The First Business Operations Super Platform

SAP sponsored this post, but I’ll start with something they don’t want me to say: SAP’s platform will put data teams on their heels. The high-value use cases data teams struggle to justify and implement come prepackaged into SAP’s current platform. Most internal-facing features on your data and AI roadmap work out of the box without additional fine-tuning or data team effort.

SAP has AI support for 39 use cases (internal processes and workflows) and plans to implement 50 more over the next 12 months. Change is coming for data teams that primarily support internal productivity and reporting initiatives. Teams will either shrink or adapt to take on a new role in customer-facing product teams.

The opportunity in that shift is massive for businesses and data teams. Firms invested in Ferrari-level data talent that they use like a golf cart. Data and AI products are the racetrack. As soon as CEOs feel the revenue acceleration, it’ll be a new game for data organizations.

SAP even nods in this direction with a new tagline: It’s GROW time. The business AI platform’s features are firmly focused on operational efficiency, but SAP believes that frees up data teams to drive growth. We’re all in the results business. The next phase of AI will look very different than the last one.

What’s The Big Deal? It’s All About The Data.

I wrote a book about what SAP presented at Sapphire. Its platform facilitates incrementally migrating the business’s operating model into a multi-technology core (apps, cloud, data, analytics, machine learning, and GenAI). It enables any business to implement a platform-based operating model and realize gains out of the box. That represents a significant leap for businesses at early maturity stages.

Access to data is a competitive advantage, and SAP has capitalized on its massive install base to build an equally massive workflow dataset. I teach this approach, and it is amazing to see a company like SAP proving it in the real world.

SAP’s platform is integrated into its customers’ workflows. As a result, the platform gathers data with business context (data about the process or workflow that generated it). 25,000 SAP customers opted into letting it use anonymized data for model training. That’s why SAP’s machine learning and AI features work out of the box and how it built comprehensive knowledge graphs.

Over the last several years, SAP has built a competitive advantage on its unique dataset. Low-cost access to workflows enables low-cost access to data. There isn’t another company with a workflow, business process, and business optimization dataset like SAP’s. Models live by their training data, so the business with the most complete dataset will deliver the most reliable models.

SAP can invest in supporting more workflows because, with its business model, returns scale faster than costs. It gets paid for the implementations multiple times.

For customers who want to implement those features independently, costs will often scale faster than returns. Their ROI comes from a single implementation, so the unit economics work for a limited number of use cases.

Most companies don’t have enough data to train models that reliably support their high-value internal processes. Data challenges create the most expensive and time-consuming barriers to realizing value from machine learning and AI. SAP has developed a solution that bypasses those challenges altogether.

What’s The Big Deal? Productivity Gains.

SAP integrated a Generative AI user interface (SAP Joule) into most of its modules and workflows. Users can manage those workflows by intent rather than task. Apple sounds like it’s reading from SAP’s playbook with App Intents. In both cases, the shift is more significant than most people realize.

Prior-generation platforms handle workflows one task at a time. My intent could be to hire someone. Managing that workflow by task means building the job description, getting it reviewed, posting it to job boards, filtering resumes, scheduling phone screens, etc. Managing by task means the user often bounces between multiple applications, and automation is limited because it can’t move across apps or link tasks together.

Managing a workflow by intent is very different. If a user asks Joule to help hire a Java Developer, the assistant works with other apps and data sources to orchestrate the workflow steps. The workflow can be managed from a single interface. The users perform fewer tasks themselves, and it takes less knowledge to do them.

Joule handles orchestration and reduces the complexity of interacting with data and multiple technologies. LLMs are excellent at detecting intents but not serving them from start to finish. Joule and Siri run on a similar orchestration paradigm. Apps are associated with the intents they serve, the tasks they support, and the data required to do it.

Once a platform manages workflows by intent, the tasks and apps are abstracted away. The user has never really cared what’s happening under the covers, and now they can focus on the outcome instead of completing the steps.

Abstraction also puts optimizing the workflow into the platform's hands. Users don’t need to be retrained when processes change, which has implications for transformation.

A New Approach To Technology & Transformation

According to SAP’s Board members, “Clearly articulating the value upfront defines a business-first approach to AI.” The platform helps customers link business goals with transformation and technology. SAP’s RISE is positioned to deliver ‘business transformation as a service.’ Abstracting the workflow steps behind a single interface means that users can get all the benefits of transformation by adopting Joule.

Signavio (SAP’s Business Process Management component) is one part of that approach. It recommends processes and workflows that could benefit from automation or technology-supported optimization. A suite of metrics helps the business prioritize initiatives based on value. It’s a powerful decision-support tool for opportunity discovery.

Clean Core is a set of standard business processes and technical best practices that accelerate technology delivery and capabilities maturity progression. The standards also lower the cost of creating technical solutions. Business leaders can benchmark their performance against industry averages through the platform.

Managing workflows by intent means that changes are implemented into the single interface.  The platform almost enforces transformation because “transforming to digital but still doing things the same way doesn’t deliver value. Value is realized by transforming the business alongside technology.”

Employees must overcome the pain of transformation, and SAP believes it must deliver an interface that lowers the pain of transformation. During the conference, SAP announced its planned acquisition of WalkMe, which calls itself a digital adoption platform. SAP sees the acquisition as another way to move employees past their initial agitation phase.

The Early Customer Story. What Are The Outcomes?

Customers want concrete metrics about tangible outcomes. Business leaders don’t know which products are real and which ones are smoke. They don’t know which promises can be delivered and which products will underperform. CIOs don’t understand how much they should be spending on AI and how returns grow as spending does.

At a high level, SAP customers outperform their peers by 7%. Across all customers, SAP’s models are 90% accurate on predictions 30 days out. SAP brought out several customers to share their outcomes.

Team Liquid is an eSports company that uses SAP to analyze game performance for its teams. They saved 10,000 work hours after making the switch. The company calls the gaming analysis data a competitive edge, allowing it to outperform other teams.

Companies from Delta to Churchill Downs discussed their results across marketing, HR, finance, and operations use cases. Each one reported productivity improvements and cost savings. They shared high internal satisfaction scores.

SAP brought up complexity and reliability use cases. Amazon’s Project Kuiper runs its supply chain and manufacturing processes on SAP. It will build 3000 satellites and ramp up manufacturing capacity to build 3 per day.

Apple runs its business and supply chain on SAP.

The company uses its own tooling internally. Its 4,000 consultants save 2 hours per day by using Joule for knowledge search tasks. The intelligent assistant has an 80% satisfaction rating.

The company also deployed a coding assistant for its proprietary language, ABAP. Users ran comparisons between Joule and ChatGPT. They reported that Joule delivers more accurate, detailed answers, including links to deeper dives. ChatGPT often delivered out-of-date information, and answers only covered surface-level details.

The message was, ‘The platform is reliable even in the most complex use cases and delivers value across business sizes and industries.’ When it comes to quantifying the value that data and AI bring, I think it’s too early. Most customers migrated for the value the cloud platform and ERP capabilities deliver.

A few, like Team Liquid and Delta, are seeing value from data and built-in analytics. Getting a sense of AI’s value is best summed up by one customer quote, “Joule was a nice to have, but it’s integrated into everything, so we started using it, and it’s extremely valuable.” How valuable? Not sure.

SAP Has A Head Start On The Operations Super Platform Race

I don’t think another business will duplicate what SAP currently delivers for at least two or three years (the time it would take to gather a similar dataset and integrate models into a similar platform). Hyper scalers don’t have workflow-relevant contextual data. They handle huge workloads but don’t have access to business workflows. Other ERPs don’t have the same scale or access to business-wide data.

SAP is the only one that can integrate a single copilot (Joule) into all enterprise workflows. It is also the only one that can implement a single interface that supports access to multiple copilots without requiring users to bounce between apps. We’re underestimating the current levels of copilot fatigue stemming from using multiple copilots vs. a single interface.

Public cloud adoption has turned a corner, and more customers are switching off legacy platforms. SAP has an advantage here as the only company that has successfully gotten customers to move from legacy to modern infrastructure.

Its AI strategy builds on that. If a business adopts the cloud, it gets data and AI, too. Those features just need to be turned on. One industry analyst delivered a dose of reality, “Customers loathe transformation and major upgrades.” They realize that the journey must start now, but that doesn’t mean business leaders are looking forward to the trip.

Nothing SAP delivered will make customers loathe the journey any less. Its platform will reduce the pain and complexity. SAP’s approach embraces continuous transformation rather than one big-bang event. That’s smart because this journey will never end.

I started with something SAP didn’t want to cover, and I’ll wrap up with an equally uncomfortable question. How long do legacy businesses have before they fall so far behind that they cannot catch up?

Data, analytics, machine learning, and AI accelerate improvement cycles. The operating model can be optimized faster, which means the business can compete on price in ways that its peers can’t. Reliable predictive and prescriptive models make businesses more forward-looking, which improves performance. How long until the gap grows to the point where legacy businesses are no longer competitive?

SAP’s leadership team repeatedly said that businesses must choose to transform because technology can’t force adoption or usage. How much time do businesses running on legacy systems have to make that decision?