Druva benefits from Host Analytics flexible modeling

SUMMARY:

Can you mix desktop and cloud in modeling environments? Druva believes it is essential given the current state of the art.

Mahesh Patel CFO Druva
Mahesh Patel, CFO Druva

Financial modeling has proven a perilous pursuit. Islands of information, ad-hoc spreadsheets, lack of the right kinds of data and lack of agility are just some of the practical problems that beset line of business planners attempting to model business scenarios. Those issues hold companies back regardless of whether modeling on a stand alone project basis or as part of a broader effort to understand how different models impact enterprise performance management (EPM.) Druva Inc., which provides cloud-based data protection solutions, found itself struggling to get the kind of agility it needs to model for a fast growing business. Enter Host Analytics new modeling capabilities.

When I spoke with Mahesh Patel, CFO Druva, it was clear that as a self confessed EPM nerd, he was keen to find the right solution. Patel says that while Druva is a SaaS business, he wasn’t comfortable with the idea of trying to get modeling done wholly in the cloud.

Everyone’s telling us that cloud is the way to go but for some things it just isn’t right – yet. There are limitations. Our people are used to having the depth of solution Excel provides and that needs to be on the desktop. Host Analytics was new to me, but it gives us a way of using the spreadsheet on the desktop while also giving us the agility we need. Data is in the cloud but the user experience is on the desktop.

That only explains part of the challenge:

Older products force us to remain in stagnant data – that wasn’t going to work. You can’t really get a history or trend out of Salesforce so don’t have good insight into which products are selling better but that is really important when you’re growing quickly. I’ve used a number of solutions like Anaplan, Tidemark, Adaptive, Hyperion and Cognos but none of them really have the modeling capabilities we need – or the near limitless dimensions that allow for experimenting with different scenarios at a price point that works for us.

Patel raises an important set of issues. Modeling is nothing new but the ability to bring finance and line of business close together has always been a challenge. Couple that with the need to use current data and traditional systems soon become limiting, especially when you’re trying to match with data coming out of SaaS systems.

What about implementation? Druva launched Host modeling for two parts of the organization. First was to review income versus expenditures based on headcount. That’s essentially a finance department issue. Recently, Druva launched for sales planning.

Financials took about six months and sales planning took about two months. The biggest issue was understanding what we needed and mapping that back to what we have available.

Again, it is interesting to note that handling the initial data requirements is always a key exercise in any EPM project. But what about the day to day outcomes?

One of the most common use cases is understanding what’s happening around quota carrying rep numbers. We need to intersect between sales headcount and results and then both model and change assumptions. On the opex side we use modeling to understand the impact via what-if initiatives. Example – what might happen if we increased head count in a particular region? That has both cost and revenue impacts. We are now iterative in our methods, both tracking and improving assumptions along the way. In turn, that means we can be a lot more accurate in the forecasts we build on top of tested and measured assumptions.

As we concluded our call it was apparent to me that Patel believes he’s found an economical solution that works for Druva’s environment and which helps both finance and line of business fulfill their respective roles but without the disruption normally associated with a project of this kind.

Launched earlier in the year, Host Analytics modeling was recently updated to extend predictive capabilities, auto-detect mobile devices and provide integration from data sources via API.