What organizations need today is a distributed order-to-cash platform that could leverage the power of data, analytics, and ML to identify and execute innovative business and revenue models with a focus on sustainability and profitability
As the global businesses got disrupted during Covid-19, it has led decision-makers and organizational leaders to rethink their approach to revenue management. An important but forgotten aspect of revenue management is the ‘Order-to-Cash process’. What I find surprising is that this very critical process that drives the value realization and literally ‘pays the bills’ is dismissed as an ‘IT Topic’. But guess what, there’s no other ‘business topic’ more important than order to cash.
Unfortunately, the O2C process is so deeply embedded in the ERP that it is difficult to even start finding its toots and tethers without some serious digging. Algorithmic businesses are agile and they need an agile solution to their order-to-cash requirements.
An article from Forbes explores the need for a digital Order-to-Cash system highlighting the inadequacies of ERP systems to effectively address the complexity involved. As highlighted in the article, “ERPs are great for backend financial processes, but they weren’t designed for managing and gathering this kind of dynamic revenue in the order-to-cash process. You do need ERPs. But the purpose of those solutions isn’t order-to-cash and never was. Order-to-cash is highly complex, and it’s going to get more complex as we move toward recurring revenue business models. ERPs are good at standard processes, but they can’t handle variability. This has proven increasingly true as business models become more complex and involve more recurring revenue.”
The real O2C process for Algorithmic Business requires a commercial value stream thinking towards solutions. Each value stream is powered by hundreds of data/AI assets that power multiple algorithmic products and services. These value streams in turn power multiple offerings, contracts, and revenue streams.
The real data-driven, agile O2C process needs to integrate with these commercial value streams and offer flexible monetization solutions to combine these products and services in multiple customer-aligned offering SKUs and billing models.
What do we really mean by an agile Order2Cash platform?
Attached below is the Boston Consulting Group (BCG) O2C platform model. On one hand, it has good segregation between data, workflow, customer journey, and teams, and that makes the concept really good.
However, a practical implementation of agile O2C with #commercialvaluestream and #datamesh and #dataproducts requires a lean, microservices-based workflow design while the O2C platform itself should contain atomic workflow components and AI/ML-based features for orchestration and governance. Also, the data layer itself need not be a part of the O2C platform and should rather interface with the organizational data fabric.
The reason why O2C has gained tremendous prominence recently is that Order Fulfilment lies at the heart of customer experience and the Australian Retailers Association seems to agree. There’s nothing more customers want than getting the right product at the right time at the right price.
“Retailers will need to achieve this at increasingly efficient speeds and effectively use data to deliver a service at a hyper-personalized level, while at the same time contending with higher expectations around sustainable delivery and procurement practices.” – ARA
In addition, an IBM article on the topic captures the essence with the following simple formula
Fulfillment = loyalty
It further explains by saying that:
“The disruption in retail we’ve seen over the past twenty years has positioned fulfillment at the heart of the customer experience”.
Due to the increasing importance of e-commerce across industry segments including B2B, what’s been true of retail is now also true for every industry. Customer experience is invariably linked to the ‘fulfillment experience’ at every transaction. Unfortunately, a large percentage of the recent investments in customer experience have targeted convenience instead of fulfillment.
The retail mindset is fundamentally changing from managing merchandise inventory “stock it and they will come” to personalizing product mix and pricing around customer lifestyle and demographics “they will come IF we offer what they want”.
While many retailers are working on integrating customer journeys and gathering customer intelligence across platforms and touchpoints, very few are able to leverage these insights in their fulfillment and supply chain strategy.
Every store requires its own ‘identity’ around its customer landscape to offer the right product mix, pricing, and service levels. Hence, the stores set the terms of the contract, not the central supply chain!
The O2C process needs to become agile, adaptive ad responsive to the diverse fulfillment needs of the stores while still being cost-efficient and robust. This is a multi-dimensional optimization problem.
Is this even possible in traditional ERPs?
According to experts, ERPs and Analytics platforms are designed for a separate set of requirements and have different jobs.
The ERP systems are about efficiency, scale, and doing things right i.e., executing hundreds and thousands of repeatable processes every day on time without any error.
The analytics systems/platforms are designed for exploring new sources of value and delivering and realizing the value at scale and at speed.
ERP systems have not been designed for this type of hyper-personalization of offerings and services with real-time scoring using millions of industrialized models. It is a different type of requirement.
The most impactful use cases, therefore, use a combination of ERP and other data where we can integrate granular financial data from ERP from other operational data (e.g. from the manufacturing floor) and customer data (e.g. from CRM) to start developing an integrated and holistic view of the entire business landscape. The business can then slice and dice this holistic ‘single source of truth’ across domains and reuse the granular data using a logical data model matrix.
And to ensure the utility of the single source of truth, it is imperative that the integration of data is done carefully based on ‘actual utilization’ i.e. integrate the data that is being queried all the time by everybody and then work incrementally from there. You don’t just want to work agile in the project phase but you want an agile solution in production where you can quickly pivot (be agile) depending on the business issues and conditions (What-if Scenarios).
It is true that the ERPs of today were designed and implemented by the Business Process Re-engineering (BPR) experts of the yesteryears and they designed them based on how the business operated at the time and what it needed. Also, the philosophy at the time was to eliminate the variability across regions, products, and functions and to transform the business to operate in ‘one right way.’ But today, businesses will themselves be restricted by the one right way. The one right way does not work anymore because customers demand personalization and variety. So now with all the agility and data, we need a new set of agile processes outside the system. The most innovative ERPs of today will become relics in the next 10 years and we will still need them for high-speed transaction processing, but they will not be the ’focal’ of the business. The new focal will be the digital O2C platforms and customer context management tools
One way of looking at this hyper-personalized product packaging based upon the customer’s predicted behavioral and performance propensities is through the lens of ‘Nanoeconomics’ coined by the Dean of Big Data – Bill Schmarzo. The Analytics systems seek to re-invent processes by:
- Understanding the sources of value creation and value impediments, and then
- Building learning systems that continuously seek to learn, adapt and respond to how value is delivered
According to another expert, there is a clear difference between personalization and individualization. Personalization is almost invariably from the corporate product owner’s perspective. It is about what is the right micro-marketing offer to make to the customer during their current interaction. It is merely an extension of traditional direct marketing, albeit on a much smaller scale.
In contrast, individualization is about the customer’s contextual perspective. It is about what is the right service to provide the customer during their current interaction. It could be a product offer, but it could also be a service, support, or something else that is relevant to the customer.
Both are powered by analytics. Machine Learning through Deep Reinforcement Learning offers great opportunities going forward. We know that personalization has historically low response rates (0.01-0-05% for digital according to the IDM). There is not much data about individualization, and we would be surprised if it were not significantly better.
Nanoeconomics gives us the power to segment user personas into the tiniest segments possible. Classical marketing stops at ’mass customization’ or ’segment of one’ but why stop there? We could segment a user based on what they need in summer vs winter or day vs night or before and after watching television. Individualization is the segmentation bit that leverages contextual data to design segments and personalization is then to figure out what can we offer to which segment and why. And both are important. The first defines ’where to build’ and the second – ’what to build’
According to the experts, personalization is clearly the dominant strategy today. But it is mostly about products. Individualization will likely increase in importance as cookies are progressively turned off and contextual analytics replaces behavioral analytics. The more we drive contact optimization through the lens of the customer, the better it will be for all concerned.
People can belong to multiple behavioral groups based upon their propensities, preferences, associations, relationships, inclinations, etc. It’s not 1-to-1 marketing, it’s not 1-to-many marketing. It is many-to-one marketing in that people can belong to multiple behavioral groups.
Just as we are all typically members of 5-7 different social networks, so we exhibit subtly different behaviors in different contexts. Although our beliefs, desires, and intentions are unlikely to change in the short-term and therefore the differences are subtle.
Key learning for us from the experts is that “individualized” experience requires that organizations focus on collecting and analyzing the transaction and engagement data to try to determine customer or user intent. Creating an “intelligent” user experience requires leveraging AI/ML to analyze a deep history of the user’s interactions to determine the user’s intentions, and then coupling those intentions with current trends, patterns, and relationships to match those intentions with a deep understanding of the available content to recommend the most relevant action.
Order to Cash (O2C), and the As-a-Service (aaS) business model.
McKinsey & Company highlights several ‘tension points’ in the O2C process as companies change from product-based deal architecture to an aaS architecture.
The key point is that customer relationship – the aaS model requires companies to become “trusted advisors” to their customers and to orchestrate their processes, systems, and people around the customer pain points and context to deliver on promised value and ROI. The setup needs to be adaptive to changing business needs of the customer. Services are not about you, it’s all about the customers. They need to integrate into the customer’s ecosystem and must solve customers’ needs and problems.
And this is the main challenge where ERPs fall behind – there is no ‘One Right Way’ in aaS business that can be implemented and optimized in the ERP. There are many paths to customer satisfaction and some of them are profitable. You need to iterate and experiment to figure out what works best for you. And in those experiments, you will find unexpected profit pools.
According to yet another expert in this domain, a good O2C platform should help you in making a successful trade-off between ‘designing customer-centric O2C value streams for each customer’ while at the same time ‘helping you in achieving company/supply-chain-wide profitability goals.’
One of the companies was able to reduce their working capital and receive incoming payments on day #1 instead of #150 and close their books at the end of the quarter in 50% of the time. That was significant.
Healthy Cashflows and invoice-to-cash cycles are the lifeblood of any aaS/subscription business. Most companies struggle with accurate metering, on-time billing, invoicing, disputes, and collections issues. Others are already in deep red incurring and accumulating operational losses. Starting an aaS business is exciting and fun but operating it profitably is not. There are hundreds of partners and consulting companies that will help you set it up, you will hardly find anyone to help you when you get into a financial and accounting mess.
One CFO remarked in this context that a healthy cash flow is like treading water, it determines whether you are swimming or sinking. And this is because we want to create awareness around these mundane yet life-critical aspects of aaS business models.
These kinds of bottom-line improvements are not overly exciting or attractive to solve. But when done right, they’ll free up the working capital that can be reinvested to further fuel your company’s growth. Small and large companies alike often lack the level of detailed metrics to decide when and how to apply AI to solve such problems. For example, consider a metric like “re-work rate” i.e., what accuracy does the ML model need to have to be “better” than the baseline. Or what is the minimum number of payments needed to generate economies of scale to invest in applying ML in the first place which makes the business cases harder to calculate?
The most basic conditions for getting returns on investments are that:
1. The initiative breaks even operationally i.e generated a significant contribution margin and
2. It can self-sustain i.e there is enough free cash flow to be reinvested.
There is hardly any company that is explicitly targeting these goals. In their view, they are ’investing’ capital to cover ’operational expenses’. And somehow CFOs are silent over this misappropriation.
Pret-a-Manger Case Study
We’ve been discussing the potential pitfalls in the subscription business e.g. order fulfillment and cash flows and let’s look at this case to understand it better.
Pret A Manger, a UK-based restaurant chain, launched their coffee subscription at £20 per month for unlimited (i.e five Barista-made drinks like (organic coffees, teas, frappes, hot chocolates, and more a day). It did sound like a cool plan.
But soon the offer got oversubscribed and as of January 2022, Pret was struggling to ’fulfill’ its obligations i.e offering good quality coffee to subscribers resulting in customer disappointment.
Based on my independent analysis, there are seven things that could have gone wrong with their business model:
- Targetting was off – the service is being availed by existing customers more than new ones
- Under-pricing – the offer looks underpriced considering the high subscription rate
- Deep red cashflows – most coupons rely on a redemption rate of 2-5%. But in this case, it’s much higher (linked to pt. 1)
- Demand was under-estimated causing inventory stockouts (linked to pt. 2)
- Lack of communication – e.g. it could have been a ‘boardroom move’ without involving store representatives and now stores are under-staffed
- No bundling – probably someone thought a coffee subscription will boost sales across the segments but it doesn’t work if coffee is already a ’best-seller’. Then you are just killing the cash cow
- Pret unknowingly enabled competition and started a price war – clearly, there is demand for coffee subscriptions and Pret can’t fulfill it. That’s an invitation to competitors.
This case is a reminder of how the Subscription Business Model could be a double-edged sword. And sometimes great ideas can get derailed by poor execution
A simple basic ‘back of the envelop’ calculation, reveals that the subscription price should not be anything below £49 (assuming ~60% redemption rate) because
- Pret is clearly targeting ’regular’ coffee drinkers
- Pret know that these people take more than 1 cup per day (let’s take 2 per day).
At a nominal cost of £0.8, the cost alone per subscription would be £35 per month. At the promised 5 cups per day (i.e if everyone redeemed 100%), it would be £88. So even if Pret followed the loss-leader strategy and were willing to offer coffee at a 30% discount, their estimated redemption rate should be < 32% for this business model to work successfully. And this is unreasonable for a staple food like coffee. Moreover, they’ve not mentioned any time frame so this doesn’t look like a promotional offer. And then there is no bundled offer on food or space.
The issue with the £20 per month pricing is that they are attracting the wrong customer segment. Pret’s real target is the ‘Cafe Workers’ i.e. freelancers, professionals, part-time workers who look for a place to work and ‘pay with coffee’. Their budget for coffee would be somewhere between £90 – £150 per month. For them, a coffee subscription for £49 or even £65 will make sense as it reduces their monthly cost and make them Pret’s regular customer.
At £20 per month, they are getting more ‘casual cafe visitors’, people who go to the cafe once in a while but are now simply the discounted coffee with no benefit to Pret.
The new demands of hyper-personalized products and services are forcing companies to explore new avenues of business creation and new business models to capture value. The traditional ERP-based order-to-cash models are clearly unfit for the job. What organizations need is a distributed order-to-cash platform that could leverage the power of data, analytics, and ML to identify and execute innovative business and revenue models with a focus on sustainability and profitability. The magic lies in achieving the perfect balance between ‘designing customer-centric O2C value streams for each customer’ while at the same time ‘helping you in achieving company/supply-chain-wide profitability goals. And as we observed from the case study, while it has become so much easier to develop and launch innovative business models, it is becoming increasingly difficult to operate them profitably and generate significant ROI in the end.
This is a co-created article based on the comments and feedback from multiple experts on LinkedIn.
Somil Gupta is an AI Strategy & Monetization Advisor, in Sweden. He specializes in developing commercialization and value realization strategy to grow data-driven businesses and monetize investments in data and AI. He has been awarded the Nordic Data and AI Influencer 2021 award for his work in this field.