Data: the Probiotic for your Gut Instinct
If you read through blogs and social posts about data you inevitably hear about the benefits of being data driven:
- Save time and money
- Increase revenue
- Sell more with the same resources or less
- Grow your business with the same resources or less
The message is clear, if you are not driving with data your are wasting money and missing out on opportunities to grow your business.
As true as these things are, the human side of the equation is often downplayed and neglected.
A good data strategy doesn’t eliminate your gut instinct, it enhances it.
Data analytics is about answering YOUR questions with the data YOU collected so YOU can see clearer how YOU can build YOUR company.
Starting point – Your Questions
If you are going to have a successful data strategy you first have to start with meditation. I’m not talking about yoga, I’m talking about thinking deeply about your business and the questions you are trying to answer.
Data analytics can seem like an IT issue, but at the core it is about answering key business questions and testing business ideas. As a business leader you have to think through what the important questions are that you are attempting to answer. This will influence the way IT will handle the flow of your data.
If you just turn the project over to IT (internal or external) with a vague requirement your project will likely end in disappointment. You and your leadership team should spend some time thinking about what questions you are trying to answer with your data and if they get at the core problems you want to solve.
Once you have your plan in place it is time to get to work getting the right data.
This can be an overwhelming task. There are hordes of data everywhere:
- IoT devices
- Legacy apps
- Web forms
- SaaS products
- Social channels
- Ad campaigns
These are a small handful of options, and most of use multiple versions of each data source. It is not uncommon to draw data from a couple dozen platforms.
Ideally you want to draw all of these data points together into a data warehouse (more or less a big database) with automation. As you are building these data pipelines you want to take a good look at the data coming in and make sure that it is structured well both in terms of data standardization and normalization.
Data standardization refers to making sure that all data points in a particular set are in the same format.
For instance RapidBI may show up in a database as RapidBI, rapidbi, Rapid BI, Rapid BI LLC, or many other iterations. As humans we know all these are the same, but we have to tell the computer to treat these as the same entity for reporting purposes. For large data sets you will want to use software tools to do this translation work for you as it can be very tedious and repetitive. The stuff computers are great at.
At this point you may also want to look at how you ingest data. Are there are lot of human errors coming through one or two systems? It will serve you well to look into validation rules in your SaaS products, or to build validation rules into the internal tools you are building. Having strong validation up front will reduce the need for standardization later.
Data normalization refers to how data is logically connected in a database.
When you are building your database/warehouse each table should serve one particular purpose and not have repeating data points. If there is repeating data, you want to further parse out your data sets into more tables. For a good entry-level explanation of database normalization, take a look at this article with an example.
If you do this work correctly you will be well on your way to data driven decision making. It pays to work closely with your data engineers here. How they structure the data model will determine how quickly your Business Intelligence tools run and even what you are able to report on.
You must transfer your vision for data analytics to your IT team if you want your project to be successful and if you want your analysis tools to run quickly. If your tools are running slowly, you likely have an issue with your data model, not the tool itself.
Visualization and Analysis
Once you have clean, structured data in your data warehouse you place a Business Intelligence (BI) tool on top for reporting.
This is the fun stuff.
There is no shortage of good options out there for BI tooling:
- Amazon Quicksight
- Metabase (open source)
A few of these, such as PowerBI and Quicksight, even have advanced analytics packages that include Machine Learning and Natural Language Processing tools. Take some time to evaluate each of them and think about which features you actually need as pricing can spiral quickly.
After the BI tool is set up you can start building dashboards and reports to answer the key questions you started with. A lot of these tools are very user friendly and even non-technical people can use them. Take a look at how you can go from Excel to BI here.
Once you have set up your reports, use your gut and the data to make sense of these reports. Your reports are only as good as the data going into them.
For instance, maybe you are selling 4x4 trucks and decide to run an ad campaign. Let’s say it is in February and your area is getting a lot of snow. If your reports tell you that you have a successful campaign, hopefully your gut is telling you that you might want to do a little more testing and research before dumping a lot more money into your ad spend.
This is an example of how your gut can fact check your data.
On the other side of things, your data needs to guide your instinct. If you were going to make a large purchase for your company you would want to look at your P&L at the very least. You would want to analyze how much money you have, your outstanding debt, cash flow, roadmap to ROI and more. Your gut says, “if I invest in X, my business will grow by Y percent,” but you fact check your gut by looking at the data and judging if your plan makes sense.
We can do the same due diligence with every part of our companies.
You have to allow your data to enhance your gut instincts and make your decisions more informed and powerful.
For case studies on how we have helped companies lead with data, check out our projects page.
To get started with RapidBI, watch this video about the RapidBI process and benefits.