Archive for October 2013

Cost of Social Analytics Products Limits Market Reach


Social media analytics software, software that analyzes social media data for useful information about a company, product, or brand, is reaching a rather limited market right now. There are two major types of customer for these products; large multi-national multi-brand companies and the advertising and marketing agencies that serve them. Think of this as the corporate one percenters. As marketing organizations strive to become more data driven and analytical, one of the better tools for ascertaining how consumers feel about a company and who is influencing their buying decisions is designed and priced for a small fraction of companies.

Pricing generally falls into two categories – expensive and really expensive. The pricing bands center around two midpoints of US$25,000 per year and US$100,000 per year. For enterprise software that touches much of a company such as a CRM or ERP, that’s not outrageously expensive… at least for a large enterprise. For a small to medium enterprise (SME), on the other hand, that is a boatload of money. SME companies sweat spending $400 per user on Microsoft Office even though the products in the suite are critical for knowledge workers. They cringe at spending $5,000 a year on software maintenance agreements that are essential to their survival as a company. Spending $25,000, $50,000, or $100,000 per year for software to scan social media for customer attitudes and influences would seem insanely expensive to them. Vendors can talk value all they want but raw dollars is important when you don’t have unlimited funds.

Unlike a lot of other enterprise software, social analytics doesn’t touch everyone in the company, especially at the moment. The argument can be made that there are a lot of uses of social media analytics aside from marketing – service, supply chain management, etc. That’s not where the market is right now. Instead, social media analytics is mostly used for brand management and market intelligence. For software that is mostly used by a few sales or marketing folks this seems like an extraordinary amount of money. While brand and market intelligence drive revenue, for the SME market more direct lead generation efforts are much more important.

Simply put, most social media analytics software is too expensive for the SME market and even many large enterprises. There are some lower cost options but they are much more limited. Often, the number of social media accounts or the types of analysis are severely restricted, or customers get analysis on data that is not close to real-time. This sends one of two messages to potential customers. First, social media analysis only matters to giant brands. Or, social media analysis matters to everyone but we (the vendor) don’t care since we can squeeze better margins form the biggest of the big. Either way, when that top segment of the market becomes saturated and vendors start looking to everyone else for growth, they may not find a particularly receptive audience.

There is a glaring hole in the SME market for social media analytics. Now, if someone can only fill that with a sub-$5000 per year offering there will be many happy customers to buy it.

Asserting Control over the Cloud

Cloud applications and infrastructure are becoming ubiquitous. There are cloud applications (SaaS), cloud software stacks (PaaS), and cloud infrastructure (IaaS) options including storage and compute resources. The advantages of cloud computing are well known including easy of deployment, turning CAPEX into OPEX, flexible and elastic resources, convenience, and the ability to focus on business value over technical problem solving. The disadvantages are also widely known. Usually, IT (and analysts) will talk about problems such as security, availability, or business continuity. These are all important but are basically proxies for the “big problem” – lack of control.

Control is an essential component of responsibility. If someone is responsible for something, whether it’s to insure privacy, smooth operations, protection of company digital assets, or driving business value, they need to be able to control their environment. In many cultures, especially western culture, it is considered inherently unfair to make someone accountable for something controlled by someone else. This sets up the single biggest objection IT has to cloud computing; they are still responsible for what happens in the IT environment even when there is a third party cloud provider actually delivering the infrastructure, stack, or application.

Everyone has to give up control sometimes and usually it doesn’t end in disaster. That’s because many IT issues are a matter of annoyance or inconvenience. It’s the difference between not having access to information right now and having it go away altogether. Asserting control over minor issues with no real risk is petty and counterproductive. Lack of control becomes a big problem, however,
when mission critical applications, the kind of applications that help run the company, are involved. When the business requires high levels of security such as exists for some information in financial services, legal services, and healthcare, then cloud problems become big problems.

There are three typical ways to assert control over the cloud computing environment. First, is to harden the cloud environment or, more accurately, get the cloud vendor to do so. This gives the illusion that all is correct. The sense of control is derived from having done the due diligence. The second strategy is to avoid cloud vendors entirely. That’s also not a great option. To forego the benefits of cloud technology, instead of figuring out how to leverage it responsibly, is not using technology to the best advantage.

Many companies will elect to have some applications on-premises and some in the cloud. Splitting up applications like this solves many of the big problems but creates a number of other ones. For example, critical information is now parceled out between the cloud and on-premises application. In addition, a lot of information is trapped on-premises that doesn’t need to be; often not all data in the same application needs to be that secure even while some does. Companies then end up with information bubbles – batches of information needlessly rendered inaccessible to mobile devices because other information needs to be managed that way.

So how does a company assert control while getting maximum benefits from the cloud? By deploying hybrid solutions. A hybrid solution has a cloud component and an on-premises solution. Hybrid solutions are not one application in the cloud and another on-premises. It’s the same application where the data is partitioned between the on-premises and cloud instances. Common information that has no serious security aspect (such as state tables) is synchronized or shared between the two. Only the information that that must be kept in the physical control of IT is kept on-site. To the average end-user, the application looks and behaves the same even on mobile device except that some data is not. This doesn’t have to be an end-user application either. It could be a virtual file share or even a database.

This approach strikes a balance between keeping certain types of information under tight control and the accessibility and other advantages of cloud services. It solves the big problem without creating too many little ones. There is a catch though. Not all that many applications, stacks, or infrastructure can be made into hybrids. Either it exists in one place but not the other – cloud or on-premises – or the instances of each are incapable of synchronizing information between them. The few there are, however, point the way to the future where on-premises and cloud applications will only differ by who manages them.