Wharton's Lynn Wu discusses her research on how firms can derive ROI from social media.

There’s hardly a company in the world today that doesn’t have a social media presence, whether it’s a Facebook account, a Twitter feed or a website with an interactive blog. But mere existence in the social media space doesn’t guarantee a boost in customers, sales or brand recognition for a business, says Lynn Wu, a Wharton professors of operations, information and decisions who specializes in the study of information technology and its relationship to productivity.

In her latest paper, “Data Analytics Skills and the Corporate Value of Social Media,” Wu analyzed a large sample of businesses to determine how they derived value from social media. The research was co-authored with Wharton operations, information and decisions professor Lorin Hitt and Fujie Jin, a former Wharton Ph.D. student who is now a professor at Indiana University’s Kelley School of Business. The results indicate that firms reap the most benefit when they can analyze and interpret data across all departments and operations, not just marketing. Wu recently shared details of the study with Knowledge at Wharton.

An edited transcript of the conversation follows.

Knowledge at Wharton: Could you give us a brief summary of the research? What were you trying to study?

Lynn Wu: The research is really to understand how firms derive value when they start using social media. That’s an important question because, according to a recent CMO survey, only 15% of executives can quantitatively see the impact of social media on their bottom lines. That means 85% of them cannot [see the quantitative impact]. Forty-five percent [of survey respondents did] not have any clue, both quantitatively or qualitatively, what value social media is to their business. That’s even a bigger problem as social media spending is projected to double within the next three to five years. Understanding how firms derive value from social media is becoming a really important question.

Knowledge at Wharton: How did you examine this question, and what did you find?

Wu: We collected the largest possible sample because previous social media studies often target marketing-related activities or firms. So, here we just look at all of the firms, whoever had a social media presence; in this case, a Facebook company page. Then we mapped all of their stock market value into our model to look at whether their market value increased, decreased or stayed the same when they started using social media.

What we found is that on average it is a positive relationship, not when they adopt but when they actually start using social media. What’s interesting is that the return is extremely heterogeneous, as in some firms derive a lot of value from using social media, some don’t, and some even have a negative impact after they started using social media. That’s really interesting because what is driving that? Why do some firms do so well with social media and some don’t? We’re trying to understand this from a perspective of an entire organization, not just the marketing department. What we found is that it’s really the data analytic skills embodied in their employees that is ultimately driving their return on social media.

Knowledge at Wharton: When you say data analytic skills, can you explain how that plays into social media? What is it that people with these skills are bringing to the table that is helping companies maximize value from social media?

“Understanding how firms derive value from social media is becoming a really important question.”

Wu: The data skill that we measure is really using the employee resumes. We culled [millions of] resumes across the United States, and then we looked at the resumes. Do they have relevant data skills, like data mining skills and data analytic skills? What we found is that these data skills cannot just be concentrated in marketing departments alone. At first, we thought since marketers use advanced analytics, maybe the value came from there. We found exactly the opposite — that it’s really data skill across the entire spectrum of functions within the firm that matters the most. That includes data skills in HR, in finance, in operation, in product development and in R&D. Only when data skills are distributed across all function aspects of the firm can firms derive greater value.

If you think about it, it makes sense because social media data does not mean only for marketers. The data, the consumer preferences, consumer sentiment and feedback can really be applied broadly to every aspect of a firm’s operation. An example would be clothing retailer Zara, which is great at tapping consumer preferences because they track preferences over time. They are actually introducing new products every two weeks.

Some firms are turning their biggest Facebook fans into employees, so that can be great for HR functions as well. You think of the wealth of social media data, and it can be applied to every aspect of firm operations. When a function within the firm — like finance, HR and operations and product development — can make sense of this data and put it into strategies, that’s when you see the greatest value.

Knowledge at Wharton: It seems the differentiating factor here is not just adoption of social media, but knowing how to use it in a smart way.

Wu: Exactly.

Knowledge at Wharton: Were there conclusions that you found surprising?

Wu: Absolutely. Some firms are just naturally fit to use social media, right? If you are a consumer-facing firm, you have products and you need to know what users are doing, then it makes sense. So, we thought most of these will be concentrated in consumer products, in retail or high-tech. But what we found was actually the opposite. It’s really the non-consumer facing firms that generate the most value. That gets us thinking maybe the social media data is really impacting every aspect of the firm’s operation. The marketing outcome may be just the tip of the iceberg. That’s interesting. Also, we thought marketing departments might be benefitting the most when they have strong analytic skills. Again we did not find that. We really find the data skill within non-marketing departments, across all functions of the firms, ultimately drives the value.

“Only when data skills are distributed across all function aspects of the firm can firms derive greater value.”

Knowledge at Wharton: What do you think companies can take from this research in terms of applying it to their operations?

Wu: I think this may not just apply to social media per se, but any kind of data-centric decision making. We have search data, we have click stream data, we have mobile data — all of this data needs some kind of sense-making. Having data analytic skills within the firms is becoming more important than ever. You can’t just hire a raw number of them and put them all in R&D or put them all in marketing. They have to be distributed, de-centralized across all departments. That’s when you can derive the most benefits.

Knowledge at Wharton: What other misperceptions do you think this research dispels? It really goes to the fact that it’s so important to have people to interpret your data, as opposed to just having data.

Wu: Absolutely. I think translating data into strategies in every aspect or operation is the key. It could potentially have an effect on firm hierarchies, for example. Before you are developing a product, you may not necessarily need to ask your boss for permission. You just go to the data directly and say, “Oh, this is where it makes sense to do this.” You just make that decision on your own.

Knowledge at Wharton: It also lends credence to the idea of incorporating a data layer into departments or areas where maybe you wouldn’t normally think that you should.

Wu: Absolutely. People analytics, HR analytics is a big deal. How do you use this data to find your next employees? It’s something maybe most firms have not thought about, what social media can do. Or using the next social media data to generate a next product development or new products for consumers. That may be useful as well.

Knowledge at Wharton: You said that some of the data you found indicated that social media in some cases impacted companies negatively. Was that companies that did not engage or that did not have as much data analytic skills on staff?

Wu: Sure. For example, if you just adopt without using it, that’s a negative effect. But the more negative effects really came from spending a lot of time and energy developing social media platforms and not really having the capability to make sense of data. You make all of this upfront investment into social media, but you don’t really get the maximum return. As a result, you maybe have a negative ROI from social media.

“The more negative effects really came from spending a lot of time and energy developing social media platforms and not really having the capability to make sense of data.”

Knowledge at Wharton: What do you think sets this research apart from other analyses on this topic?

Wu: I think one important factor is that we really look at all firms across varying industries. In manufacturing, in mining. Not just consumer-facing industries. It’s really a very large, comprehensive study on how social media affects firm value. More importantly, we are not just looking at the marketing outcomes or things that maybe all of us know that social media could impact. But other things, that sets of data skill and how it will affect every aspect of the firm, is not something that we or other researchers have really delved into.

Knowledge at Wharton: How will you follow up this research?

Wu: I think data analytic skill is a really big, emerging area. Here, we show the impact on social media. I think it will be really interesting to see how data skills can affect various other parts of operations, such as innovation activities. What kind of innovation does data skill engender? Is it going to be incremental? Is it going to be novel? Is it going to be a brand new innovation that will change your world? That’s going to be really interesting to see.

Knowledge at Wharton: Maybe data analytics will be one of those basic-level skills that everyone is going to be expected to have? It seems that is the direction we are moving in.

Wu: If possible, I think everybody should have some skills. But I think there still has to be some specialized skills involved to really make sense of this data, to make sure that this is really what it is telling you, because if you make the wrong strategy based on the wrong interpretation, that can be very detrimental to a firm. So again, people should have the ability to understand the results, but people who actually generate those results require some special training.