By Dayna Sessa
If analyzing numbers makes you anxious, relax. A majority of the value you gain from data comes from things you already do as a savvy business person.
I waited eagerly for my daughter to turn six months old and sit independently so my husband and I could begin bringing her to a neighborhood play area near our regular coffee shop. Witnessing her play independently alongside children her own age seemed like a milestone (we're first-time parents and geeky engineers). Finally, with upright baby in tow, we began stopping by on weekends in search of other babies able to play with rattles and balls but not yet ready for the firehouse and train sets. We'd dip the nose of her stroller in and out, randomly trying to catch what seemed like an increasingly elusive clique of babies with an entirely different coffee run.
You're probably wondering what, exactly, a search for playmates has to do with data. A great deal, it turns out.
Before long, I began asking the staff and owner of the play area the same kinds of elementary trend and cluster analysis questions my clients often ask me. Put simply, how can I find the other babies? How many babies her age come here? What times and days do six- to10-month-old babies most frequently play? I quickly realized that the owner could not answer these questions and felt too overwhelmed and ill-equipped to consider them. Yet, if answered properly, these questions could have brought new customers and more revenue. We and other parents represented a missed business opportunity.
Let's use this case as a lesson in the six baby steps to data competence. Data at any scale - even the "small data" of a new company in its infancy - can have significant commercial and marketing implications. If contemplating numbers intimidates you, note that something similar to the Pareto Principle, or the '80/20 rule' applies because unexpectedly only 20% of data analytics is actually the strict number crunching. The other 80% is conceptual and perhaps more approachable: figuring out what you'd like to know and how to approach the problem; finding, properly storing and cleaning up the data; implementing change management and training staff.
Nor should a small business get intimidated by buzzwords or technical jargon. Hadoop Clusters and SQL Servers are simply tools for capturing and manipulating very large data sets, which you can grow into as your business expands. That step will be far easier when you put great data practices in place early on. And you'll benefit from keeping your finger on the pulse of the data in your company all along, rather than waiting until you need enterprise tools that may require a data scientist to use. Let's get there, growth is a good problem to have.
1. COLLECT: Whether you are starting up or established, a brick-and-mortar store or e-commerce website, selling products or services, there is always data to collect about customers, sales, leads, suppliers and more. Begin by considering what you'd like to know, the commercial implications of that information for your company and how easily you could acquire it. Often you'll find that you're already collecting some data on social media platforms, through Google analytics or at the point of sale. Sometimes you need to collect data from scratch, which isn't necessarily hard.
To determine what you might want to collect, consider:
1. What questions do customers often ask? What services do they seek that you do not yet offer?
2. How effective are your marketing campaigns? Do you have a way of tracking and attributing sales to individual campaigns (often referred to as "marketing attribution")?
3. Does anecdotal observation show any positive or negative trends? Is there any way to prove or disprove your hypothesis (or gut instinct)?
Once you begin collecting data, you need methods to pull it all together to make it usable and actionable for your needs.
2. COMPILE: Putting aside the hype, you could say that "big data" exists when you have so many pieces of information to store that you outgrow your old tools. I would posit that nearly every company with a store of "big data" started with a ledger pad or a computer spreadsheet and perhaps still stores or manipulates some of its "small data" the same way.
For an entrepreneur, Microsoft Excel or Google Docs offer a very low cost or free way to put data in one place and tabulate it. If you've used a ledger or balanced a checkbook, you can use the basic functions of a computer spreadsheet.
3. CRUNCH: This is where data analytics, the science of drawing conclusions from raw data, comes into play. I highly recommend making a minimal investment in a course on deeper spreadsheet manipulation for yourself or an employee. You can find such courses at local libraries and online through organizations like Coursera, General Assembly or Khan Academy, among others. These courses help you learn how to sort and filter and create graphs and pivot tables if you don't already know how to use a tool like Excel.
4. COMMUNICATE (AND OVER COMMUNICATE): Talk to your team about trends they are seeing, what needs improvement and how data can become an integral part of the company's overall success. Educate your staff about the institutional changes you are going to make as a result of your analysis. Teach them methods for collecting data and make them aware of the importance of building a clean and extensive data set.
5. CREATE: Evolve with your company. Look at the data to find new ways to market and build new business and to service customers better. Use data to evaluate success. Make new policies. Perhaps this means asking clients how they heard about you (or having a code for each type of marketing campaign to determine the same systematically) or creating a promotion.
6. CHECK: Have a means of measuring success for your program, perhaps one that translates to the number of new customers (acquisition) or the number of customers who repeat business (retention). See if you've had impact. If you have, keep measuring and keep making sure that every policy stays relevant. If something hasn't worked from the start, figure out what other data you might need to better answer the question. Talk to customers. Talk to your employees. Any form of knowledge is data.
What would this approach have meant in the case of the play area? Let's look at what information we'd need to answer my question about suitable playmates. To know how many children my daughter's age visit the play area, we should collect every child's birth date and timestamp each visit. Then we'd use columns of a spreadsheet to compile that information into a usable format. We'd manipulate some filters and graphs to come to some conclusions and communicate those conclusions to our staff. Finally, we could create promotions and email campaigns related to high-use times for specific age clusters. (Come to think of it, that means perhaps we also want to collect email addresses.) Then we watch and measure whether these initiatives result in new customers, higher revenues, more consistent business or happier parents.
And, most satisfyingly, some of the feedback may be in the form of qualitative data - like smiling parents with coffee cups in hand, saying "Thank goodness we got that email."