Businesses have been collecting data for ages whether they realized it or not. But it’s been associated with boring, monotonous work in the form of hand written ledgers; filing documents, searching through filing cabinets; storing material on floppy disks; numbers for financial purposes; records of clients; all information would be excavated ever now and then in extreme circumstances. It was more of a chore than anything else, but a necessary chore. However, over time and through technological advances data is being viewed a precious, untapped resource. As technology progresses, the costs of new tools to tap into this precious resources of data is becoming more affordable for smaller businesses. This is great for smaller and medium sized businesses because secrets from their data can be brought out into the open. A new, hidden story through their data can now be told. Some examples of data are purchase history of customers; general location of customers who purchase; call center interactions; numbers of online vs in store purchases; and so much more. This article I came across “What 3 Small Businesses Learned From Big Data”, [1], is great because it looks at three small businesses and how they used data science, data analytics and story telling with data to help improve their business. The first story is of a zoo.
Story telling with data for a Zoo
The weather affects zoo attendance greatly, and thus staffing of the zoo. Rainy days, mean lower attendance which means not as much staff required. It would be ideal to be able to predict attendance based on the weather then in turn this would give a better idea of how many staff are required. That’s exactly what this zoo in Tacoma, Washington did. Prediction of any kind requires historical data. This zoo looked at the historical attendance records along with detailed local climate data, which can be found everywhere as there are weather services collecting this data all the time, hourly in some cases. Through looking at this data the zoo was able to anticipate, not necessarily predict, how many visitors would be coming to the zoo on a particular weekend. That coupled with knowledge of how many staff are required per number of visitors allowed the zoo to now be able to better determine how many people to staff on a given weekend; where increased staffing should and should not be placed. The big test for the zoo was on a long weekend, and in the US it was Memorial Day weekend. It was a rainy weekend and the temperature was low but with this new information from historical data the zoo was able to predict the attendance to within a couple of hundred visitors out of thousands for that weekend. This was great for the zoo because they weren’t short staffed nor were they over staffed. The result is not only a happier experience for all but also costs were more efficiently used and closer to being maximized which in turn means increased profits. In the years to come, the prediction became more accurate for the zoo. Another data science application that paid off for the zoo was when they considered people who purchased tickets online. From looking at the data of online customers they learnt that these online customers were buying tickets at particular times of the day, in the late evening or early morning. This makes sense once we see it through the data. This is when kids are in bed, parents have some time to themselves and can start planning family time and weekends. With this new found knowledge from their data the zoo started offering time limited sales and deals to increase ticket sales at these times. As a result, online ticket sales rose considerably for them. I think the article quotes that “online ticket sales have risen 771 percent over the past two years” [1]. The zoo also started targeted campaigns by region based on where the most frequent visitors lived. Again, this could only have been done by exploring the data and seeing what story it tells.
Summary
Entertainment or attraction type business – Zoo:
- Optimal staffing based on temperature and attendance;
- Noticed most frequent times of online sales so optimizing sales at these times through sales and deals;
- Determined frequent visitors and their location thus offering location specific sales and deals.
Realtor using story telling with data
Another example in this article was of a small town realtor company. The story goes as follows, in the small town a family run local business manages about 1000 homes on the islands of Kitty Hawk, North Carolina. The homes managed are cottages to mansions. The challenge in this business is to satisfy guests with a rental spot for the dates they are looking for at an affordable price and making sure the homeowners rent their properties out as profitably as possible. It seems this small town realtor kept all kinds of data. However, “reading”, “looking into” and “interpreting” the data was the obstacle. Trying to determine what story the data has to tell in order to make better decisions was the challenge. Through the use of analytics the company was able to take their data and format it in a way that they could now understand a story from the data. In the past the best the company could do was tell property owners when their property was available, so empty and no one renting it. Now with the new stories from the data they were able to offer pricing recommendations for particular time periods based on all kinds of criteria including market conditions, seasonal trends, size and location of the home. An example cited in the article is that of a Fourth of July weekend. Through the data, they noticed that there is a decrease in demand for rental after this long weekend. Consequently, they the realtor starting adjusting prices for this time period earlier in the year in January. This increased bookings because now guests planning out their summer holidays as cost effectively as possible could book well in advance making everyone happy, home owner, small town realtor, and visitor. Because this small town realtor was able to offer better service to its existing clients, they were getting more referral business. Another way this small realtor used their data was to start looking at contractor’s maintenance costs to try and find the best contractor, eliminate invoice errors and automate servicing. This helped minimize costs on the maintenance front of the business.
Summary:
Service type business providing the service of renting out property for home owners – Realtor:
- Used data to optimize rental prices at certain times of the year to maximize profits for rental property.
- Minimize maintenance costs by using data to find and then eliminate invoice errors and automate servicings.
Car dealer in Phoenix using data to tell his story
The third business mentioned in the article was a used car dealer in Phoenix. Who wants to buy a used car online? Some people do and this company, Carvana, was providing this service. This online car dearler Carvana did something interesting. They went to Kaggle. Kaggle offers competitions online to people who are interested in solving problems through data methods. Carvana submitted a problem to Kaggle with a prize for the winner. Whoever was able to solve the problem, won the prize. The problem Carvana proposed was, “a better way to predict if cars purchased at auction would be lemons-‘kicks’”, [1]. The result of this competition was a means of allowing Carvana a better way to bid on cars at auctions; bid less and purchase good cars for less; avoid the ‘kicks’. This saved Carvana money, a savings they were able to pass onto their customers. Another way Carvana used their data and the story the data told, was to minimize the risk on the financing side of their business. Looking a potential buyer’s credit score is the usual way of determining the risk of a particular customer. Carvana considered many more variables and across many databases to help predict the likelihood of default payments; weed out the potentially buy customers; offer better tailored interest rates for each customer. The result for them was “…`meaningfully’ fewer defaults and not one car stolen through fraud.” [1]
Summary:
Online sales of large priced items business – online car sales:
- Platforms like Kaggle can be used by small businesses to find a way to use data to answer a particular question or solve a particular problem they may be facing. In this case it was to be able to avoid purchasing the lemons at used car auctions.
- Ability to consider a greater number of variables to determine the risk of a potential buyer.
- Once the risk of a potential buyer is assess, offer the buyer interest rates based on their assessed risk.
References:
- What 3 Small Businesses Learned From Big Data, Kelleher, K., 2014, Inc. Magazine, Iss. July-August, URL: https://www.inc.com/magazine/201407/kevin-kelleher/how-small-businesses-can-mine-big-data.html
- com, https://www.kaggle.com/
- Negative Google Reviews, why it’s important to respond - December 10, 2024
- How to design Website Architecture or Website Structure? - July 26, 2024
- 2024 Marketing Trends to stay Ahead of - July 25, 2024
