How the Tech Industry’s Analytics Success Depends on Time, Talent and Treasure

John DeSarbo

One might think that high-tech companies would be further along in their use of sales and marketing analytics. After all, these companies live and breathe technology, and they have an institutional understanding of data. Yet that’s not necessarily the case.

With the help of the Economist Intelligence Unit, ZS recently completed a cross-industry study to assess the state of analytics implementation and integration and found that companies’ results have been underwhelming. Among 448 respondents, some of whom hail from high-tech organizations, 70% rate sales and marketing analytics as “very” or “extremely” important to their competitive advantage, but just 2% report that they’ve managed to generate a “broad, positive impact” from their analytics investments thus far.

John DeSarbo, a principal at ZS and leader of the firm’s high-tech practice, talks about why companies in the industry haven’t made more progress and what they can do to improve.

Q: The study’s results showed that companies in all industries haven’t yet made much traction with sales and marketing analytics. Is that what you’re seeing in the tech sector?

A: Yes, and I’m not surprised by these results. People in the technology industry are very quick to adopt the buzzwords—“social selling,” “big data,” all of these things—but very few companies really understand the level of investment required to deliver the promise.

Q: Is it wrong to assume that technology companies should be further along than other industries?

A: They may actually be further ahead than some industries, but not by much. Their products lend themselves more readily to data capture, but that doesn’t mean that their capabilities have kept up with that.

Some companies are just starting to do basic reporting or maybe some form of business intelligence based on simple metrics, but those all tend to be retrospective. I think companies are starting to realize that they have to go beyond that and develop more actionable recommendations from analytics.

Q: What are the biggest stumbling blocks?

A: In part, the challenge is time. Companies face a lot of pressure to meet short-term requirements. If someone’s boss is screaming for something in the next quarter or two, even in the current fiscal year, that’s a pretty short window of time to build a very robust analytics capability, particularly if the company still has issues with its data. They might not have clean data, they might not have integrated their data or they may not even have identified the right data streams. That data effort, itself, could take a year before they can even start applying some of the analytics.

The sheer availability of data is another issue. Storage is cheap. Whether it’s marketing interactions, sales interactions or customer behavior, there’s no shortage of it. And because it’s easy to capture data, everyone captures it, but no one has taken on the responsibility of holding all of it together and stewarding it, using the same data structure and definitions, and with consistent metrics to measure.

As our colleague Bhargav Mantha has written, there are “four Vs” to big data—volume, variety, velocity and veracity—and without mastering these, it’s impossible to deliver the most important “v”: value.


READ REPORT: Broken links: Why analytics investments have yet to pay off

Q: What about the distribution aspect of tech?

A: It’s an issue, as well. So many of the sales and marketing interactions happen through third parties—meaning channel partners, which touch a significant portion of the revenue for many companies. That complicates the problem because the manufacturer may not have all of the data. They may have to collaborate with a channel partner to collect some of that data, or they may need to offer the analytics capability to the channel partner to entice them to share the data. And, in fact, they may need to apply analytics to understand the value of the channel partners, themselves.

Q: Are there specific applications of sales and marketing analytics that are more relevant for tech companies?

A: One thing that may be important for tech, though may not be unique, is telemetry data. When customers interact with a piece of software or hardware, they generate a ton of data. Assuming they opt in to make that data available, companies can track it.

There’s immense insight that can be gathered around customers and their behaviors that can be incredibly predictive. What’s the best offer? How do people use a particular feature? What’s the likelihood of churn? And what are the things that happen just before the product “breaks”? That’s a tremendous opportunity, but very few companies are using telemetry data to their advantage.

Q: The study also identified problems at the front end of the analytics value chain in how people frame the problem, and the back end in how they translate insights into meaningful actions that affect customers. How do those challenges affect tech companies?

A: In both cases, it comes down to context. On the front end, I think that has more to do with the analytics team really not understanding the context of the problem. Most analytics teams have been built as a reporting function. They don’t really have the ability to engage the marketing and sales teams to understand the context of the business issue and, based on that, identify the data they’d need to address the question through analytics.

And similarly, on the back end, the analytics team may consider it a success if they get a good, precise answer, but maybe the high degree of precision isn’t as important as putting it back in the context of specific actions to prescribe for the sales or marketing team.

Q: So how should tech companies begin to address these challenges?

A: The first point to understand is that there really isn’t a silver bullet. Tech companies—and tech executives in particular—are always keen to find some new, whiz-bang application that’s going to solve their problem. There are thousands of companies out there that offer analytics platforms and tools, and companies have spent billions of dollars, yet we’re really not that much better off.

When you look under the hoods of those systems, the “black-box predictive analytics tool” is usually a bunch of Ph.D.s sitting in a back room configuring analytics. You still need someone to figure out the right model to frame the problem, and you need someone to interpret the results and translate them to specific actions on the back end.

Success requires time, talent and treasure. It takes time just to build an understanding of the business and aggregate the data. It takes talent: finding those people who are really good at the analytics, people who really understand the business, and that rare intersection of people who understand both. And it takes treasure: Building an analytics capability is not cheap.

Most companies are constrained in at least one of those areas, meaning they need correspondingly more of the other two, so it’s a question of mapping out the most important questions you’re trying to answer and then building a road map of how you get there.

Last, executives need to manage expectations. Pick some proof-of-concept project in one area and work on it until you have some success. That way, you can start to build momentum based on wins, rather than starting a bunch of initiatives in every different part of the company.

For more analysis and insights from the study, read “Broken links: Why analytics investments have yet to pay off.

About the Expert

John DeSarbo is a principal in ZS’s San Mateo, Calif., office and leads ZS’s high-tech and sales channel strategy and management practices. He has more than two decades of consulting and industry experience focused on sales and marketing strategy and operations. His areas of expertise include multichannel go-to-market strategy, channel analytics, alliance and partner programs, and sales force design. John works with clients across a variety of B-to-B industry sectors with a particular focus on the technology industry.

Contact ZS