Social Data Lab

Summit 2011.Q3 -- Social Media for B2B

Summary and Purpose

We explore the implications of the social data revolution for a leading enterprise software firm (SAP). Enterprise software companies are always trying to capture more of the long-tail small-business market, however the typically high customer acquisition costs in this space impede growth. We hypothesize the best ways to use social media channels to reach new customers. Our discussions touch on influence, identity, data visualization, robust metrics, and insights from the B2C space. 

The goals of the summit are encapsulated by the following questions:

  • How does a B2B enterprise software firm use social data? What can they learn from C2C and C2W (customer to world) use of social media?
  • What social experiments can be carried out at the major annual user conference? What are hypotheses? And what learnings can be expected (1) for SAP, (2) for their customers, and (3) for third party providers in the ecosystem?
  • How can the effect of different incentive schemes to increase customer engagement be measured, in the short run, in the long run?
  • How can Social Media help us with lead generation of those small and medium sized businesses? And can it help us find key decision makers in those organizations?


Social Data Revolutionary/SDL Director

- Andreas Weigend (

Social Data Lab Members
- Bhaskar Garg
- Suthinand (Pao) Jirakulpattana
- Sebastiaan Boer

SAP Representatives
- Ted Sapountzis
- Alessandra Carvalho Nunes
- Natascha Thomson

Initial Questions

What insights can we get from the social data we are currently collecting?
What novel ways can social data be used for B2B marketing? What can we learn from B2C?
How can we effectively determine topical influence?
How can we visualize and interact with big social data?How can SAP penetrate the small business market effectively using Social Media?


Workshop Overview

Central problem #1: Selling to small-market businesses

SAP needs to become more prominent in the small business market for their key software products. Marketing and selling to small-market businesses have the following difficulties:

  • SAP cannot use the same direct sales approach as with mid-market and large-market businesses, because the revenue does not justify a high lead-generation cost.
  • Because of the need for low marginal cost of acquiring leads, SAP needs to get new leads cost effectively and increase the conversion rate for these leads. 
  • SAP's main lead-generation event is a large conference (Sapphire); small businesses typically cannot attend.
  • How do we get quality SMB leads?
    • Need to engage with the decision makers of small-market businesses: CIOs, CMOs, VP Sales, IT Directors, IT personnel, etc.
    • Notion of top-down influence vs. bottom-up influence: do we target the people using the software, or the managers?
  • How best to employ low-cost online lead generation? Which social media channels to use (Facebook, Twitter, LinkedIn, website, or some combination)? How to increase the conversion rate of leads acquired online?

  Finding and Influencing the Influencer

 Use Social Media to lower the cost of finding leads and increase conversion.

  • Who is the SMB employee in charge of the SAP purchasing decision? 
  • Who is the influencer?

 First, we define influence. Klout chief scientist Ash Rust called in to share their perspective on effective metrics and   methodology. 

Klout: The standard for influence

  • Klout definition: influence is a function of the ability to drive action. It is not just the follower count on Twitter. 
  • We questioned the divide between online and offline influence. Does Klout really extract true real-world influence from social media channels? Thought exercise: who is more influential, President Obama, or Justin Bieber? 
  • Influence = function(number of people you influence, how much you influence them, how influential they are and much more). Influence is also affected by:
    • time (Klout currently only measures influence within 30 days; future algorithms may involve continuously decaying influence with time) 
    • proximity (i.e. how many degrees of separation are there between the influencer and influencee)
    • online versus offline engagement (political figure vs. pop celebrity example)
    • identity (same across various platforms, or some identities more influential than others?)
    • topic (intratopical influence is very different than global influence)
  • Influencers in Social Media: someone who causes change of action as a result of the content they create. 
  • Duncan Watts philosophy ( target the masses or target the influentials?

What data can we use? We track 'direct actions':

  • Tweets
  • Retweets
  • Responses/mentions/replies/conversation (backward & forward)
  • Clicks

How can we effectively find the influencer on a SAP purchasing decision? Currently, Klout only features one combined score for overall online influence. This one-dimensional score is of little importance to SAP. Instead of using Klout to determine influence, a valuable source of information lies in the SAP conferences themselves. By integrating Social Media (i.e. Facebook and Twitter) in to the fabric of the conference, we can create an environment where influencers will pop out from the crowd. The unique benefit of this scheme has many aspects. 

  • SAP will be able to identify influencers in specific areas of interest (i.e. business software). 
  • It is possible to provide a real-time feedback loop during the conference, and increase engagement
  • The impact of each individual video posted on the company’s facebook page can be measured, given the live feedback loop.

Be conscious of

  • "Influencer code of ethics"
  • What is the difference between advertising and spam?

Notion of Identity

To truly personalize customer experiences, we need to collect data that customer shares across as many possible channels as possible. But these are often not integrated in to one customer identity, even though the person is the same.

  • - Basic choice: Real Name vs. Pseudonyms (with Twitter, for example)
  • - Reconciling identity across disparate social networks is one of the most complicated problems
  • - Easiest way is to let the user to help you: they have an incentive!
  • - Users can add all of their networks to one profile → strong sense of identity
  • - They can post their Twitter or Facebook profiles to their blogs

People view their identity differently across cultures:

  • Everyone in Japan has 2 identities - online and offline
  • Western culture often promotes a more integrated identity

The user wants the highest influence score possible by integration, but in the future as many different social networks proliferate, pseudonyms are inevitable. We need to adjust by providing mechanisms to accomodate different ways to interact, across multiple "identities" and levels of anonymity.

Central problem #2: Measuring and Increasing Engagement at Lead-generation Conferences

SAP proposed an infographic that displayed various metrics such as number of tweets and blog posts that the conference generated. This graphic is mostly a reporting mechanism useful after the event. We want people to engage during the event via social media channels, so we need different incentive mechanisms.

Robust Metrics for Social Media and Data Visualization

How can we measure the effectiveness of Twitter use? The absolute number of tweets/Facebook likes/blogposts generated during an event does not capture the utility of a Social Media. Not all tweets are created equally. We also want to emphasize real-time measurement which creates a tight feedback loop that influences behavior AT the conference, not after it.

  • A tweet from the SAP account has different marketing value than a tweet by a conference participant. Therefore, we have to differentiate between SAP's tweets, participant tweets and their respective retweets. 
  • We can create a 2 x 2 matrix to group participants to SAP/Non-SAP, tweet/retweet quadrants. This matrix will visualize the effectiveness of the Social Media campaign.
    • How many tweets and retweets did SAP employees have to push in order to generate Social Media buzz?


Outside World


Empoyees tweeting or retweeting each other

External Community engaging SAP

Outside World

SAP engaging external community

True, autonomous engagement

P2P sharing of SAP-relevant information outside of the SAP community is the most powerful. We can even leverage these metrics to help us find key influencers amongst conference participants, and increase use of Social Media during the conferences. 

  • Set up large screens across the conference that display the matrix described above. Each quadrant of the matrix contains the most active tweeters in that quadrant at that moment in time. This visualization is competitive and fun, and creates icebreakers encouraging participants to mingle and share at the conference.
  • The screens could be topic-based across different zones of the conference

Additional interactions during the conferences can increase Social Media participation. For example: use Bump and/or QR Codes to make information exchange more effective. If you scan my code, you will get my contact info. And I will get yours. These interactions are

  • Symmetric, you scan my code and I scan yours
  • Ice breakers
  • Way more efficient

In general, we recommended:

- Getting more people to participate via gamification, proper incentives, and rapid feedback on their level of participation.

- In a given event, understand who the influencers are and why they are influential, perhaps by performing a post-conference analysis.

- Allow people to tie their comments at the conference to their identity.

- Make engagement topic specific so influence is as well-defined as possible, i.e. booths for each product

- Create incentive systems for SMB decision makers to influence their peers (referral rewards, etc.)

What can we learn from B2C about buying behavior?

Businesses can map out a professional social graph connecting sales representatives, friends, partners, accountants, peers outside the firm, and other “influencers." They can delineate the strength of the social connection as well as the degree of expertise each person has. This locus of decision-making can then be generalized and scaled.

For marketing: segmentation is a step forward, but still not optimal. Resist simple categorization of the SMB market into buckets. Companies are moving toward personalization; a truly individualized marketing pitch and product/service for each customer. i.e. Amazon -> personalized product recommendations.

Other marketing ideas to show case products include celebrity sponsorship, for example, a B2B equivalent to Lady Gaga to promote SAP products to SMB customers. B2C is more algorithmic and data-driven so that more people spending less on software per capita can be supported cost effectively. Enterprise accounts, on the other hand, can justify named dedicated support teams because of revenue scale: a service relationship is essential in B2B vs. a one-time purchase in B2C.

Potential advantages of B2B market over B2C market:

  • Longer-term relationships
    • Reverse existing relationships to advertise products to their network
    • Make it trivially easy to provide reference to to other companies (incentive schemes)
  • Rational decision-making
  • Smaller market, less variation in needs
  • Content has longer tail

Some General Learnings

  • 140 characters of Tweet space does not allow the audience to distinguish real and fake knowledge.
  • Influence is relative: there is no true standard, but we need to reconcile our own intuitive, qualitative sense of it with a quantifiable index
  • Need to proactively bring offline customers online
  • Managing expectations across various communications channels is essential
  • LinkedIn is a great way to reach SMB decision makers: search for certain titles directly, then add them on Twitter
  • Embedding virality into the product itself is not viral marketing - it's better

Detailed Agenda


09:00 Coffee and fruit: Mingle and get ready
09:15 Kick off: Expectations for the day (Ted)
          How to get Marketing and Sales to work together at SAP
          Goal: Sell more stuff to existing clients
09:30 Each participant (to make introduction useful) shares one concrete thing that is difficult and important, where he/she hopes to make progress during the workshop day. (ASW to comment in real time, contribute relevant examples from other companies)

TOPIC 1: Defining and influencing influencers

10:00 ASW
          Social media purposes
10:15 Overview
10:30 Data sources
10:45 Metrics
11:00 Discussion (Gaming)


Purchasing behavior: What can we learn from B2C about buying behavior? 
Robust metrics for social media and data visualization

13:00 Define and communicate break-out topics and form groups
13:15 Lunch / Break-out working session in small subgroups (2 groups(
14:15 Report back to entire group (each group 10 min max)
14:45 Coffee

15:00 SAP presents framework on measurement

15:45 Ash Rust, Director of Ranking at Klout
1630 Coffee


1645 Quiet time where each participant creates concrete project ideas and outcomes
1715 Collect of ideas, benefits and risks, rough prioritization
1730 Summary (ASW)
1745 End

Last modified: July 1st, 2012 at 21:22

Social Data Lab

Powered by GroupSpaces · Terms · Privacy Policy · Cookie Use · Create Your Own Group