So this post is for the planning and analytics geeks out there. I enjoy watching software applications categories undergo fundamental change, where real innovation starts to appear. And the nexus of planning and analytics is where this is happening.
Often the drivers behind these kind of changes can be technology based – such as the rise of mobile, or perhaps social or economy based – such as the rise of the self-employed economy. But when these external drivers gather momentum they often disrupt software categories. Some people go with the flow, others try and fight it.
With that in mind I read with interest an article recently that made the case for Analytics (or for the old school among us: Business Intelligence) and Budgeting/Planning applications being two separate worlds and really not needing to be unified together in a single application, that putting them together is just hype, not useful.
It was on the heels of a set of announcements from SAP, with SAP Cloud for Planning bringing together both analytics and planning – large scale analytics, data visualization, modeling and planning under the same unified hood, underpinned by SAP HANA. I personally think there is real innovation to be had at this nexus of analytics and planning, but more on that a little later.
The crux of the case in the article, was that Analytics is for the tech guys who get big data, data prep, data warehousing, SQL, unstructured and structured data etc., while Planning is for Finance, who worry about drivers, financial allocations, forecasts etc.
Different Disciplines, Different People?
I get it, my background hails from the data-warehousing, Business Intelligence, and Online Analytical Processing (OLAP). And to be honest, financial planning was a different world. When I built dashboards and analytics for organizations, (typically for the crew in IT) there was often a separate planning implementation going on in the room next door for Finance. Each side looked with somewhat distain over at the other (I preferred to write SQL than think about cost allocations).
When Business Intelligence first emerged in the mid-90s it was built by tech, for finance-IT – we’re talking star-schemas, semantic layers, and all that good stuff – distant from the world of finance. While when the first packaged planning apps for finance appeared, they were built as apps. New technology at the time like OLAP databases were optimized for modeling and what-if? analysis for finance – but had fewer dimensions and detail for analysis and weak ad hoc analysis, while big iron driven data-warehouses were optimized for large scale analysis, but couldn’t handle changes in assumptions inherent in the modeling and planning process.
So when the rounds of vendor consolidation with Business Objects, Cognos, and Hyperion happened in the mid 00’s it was two (often more) Business Intelligence (BI) and Corporate Performance Management (CPM) stacks. Two different categories, different skill sets, different code-bases. Vendors glued these two stacks together with a veneer of branding and single sign-on to make them like a suite, but they were really different code-bases and experiences beneath the thin integration.
Change is Underway.
But just the same way that NetSuite and Workday are reimagining their respective categories in ERP and HCM for the new economy, the same is beginning to gather pace in CPM. In ERP for example, eCommerce capabilities increasingly need to work with the ERP seamlessly, from web storefront to order -- because a digital storefront is often strategic. And HCM apps need to be mobile-first in an increasingly self-service world. So, CPM is undergoing a similar transformation, just differently.
CPM is changing because planning itself has to be more responsive, more in tune than ever before with the operating environment. And that requires analytics.
A recent Hackett Group survey showed that about a third of companies intend to implement rolling forecasting over the next few years. Combined, Hackett saw over half of companies building some kind of rolling forecasting process. Hackett attributed it to increased competitive pressures on companies, and faster moving markets. Companies want to see now just further out, they want to see their forecast adjusted based on a continually changing environment.
So doing a yearly plan/budget isn’t good enough anymore either. And because organizations are increasingly moving to rolling forecasts, it means ingesting ERP, HCM, and CRM data increasingly frequently. And more frequent planning and the push for more accurate forecasting means responding to external data too. Not all of this data needs to be in the plan itself, but the planning professional must be able to update planning drivers, change assumptions, and make course corrections in the face of the larger data landscape that they are expected to respond to - and they need to see that environment clearly.
The data landscape they're making decisions on is larger than before, and they’re being asked to re-plan and respond to that landscape faster. Planning no longer takes place in a vacuum, and it takes place more frequently, and closer to the business.
The dashboard vendors don’t have it easy either. Because standalone dashboards aren’t really good enough anymore either – they don’t have a call to action in them – just seeing a chart isn’t good enough – the expectation is you’ll do something about it. You either take action in your system of record – that’s why providers like NetSuite, Workday, and Salesforce provide embedded analytics. Or you plan and adjust based on those insights, using engines that combine analytics and planning, like SAP Cloud for Planning, Anaplan, and Adaptive Insights. But a standalone run-of-the-mill web based dashboard environment (and standalone planning environment) is deteriorating in value.
But really reimagining planning and analytics as a single unified solution means starting with a clean sheet of paper. Providers like SAP are taking the lead. Remember those data stores I mentioned earlier, one optimized for planning and the other optimized for large scale analysis? Well in-memory columnar databases like SAP HANA offer the opportunity to do both in the same database and data model, which makes it easier to model and plan in the context of large scale analytics. With data visualization operating on the same data store that's being used for analysis and planning, it's a potentially potent combination, blurring the lines between analysis and modeling.
So to do this right, it really helps to have a unified system – one database engine and model – the same engine serving both the analytics and the planning, one set of common definitions, one unified user experience, one business dictionary across both. It’s no longer just gluing these systems together anymore - like what happened over a decade ago, they have to be rethought in the context of where planning and analytics are headed, and designed together.
For once, this isn’t just vendor hype. As the nature of planning changes, a new opportunity opens up to rethink the systems that enable it.
Now time for that PB&J.