If you work in marketing, sales, finance, or operations, you probably notice data creeping into your day-to-life. And with good reason: businesses that run on data are more successful, reporting 33% higher revenue and 12x revenue growth. Problem is, in order for data to get these big results, it has to get out of the hands of IT and into the hands of people who need it to make decisions every day — that’s you.
The good news is this, you don’t need to be a data analyst to get results from your data and start making data-driven decisions. In this post, we’ll lay out the different types of tools that you have at your disposal, and break down what they do, what they do well, and what they don’t do so well.
Spreadsheets are hands-down the best-known data tool. They’re fast, easy and amazing for certain tasks. The flip-side is that because of their popularity, they’re often misused for things they’re not good at.
Where spreadsheets shine:
- Financial Modeling. Spreadsheets are great for the kind of assumptions and testing needed to put together month-by-month forecasts of financial performance.
- Brainstorming. Spreadsheets are fast and easy: perfect for back-of-the-envelope calculations or preliminary work on a new data set.
- One-Time Analysis. Spreadsheets are great tools for one-off investigations. Grab your source data, analyze, and draw conclusions quickly.
Where they fall short:
- Operational Reporting. Spreadsheets don’t automatically update. This forces decision-makers to use stale information, and wastes your time manually updating and maintaining reports.
- Complex Analysis. Every layer of complexity increases the potential for error.
- Sharing and Collaboration. Spreadsheets offer little insight on what numbers mean or where they came from. Once multiple people are collaborating, the risk of mistakes skyrockets.
Mistakes are incredibly easy to make in spreadsheets, but difficult to spot. It’s estimated that 88% of the world’s spreadsheets contain errors. This is a big deal, because an error in a single formula can flow through to other unsuspecting cells. Proceed with caution!It’s estimated that 88% of the world’s spreadsheets contain errors. http://ow.ly/zQzQm
Structured Language Query (SQL) is the (nearly) universal language used to interact with databases and query large data sets. Because it’s language, not a program, you’ll need a SQL client to be able to write queries. SQL excels at one-time analysis, and its English-like syntax makes it more accessible to non-technical people.
Where SQL shines:
- One-time Analysis. For quick and dirty answers, there’s nothing quite as effective as writing some SQL and exporting the data to a spreadsheet or statistical tool for further analysis.
- Basic Data Sleuthing. SQL is great for basic questions like “How many orders did we have yesterday?” or “How many active customers do we have?”
Where it falls short:
- Operational Reporting. Most effective operational reports require complicated SQL statements that are difficult to write, debug, and maintain. It can be extremely frustrating!
- On Production Servers. Running SQL queries against your production database is an absolute no-no. A long-running SQL statement can cause your website to grind to a halt.
As the complexity of a query increases, SQL becomes harder to write, debug, and maintain. Complexity also makes execution take longer. If you’re running complex queries against large data sets, expect to wait quite a bit for the results.
Most business reports rely on sums, averages, and counts. But if you need to do something more analytically rigorous you’ll need specialized software. A spreadsheet will tell you your sales for last quarter; a statistical tool will help you determine the effects of a branding campaign on your sales.
Where statistical software shines:
- Discovering Causality. If you want to answer questions like “What characteristics predict purchase?” and “What behaviors predict churn?”, you need to use statistical methods and tools. Attempting to answer questions like these without statistical rigor can set your business on the wrong path.
- Deciding Between Competing Options. Form a hypothesis, collect data, and then accept or reject it. This process is commonly applied by digital marketers doing A/B testing, a simple (but incredibly effective!) use of statistics.
- Advanced Market Segmentation. Take what you know about your customers and cluster them into segments so you can develop targeted products.
Where it falls short:
- If You’re a Novice. Incorrectly run tests will lead you astray. If you don’t have the background to confidently run a statistical test in a tool like R, STATA, or SPSS, it’s better to stick to something more familiar.
- Post-Hypothesis. You must create your hypothesis prior to conducting a test. If not, you run the risk of cherry picking results that support your preferred outcome.
- When an Easier Tool Will Suffice. Statistical tools are the hardest of all a business user’s tools to master. If an easier tool will suffice, use it. The point of data analysis isn’t to prove your analytical muscle, it’s to drive business decisions.
For strategic decisions, be rigorous in your analytics. For day-to-day work, getting to a decision is more important than flexing your analytical muscle.
Visualization and Dashboarding
Visualization and dashboarding tools provide a graphical layer on top of your data. They shine when you’re trying to tell a story with data. These types of tools encompass everything from chart building tools and interactive visualizations to plug-and-play dashboards.
Where visualization and dashboarding shine:
- Complex Stories. Some data is inherently complicated, and advanced visualization techniques can help you build a comprehensive story.
- Making an Impact. Most of time you can get by with basic tools like line and bar graphs, but when you need to make a big impact fast, use data visualization tools.
- Showing the World. Dashboards excel at keeping critical metrics in your line of sight at all times, helping every member of your team stay focused on the numbers.
Where they fall short:
- When it’s Unnecessary. If you can just as easily accomplish what you’re looking for with a different tool, stick with that. No need to add unnecessary tools to the mix.
- Number-Crunching. Visualization and dashboarding tools do not perform data analysis—they provide a lens to look at data that you’ve already analyzed. If you need to analyze data, do it in another tool.
Know what you’re paying for. Visualization and dashboarding tools can sometimes look very similar to business intelligence tools, but they’re out to solve very different problems. The key difference is that visualization and dashboarding tools are an overlay of your data (front-end), while BI tools allow you to manipulate your data in a data warehouse (back-end).
Google Analytics is the most well-known analytics tool. But today, just about every piece of business software includes an analytical component. Tools like Silverpop, Hootsuite, and AdRoll all provide analytics related to their product. The basic reports that typically come standard make it easy to answer single-domain decisions like “How many website visitors did I get?” or “What subject line performed best?”
Where analytics tools shine:
- Data Collection. Analytics tools often collect the data that they analyze. If you need to gather the data, don’t do it yourself—get a tool!
- Single-Domain Decisions. Analytics tools are purpose built. If you want to know how to optimize your website or email marketing, an analytics tool is your best bet.
- Answer Common Questions. An analytics tool is built to answer a specific set of commonly-asked questions, like: “How many website visitors did I get and where did they come from?” “What subject line performed best?”
Where they fall short:
- Cross-Domain Decisions. Most business decisions require data from multiple data sources. For example, optimizing your product prices means you need to integrate your web, order, and accounting data and start running experiments. Analytics tools can’t do this.
- Answer Unusual Questions. Analytics tools ultimately they force you into one particular way of looking at the world. If you want to look at revenue excluding sales tax but your analytics tool doesn’t have that option, you’re out of luck.
Analytics tools excel at collecting data related to their domain and showing it via simple, templated reports.The growth in the analytics space has created a lot of data, but has left many business users still looking for answers — companies cannot run on analytics tools alone.
After evaluating the strengths and weaknesses of these 5 tools, we built RJMetrics, a complete analytics platform that allows you to find the answers you need without digging around in spreadsheets, databases, and disparate analytics tools.
Real insights begin when data works together. To this end, RJMetrics extracts raw data from each of your transactional systems and loads it into a data warehouse. We then prepare your data for analysis, and serve it to you with a visualization layer including customizable dashboards. We can take data from any source, and analyze however much of it you have.
Where RJMetrics shines:
- Operational reporting. RJMetrics automatically replicates your data from where it lives, so that it’s always up-to-date. With RJMetrics, you no longer have to waste days updating static reports!
- Data exploration. With RJMetrics, all of your data lives in a single data warehouse. This means you can easily explore any aspect of your data seamlessly. Ask any question about your business and get an answer back immediately.
- Collaboration. With traditional tools, it’s hard to get everyone in your organization on the same page. RJMetrics allows you to share data with your whole team. Publish a finished dashboard, or let others slice and dice the data themselves.