What is data science, anyways? There are a lot of correct answers to this question, but here’s a few to get you started:
The chatter surrounding data science usually focuses on enterprise-level initiatives and enterprise-level amounts of data. But even without petabytes of data, or a data science department, small businesses can still get a lot out of data science.
Even if you don’t have the money or degrees to do enterprise-level data science, you can still achieve the same results with every capitalist’s best friend: good old-fashioned division of labor. Third-party companies who find and crunch your data are an option. Sometimes called “data brokers,” these companies can help make your business decisions more intelligent.
Brad Plothow is Head of Communications at Womply, a company that helps SMBs use big data. Womply has already helped tens of thousands of SMB clients find value in their information. For example, they’ve helped one restaurant chain wrangle a particularly potent type of data: customer reviews.
Customer reviews on sites like Yelp, Facebook, and Google represent a large chunk of data about a business. That data’s not only uncontrollable, it’s also potentially destructive: 86% of consumers say they’re swayed by negative reviews. Customer reviews are also potentially helpful. Favorable reviews, or good responses to negative reviews, can make a paying customer of a potential one.
Womply helps the restaurant chain collect all the reviews for their six branded restaurants into one manageable system. Womply created a central platform where all managers have to do “is log in, get a notification when there’s a review from Yelp, Facebook, or other sites with business ratings, review it and respond or manage reviews directly from the platform,” Plothow explained. “Instead of different logins for each place, you’ve got a central place to answer customer reviews.” Womply’s automated data science removes the logistical headache and lets the restaurant focus on the delicate business of customer relations.
For another client, Womply turns their daily data into actionable information. A California auto shop was interested in their profit and loss but needed a closer look at revenue. Their P&L data was in back office software that was great for record keeping, but not for easy consumption. Womply packaged that data in a dashboard that let them check on daily revenue. As a result, the auto shop can now easily check revenue trends and compare it against other factors like weather, mechanic attendance, and payment history. Data science turns dollar figures into understanding.
At $39.99 a month, Womply’s services are affordable to most small businesses.
Traditional data sources like revenue and P&L aren’t the only focus of data science. Even a customer’s movements around a store can become useful information when viewed through the data science lens.
RoomSignal does just that. By installing smart sensors around a retailer’s shop, they can analyze the path a customer takes. Analyzing that physical journey helps retailers improve the journey a customer takes with the brand, as well.
“Our approach shows where customers go, and how many visitors convert into customers,” said Michael Lewis, CEO of RoomSignal. RoomSignal installs discrete sensors that let owners “see where customers walked, and what they bought from those areas, so they see how customers make buying decisions.” As a result, a regular store becomes a smart store.
This location intelligence aids in analyzing customer behavior. For instance, RoomSignal’s data analysis can help SMB owners “ draw illustrations of how people behave in the store and lets them see how it might change if they move popular items to front.” RoomSignal’s approach to data science makes even foot traffic into a lucrative resource.
Smaller retailers (one to two locations) can get started with RoomSignal for $250-300 a month.
Another data science option for SMBs is Kaggle. If you need answers but aren’t looking to invest in full-time data scientists, Kaggle’s unique approach is worth checking out. How good is their data science? Google bought Kaggle in March, 2017.
Kaggle is a site that crowdsources data science. Companies, universities, and interested parties post their data needs, called “competitions,” and challenge teams to solve them (say, predict whether the used cars you buy at auction are worth it). Data scientists compete to see who can solve those problems. The winning team gets the prize (usually cash), and the company gets their insights.
The prices for Kaggle do run higher than the previously listed options. The prices on the current competitions page run the gamut from $25,000 to one million dollars, though some competitions have had prizes as low as $150. There are numerous possibilities for small businesses. One past competition analyzed customer searches to predict what products they were likely to buy. Another analyzed shopper behavior to find likely return customers. One insurance company even used Kaggle to determine a customer’s likelihood of fires. Some contests pay out in ways other than cash (one company is paying for a competition with jobs).
There’s also a range of other data brokers available. These data brokers find a dizzying array of facts about potential customers, from their age and marital status to the type of food they buy, or even whether they have a “Biker/Hell’s Angels” lifestyle. No one can fault the brokers for being less than thorough.
USA Data is one broker that works with small businesses. They offer email and social media marketing services, as well as new leads for business campaigns. USA Data also offers extra information that can expand your understanding of your current customers by adding extra demographic knowledge. For instance, USA Data offers a service that will find the email addresses of your current customers. Alternately, if all you have are email addresses, USA Data can find further demographic information about those people.
If you’re still curious about data science, feel free to ask me about specifics in the comments section! Or, if you’re an SMB who’s used a data science company, tell me (and other readers) about it in the comments.