Analytics as a Portfolio


Share on LinkedIn

In the first post of this series, I laid out a set of principles for standing up a truly differentiating analytics capability in an enterprise analytics Center of Excellence:

  1. Business driven not exploratory analytics
  2. Investing in analytics as a portfolio
  3. Agile methods for analytics
  4. Creating operational strategies for every analysis
  5. Coupling analytics and reporting to drive data democratization
  6. Making collaboration a core capability not an add-on
  7. Creating virtuous cycles inside analytics

Before Thanksgiving, I wrote about the need for hypothesis driven analysis as the primary discovery paradigm for the enterprise. As that post makes clear, I’m no fan of unguided data exploration as a research paradigm. In that same post, I laid out some basic thoughts on the principles behind the selection and prioritization of analytics projects. Which leads me, fairly naturally, to the topic of today’s post – investing in analytics as a portfolio.

The point of having an Analytics CoE is to create a robust centralized capability that can provide mature, deep analytics across the organization and across data siloes. That cross-organization view is often an essential component of the CoE, but creates both problems and opportunities. The problems range from data quality to data integration to project selection. I’m interested right now in the last of these.

In the CoE world, you have to pick from lots of different projects across the enterprise and each, in its own way, is like an investment opportunity. Because of that, I think it’s both plausible and appropriate to borrow aspects of investment portfolio management and apply them to the problem of prioritization and selection.

Let’s start with the idea of risk. Whatever analytics folks like to imply, a flat-out majority of analytics projects bear little or no fruit. There are, of course, ways to improve the odds (which I’ll talk about in the next couple of posts). And knowing which projects have a higher or lower risk profile is critically important to building a balanced portfolio of analytics projects. Just like in the investment world, higher risk projects tend to have a higher upside, but what’s most important in this context is the recognition that projects often fail. This puts a premium on having multiple projects in hand. Analytics projects tend to be like venture capital investments. VC’s know that a fair number of their investments will fail, many will generate modest returns, a few will payback big. That’s why VC’s always have multiple investments in the pipeline. If you think analytics is similar (and I do), then a conservative strategy that selects and invests in projects one at a time may doom a CoE and is, because of it’s very conservatism, a poor investment strategy.

There’s another aspect of portfolio theory that also bears consideration: diversification. Diversification in investing is simply another strategy for controlling risk and it’s standard practice in building retirement plans. But not every type of investment manager focuses on diversification. VC’s, for example, usually focus on very specific types of business at very specific stages. It isn’t because diversification might not reduce their risk, it’s that there’s a premium in high-risk investing on knowledge. It’s much harder to be expert and to find synergies across multiple industries and stages of a business.

I think both dynamics are at work when building an analytics portfolio. On the one hand, selecting projects from across the enterprise legitimately reduces risk. It makes it more likely you’ll find the areas of your business that represent significant analytics opportunities, it reduces the dependence on a single (possibly flawed) data set, and it spreads the benefits of analytics across the organization (reducing political risk). On the other hand, you’ll likely sacrifice real synergies in data preparation, data quality, and analytics expertise.

Part of the answer to portfolio management lies in Agile team-based analytics (guiding principle #3), but I don’t think there’s one right approach to diversification.

That being said, I do think that a critical part of constructing a really good portfolio is being open a to wide range of analytics opportunities. We all tend to over-invest in what we know. Marketing folks tend to focus exclusively on acquisition opportunities. Operations folks tend to focus on cost-reduction opportunities. Whatever your primary focus, it’s easy to miss high-value opportunities that come from other disciplines or areas of interest.

If you’re really trying to build an enterprise analytics capability, make sure that you have people invested in a full-range of disciplines as stakeholders in the process. You want marketing, customer, operations, pricing, strategy, supply-chain – pretty much every discipline across the enterprise to have some kind of say. Having a wide range of interests and disciplines represented makes portfolio selection and evaluation more complex, but it makes it far more likely that you’ll find new, interesting, and impactful analytics projects to deliver.

Whatever your approach to enterprise-wide diversification, building an analytics portfolio lets you balance a number of critical factors inside the CoE. It allows you, for instance, to invest in long-term projects without sacrificing momentum around continuous improvement projects. Having an analytics pipeline lets you blend short and long term projects effectively.

It also lets you blend training needs with research needs. With multiple teams working on different projects, you have the ability to integrate new analysts and help mid-level analysts build new skills without necessarily slowing down deliverables.

It’s even (and this is obviously important to us) provides a seamless way to blend internal and external resources while facilitating knowledge transfer. What better for way for you internal analysts to learn new methods than by working side-by-side with consultants on a portfolio project.

Of course, many of these benefits come from a combination of a portfolio approach to analytics in conjunction with an Agile team analytics approach. The rationale and benefits for borrowing Agile development methodologies and applying them to an Analytics CoE will be the topic of next week’s pos!

Republished with author's permission from original post.

Gary Angel
Gary is the CEO of Digital Mortar. DM is the leading platform for in-store customer journey analytics. It provides near real-time reporting and analysis of how stores performed including full in-store funnel analysis, segmented customer journey analysis, staff evaluation and optimization, and compliance reporting. Prior to founding Digital Mortar, Gary led Ernst & Young's Digital Analytics practice. His previous company, Semphonic, was acquired by EY in 2013.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here