insurance, healthcare, utilities,) there’s more need for marketing to align with customer experience teams to make better decisions that help customers realise value over time."

"These verticals are more likely to look at CLV from a net present value perspective, which is a forward-looking profit contribution model that measures return on investment. Transaction-oriented verticals tend to focus more on revenue uplift, acquisition, upsell, and cross-sell."

Banks and telcos do that, too, he acknowledges, but are also adding genuine utility to demonstrate their commitment to their customers.

Stack slicer

Pete Avery, a former Program Director, CXM for Suncorp turned independent consultant who has worked on some of Australia’s biggest decisioning implementations, has seen firsthand the limitations of marketing clouds when stacked up against the decisioning model.

"There's really two religions – who do I target versus what do I recommend," says Avery.

"Marketing clouds can deliver 'send these offers to these channels based on these segments' [solutions], but when you get to the arbitration and the real time context they can't really do it. So for example at Suncorp when I issued the RFP for decisioning, Adobe’s first response was 'do you mind if we partner with Pega?' But the answer was no – because architecturally they are not designed as a real time decision engine. Part of the problem here is around the paradigm – which [for decisioning] is customer-first with a unique approach inbound, authenticated interaction management.

"It's really around the context. If you think of a chatbot – back and forth back and forth in a conversation – I speak, you speak, we listen – the decision engine doesn't just ask you the question, it gives us an answer, and it does this in real time."

Avery believes Adobe probably should have bought a decisioning engine ten years ago. He says Salesforce thought it had with Evergage, acquired in 2020, but suggests it eventually came to the conclusion that the CDP platform was more of a personalisation tool.

But harnessing decisioning engines is also not yet a mass market approach. Avery says it is all about scale.

“Where decisioning works is really simple: Billion dollar organisations, millions and millions of customers, a broad range of products and services so people can buy more than one thing from you and you typically interact frequently with your customers.”

Hence decisioning starting to cut a swathe across the top end of town across telco banking and finance.

Culture shift

But for decisioning to go further requires more than just a willingness by corporates to make heavy technology investment. Avery says it necessitates a fundamental cultural reset.

Simply applying an operating model that combines marketing operations, inbound and outbound digital and paid media to a traditional financial services model wouldn’t work for an very important reason, per Avery.

“Historically, the financial services sector is built around product-centric, channel-centric, and business unit-centric thinking – it’s customer-last by design. Customer decisioning is about customer-first,” he suggests.

“Customer obsession must first have a strategy and then you want to build something once, and have it work across multiple channels," he says. (Which is NAB's stated intention.)

Avery gives a simple example – you are talking to a friend on the phone, sharing a tweet with them and sending a Whatsapp message, all in the one conversation seamlessly. He says that’s the experience brands need to create for customers using a decisioning engine.

“When you talk to friends across three different platforms you don't stop, and then start again, you keep the context.”

But brands must also better predict intent. “Why is someone's spending their time to come in and out, to look at my banking app, to come into my branch or to call my call centre? You want to service the customer. The first thing you're trying to do is ask 'how can I help you? What job are trying to get done?' This is not about sell, sell, sell. It is about managing the relationship.”

Cloud halfway there?

Marketing cloud can get you part of the way, says Liz Miller, VP and Principal Analyst, Constellation Research. Based on the customer's history and their transactions, marketing cloud can make real time decisions about which offers and promotions might be best."

“Take the example of a media company that has telephony, and broadband as well as basic cable. The decision to be made as I am delivering all of these emails and promotions is rather than giving everybody $10 off their next review, I want to give that [only] to my most valuable, most loyal customers. A lot of marketing automation solutions can do that for you.”

But the decisioning model goes further.

Where the complexity comes in, she says, is when organisations are smart enough to understand that loyalty and engagement are not necessarily the most profitable route to follow. 

“What you want is a decision engine or data layer that can sit on top and that to some degree is abstracted from the rest of your stack. [It needs to determine] ‘Okay, but are they really my most profitable customer?' Because you could have your most engaged customer who's been with you for decades, but never pays their bill on time. The cost increases, right. Don't reward bad behaviour.”

Hence according to Miller, “Where there are lots of different permutations to the decisions you can make, you need the type of advanced system that can bring in the data from across the enterprise, and make those complex real time calculations at scale.”

Regulations

The other particular circumstance that adds appeal to the decisioning model in the financial services sector is the impact of regulation, says Miller. The fact that Australia's Attorney General has raised concerns about decisioning suggests it is rising the regulatory agenda. 

“It varies drastically by country, based on the regulatory demands you have big questions around their model dynamics. Folks in finance are very used to having the regulators say 'No, your models can only do this, your models can only do that.' So models that determine who gets what type of loan, who gets what interest rate, those are already automated ML models, largely not driven on self-learning AI.”

Financial services firms need to demonstrate transparency and the measures they have put in place to make sure that there is no bias model that could lead unintentionally to discrimination. 

“There is a lot of focus models within financial institutions. You have to have a solution that keeps your data private … No financial institution in the world is going to be okay with their data just freely roaming outside.”

Closed shop

For now, traditional martech vendors have limited cause for panic about the rise of decisioning rivals. It is prohibitively expensive for many verticals, says Forrester's Warner, while silos may obscure the opportunity for those firms that could afford to play.

"Banks are going to spend three, five, eight million dollars a year, every year for the next five, seven or ten years. I've talked to banks that have been getting half a billion dollars in return. So the returns are huge, but it's massive investments," he says.

"You just don't see that kind of enterprise-wide thinking in some of the other verticals like retail, unless it's someone huge, or even in the hospitality industry, unless it's a big airline, or a big hotel chain like a Hilton.

"You just don't see it in those other verticals because of the investment – and the need to come together and drive that enterprise project."