Tech major IBM has reimagined how data intelligence can underpin procurement. Rather than roll-out a specific tool, the firm focused on technologies that would enhance the existing skills of the workforce to result in time and cost savings.
When setting out on its global procurement digitisation journey, IBM’s goal was to be the premier cognitive procurement enterprise. Described as a multi-faceted approach to turning the procurement organisation into ‘one people’, the project aimed to transform the skills IBM already held within procurement.
This would rely on people, process, technology and data coming together through intelligent workflows to enhance the speed of work, create engaged employees and benefit clients. But reinventing and building a cognitive procurement enterprise would require a systematic approach across four dimensions: trusted data, intelligent workflows, exponential technologies and the experience of employees.
Communicating the strategy
Marco Romano, procurement data and analytics officer, IBM Global Procurement, explains: “This transformation is not purely a technology project. It’s encompassing of people, it’s encompassing of the processes, so we’re reimagining the process of how we execute work to be what we believe is best-in-class.
“We wanted to go well beyond the traditional procurement skills; to include business acumen, understanding of the markets that we operate in, our clients’ operating suppliers, developing a growth mindset, and also the soft skills typically associated with consultancy firms.”
At the initial phase of the project, the IBM Garage approach was applied; this is the company’s vehicle to enable businesses to accelerate by behaving more like start-ups. Discovery sessions were held with stakeholders at the centre of the process to ensure they were critical and fundamental to the design thinking behind the project, getting their voices heard and input made. What’s more, Romano says the Garage approach that delivers a 75% reduction in design time and 33% decrease in development time.
Not only are development costs lower, but clients or stakeholders are engaged at every step of the process, which increases the likelihood of user-satisfaction, and translates into improved ROI. The response from internal stakeholders at IBM has been “phenomenal” says Romano, because they’ve been involved from day zero.
“They understand what they’re getting, and they design what they’re getting for their business needs. And it really helps with organisational change. Having stakeholders ingrained in the process means they become part of the organisational change management. So, even though we still put the rigour behind change, this plays a significant part,” says Romano.
Intelligent workflows are critical for delivering as both a cloud and cognitive enterprise. In this way, IBM has achieved a comprehensive overview of how work gets done, allowing intelligent information to underpin the external client and employee journeys – and then serve as the “enterprise source of truth” for workflow, cognitive assets and metrics, from source to pay.
“The basis is to get the data right, so data foundation is absolutely critical to intelligent workflows, and it’s absolutely critical to the cognitive enterprise,” says Romano. “It’s about creating trusted data that’s timely and helps drive business outcomes, that’s the key thing. The role of procurement is not to process transactions; systems can process transactions. Our role as a procurement function is to influence and shape business outcomes, together with our stakeholders, and that’s what the cognitive procurement enterprise is about.”
It’s not just decision-making, he says, but this with added speed, urgency and transparency. It provides clients with unprecedented speed compared to a traditional procurement environment “where things tend to go into a black hole and they’re having to navigate the complexities of procurement”. While the knowledge derived from the intelligence in the workflow provides insight into areas such as competitiveness, market trends and analysis into market sentiment, such as analysing what pricing will do over a period of time for certain categories, these are all insights that can be collectively brought together in a meaningful way, to deliver the value to the client that helps them make decisions.
Lead with process, not tech
This is not a single system, Romano points out. In order to create intelligent workflows, IBM is implementing a platform which takes a thin layer of technology and puts that on top of the enterprise application or the point solutions that exist today, in many procurement organisations, before connecting the processes and the data to give an integrated workflow and a significantly different user experience.
It was about understanding what the processes were, identifying the pain points in those processes, and then redesigning them, he explains.
“We didn’t lead with technology, we rather led with reimagining how we have to execute ‘work’ to make it smart, and then you start to lay the technologies,” says Romano. “On top of that, we needed to build the architecture, because the platform is the glue that holds it all together, but from a procurement perspective we started with the processes.”
And if those traditional enterprise applications or procurement applications tend to be standardised processes that live in very siloed environments with artificial barriers around the procurement function, the project here was to traverse those traditional organisational boundaries upstream to clients, so that the exchange of data and information could be far more pervasive and transparent. And then it goes down through to suppliers, where they can achieve visibility and understanding of a client’s demand. Not waiting for the client to ask, but helping them execute and develop their business.
“You have headlights,” says Romano, “as to what their third-party requirements are going to be, so that you can act and shape as they build out their business, rather than reacting to a business that’s already been built and now you’re just executing the sourcing process. This enables suppliers to be proactive and go to the clients – to help influence how they might structure their process and their strategies. It may be a bit of a cliché, but it means really being a business partner to the stakeholder inside, rather than an execution operation, and helping them to shape their business.”
Analysing data insights
Typically, understanding competitiveness for quotes or complex RFPs could take a lot of time for a buyer to execute, particularly when it involves many thousands of line items in complex environments. Procurement professionals are having to do market research, which can take anything from hours to multiple days. Now, with the exponential technologies that IBM is applying as part of its cognitive procurement strategy, the pricing analysis cycle time is reduced to between five and 15 minutes, depending on what the buyer is trying to execute. That means the buyer is not spending time putting spreadsheets and data together, but analysing the insights that the data presents to them. Then they are free to work with stakeholders to execute change and to influence that track.
As part of the project, IBM implemented advanced contract analytics to manage contract-authoring and risk; cognitive pricing to negotiate better rates using market insights of inflight deals; blockchain to seamlessly onboard new suppliers and eliminate disputes; and prescriptive analytics to enable self-service for buyers. This all led to a better human experience, says Romano, pointing out that the project achieved a 90% reduction in time spent on batch analysis of multiple contracts through AI-supported search tools. For a company with more than 175,000 active contracts, that was a huge leap forward.
“For our practitioners to go and manually perform that exercise would be absolutely horrendous. But with the contracting capabilities that we’ve put in place, it really allows practitioners to analyse the corpus of contracts and the impact that a clause change or legislation change has on that corpus. It automates the interrogation of the corpus of contracts, versus having to do that manually, so they can very quickly understand which contracts need a change made to them,” says Romano.
The new system also led to better business outcomes, such as $180m in total cost savings through supplier negotiations, enabled by data insights and a supplier onboarding time that’s 10 times faster thanks to a blockchain solution called Trust Your Supplier. This is a digital passport for suppliers, which contains the typical credentials that an organisation needs to know in order to onboard them, such as certifications, and complex legal and regulatory compliance documents.
The system allows for the upload and continual secure update of those details, which can be used by multiple customers throughout the IBM business through the blockchain to onboard a supplier, without having to recreate or asking the supplier to redo the whole onboarding process.
However, despite the resounding success of the cognitive procurement project, Romano says the company was acutely aware of the need for facing up to risks and lessons learnt along the way. For example, in order to combat data challenges it created a centralised data lake and standards for it, so a unified data taxonomy and strategy could be maintained. A cloud enterprise data platform acts as a central data lake for all mission-critical AI assets.
There was also a perceived risk that people would simply default to old ways of working, so IBM prioritised the need for outcome-driven innovation by committing to well-defined and agreed-upon benefit targets. To ensure visibility and buy-in across the enterprise, they created compelling narratives that were frequently shared across the business.
Having almost completed the rollout of this programme within IBM, using itself as client zero, the firm, which operates in 175 countries, will now look to provide similar services for external clients. And looking to the future, IBM Global Procurement is planning to implement a best-in-class accounts-payable intelligent workflow through invoice analysis and dynamic risk modelling, in order to automate payments and reduce errors and fraud.
It also aims to build what it calls an ‘AI- infused virtual assistant’ to streamline the requester-buyer experience and reduce manual efforts, while digitising the middle office for efficient requisition handling. The end goal, explains Romano, is to create a symbiotic relationship, where the machine augments the human and the human in turn augments the machine.