It's become increasingly clear that Big Data, and the tools for manipulating, visualizing and analyzing it, are transforming the business landscape. McKinsey released a report in 2011 that projects 40 percent growth in global data generated per year. This is all well and good, but more and more companies are finding that their toolbox for dealing with all of this data is antiquated and confusing.

Indeed, 58 percent of enterprise decision makers surveyed in March 2012 by DataXu felt they lacked the skills and technology required for marketing analytics. Marketers should be chomping at the bit to fruitfully employ the data they have. Successful marketing requires proper segmentation of the customer base to create more targeted campaigns. Real-time insight into the performance of existing campaigns and a clear grasp of where to redirect efforts can also turn a campaign that would have failed into a success. These are the promises made by the drivers of the current "data movement." The unfortunate reality, however, is that the accumulation of data just adds to the costs of an organization as it struggles to merely store the incoming torrent of data, let alone harness it and allow non-technical individuals to explore and understand it.

Luckily, this isn't the first time that industries have experienced this type of problem. The data movement is just like any other one that starts out as a niche interest to a select few people, eventually growing into a commoditized marketplace that competes on usability and ease of access.

Of all metaphors to pick for this process, the restaurant is an apt one. Cooking is something everyone can do. Mix up some batter, put it on a hot skillet, and you'll get pancakes. Add some eggs and a glass of orange juice and you've either got your brain on drugs or a complete breakfast. You can also go to your local IHOP and order the same thing. If you make it yourself, you know everything that's in it and can control the various aspects of the meal. But you also have to deal with acquiring the ingredients, having the facilities to cook, and doing the cleanup. If you go to a restaurant, all you have to do is show up, tell them what you want, and pay.

Similarly, the analytics space has two types of offerings. You can choose to do it yourself or you can use a hosted service to take care of things for you. As with cooking versus going to a restaurant, there are costs and benefits associated with both, but my biased opinion is that a hosted solution is the best choice for tackling the current influx of data.

Economies of Scale
Restaurants provide the benefits of economies of scale to their patrons, allowing customers to consume and enjoy foods that they normally wouldn't be able to at home. High-quality tuna is rather expensive and generally comes in quantities that no individual person could ever consume before it goes bad. Yet, you can go to a sushi restaurant and get various parts of the fish. This is economies of scale in action. The restaurant can afford to put down a significant sum of money to acquire the whole tuna and resell it in pieces to its patrons.

Hosted analytics presents a similar case. A hosted analytics provider is able to pay more money upfront for hardware than any one of its customers would. The reality of data processing is that there are physical limitations to the amount of data a computer can process given a certain amount of time. This problem can only be overcome with more and better hardware.

Because it serves multiple users, a hosted system is actually incentivized to provision enough machines to answer questions quickly. The compute resources are only required for the duration of a query against the system. The faster a query gets answered, the quicker those resources are freed up to answer someone else's query. Responding to queries fast enough to free up resources for the next query is actually the only way to achieve high levels of concurrency. Because the hosted provider is building their business on the idea that multiple customers will share the same infrastructure, they have to support more than just one query at a time and thus are naturally forced to provide their users with a faster querying experience. Economies of scale work to the users' advantage.

Integration of Diverse Data Streams
Another benefit of hosted analytics systems is that they can provide overnight integration with other data sets, both public and private. Taking this back to the restaurant analogy, restaurants add new items to their menu on a regular basis. If they find a supplier that will give them Alaskan king crab for the same price as a lesser form of crab, patrons will all of a sudden start eating better crab without having even known it was coming. The hosted analytics case is similar in that users can take advantage of new data sets that the provider has integrated.

Consider the following scenario. A marketer might normally have access to customer profile and engagement information through their analytics system. Companies like Amazon Web Services offer up data sets from the human genome, the U.S. Census Bureau, and Wikipedia. If a hosted analytics company integrates a public data set like one of these, they can then expose it to all of their clients. This means that if there are 1,000 customers of the hosted offering and only one of them asks for the integration of the public data set, 999 customers get that same integration overnight. All of the participants reap the benefits of having more data sets available. Through the process of overlaying various data streams, marketers can learn more about their customers and their behavior in order to better target their campaigns. This is just one more benefit hosted offerings provide to ensure that companies can maximally leverage the value of their data.

Useful Analytics
Analytics are only good if they are understandable and actionable, just as restaurants are only good if their food is edible and delicious. There are thousands of ingredients that could be mixed in with fried eggs, but some will taste delicious and some will just result in an inedible concoction. As patrons of many restaurants, we often come to a consensus on what various restaurants do well, personal taste notwithstanding. This knowledge can be employed to eat only the best meals. The same mechanism of collective understanding will play itself out in the hosted analytics space.

Any company that provides hosted analytics to a variety of businesses wants to give its customers only the most useful analytical metrics and functionalities. Marketers may not have the specific training to pinpoint exactly which analysis methods to leverage for maximal effect. That's where the multi-tenant properties of hosted analytics work to your benefit. The hosted analytics provider will be sensitive to which of their tools are providing the most value across their entire customer base. In other words, the individual customers all come together to form a collaborative filter to ensure that the less useful analytics features will be cast aside in favor of those that yield valuable insights. As with the integration of public data sets, this filtering mechanism ensures that benefits cascade throughout the entire system of analytics users. Even for features that do not seem to be immediately relevant to your company's success, as a customer of a hosted provider you can rest assured that once your company turns that corner in its business growth, the hosted provider already knows the kinds of analysis you'll find yourself needing and has the tools available. Newcomers to the platform are thus quickly able to reap the benefits of an analytical toolset that has been vetted by the crowd.

In the past few years, Big Data has exploded in importance. Marketers must learn how to take away useful, actionable insights from the mass of data at their hands in order to create a competitive advantage for their companies. Hosted analytics systems will truly prove themselves to be a staple choice for deciphering the increasing amounts of data that companies have to deal with, just as restaurants are a ubiquitous presence in our current lives.

In closing, we can stretch the restaurant metaphor just a little bit more. In both a restaurant and a home kitchen, there's an able cook who knows how to turn raw ingredients into a delicious meal. Similarly, the future still includes analysts who understand the intricacies of your business. You will, however, achieve much more efficient use of your analyst's time by leveraging the benefits of a hosted analytics provider: improved performance, "free" integration of external data sets, and collaborative vetting of the analytical feature set.