data analytics for small business

As a small business, how much investment are you willing to put into data and analytics? If significant returns is time saved in performing your current decision making, would that not be enough to justify effort? From experience in talking to small and medium sized business, the awareness of data and analytics is somewhat low. Fortunately, that is changing today and will continue to improve thanks to the media spotlight on data analytics in general.

Status quo is not an option!

One business was much taken in by the buzz on analytics. They wanted to start off with major changes to their current (inefficient) quote management process. Their business is building custom machinery for testing of manufactured components.  All their products have a basic mechanism combined with a whole set of options for measurement, settings etc. The cheapest product they build costs tens of thousands of dollars and goes up from there. Clearly their product is quite complex and accordingly it requires several weeks to develop a new quote for their customers. They need to explore past work orders, current commodity costs, labor costs, engineering costs, before arriving at a profitable quote. There are obviously several problems with this current state of affairs:

  1. very time consuming
  2. prone to errors
  3. difficult to quantify the likelihood of success of final product – if the quote will turn into an order or not

A challenge they focused on was understanding from past history of successful quotes and using that for developing new quotes. Clearly this was a very complex challenge for a company that is just now starting to look at analytics. Our suggestion to them was to first invest in cataloging all their past quotes, work orders and invoices. In other words, to spend time on managing and organizing their existing data before stepping on to the next level.

The next level for them was using analytics tool such as basic dashboard to understand their sales history. To detect patterns around lead times, success rates, types of customers, options on successful quotes and so on. From that point onwards, marching towards using advanced models for quoting process. Deploying machine learning to improve their ROI then becomes a natural progression.

Where can data analytics help?

We are very sensitive to the analytics maturity level of small businesses. We recommend the following five step process to get the full benefit of investing in analytics.

  1. Understand which decisions you can make better, cheaper or faster using data
  2. Develop processes to consolidate data from different silos
  3. Build appropriate models – descriptive or predictive
  4. Develop an app to deploy the models and monitor the measures that are relevant (based on 1)
  5. Regularly feed new data to the model to maintain it

This process is encapsulated in our affordable analytics model shown in the graphic below. With the availability of open source data science software, the most expensive step in this process is the first step.

Simple but effective solutions

Several of our customers are successfully using this scheme in one or more of the following ways.

Monthly budgeting (predictive): The challenge was to come up with monthly transportation cost forecasts. One customer is using forecasting models to develop rational and data-driven budget estimates. The earlier process was simply guess work. Now they have a rigorous data driven approach to make high quality estimates. 

Production scheduling (predictive): The challenge here was to plan production schedules. Another small business makes multiple product lines. It is essential to have an idea of likely demand for each of the products. Planning for equipment and labor is simply too time consuming, if not. With dashboard to predict demand for products, they are able to come up with solid schedules for their limited resources. 

Manufacturing overhead tracking (descriptive): The challenge for this small manufacturer was to keep an eagle eye on spending and costs. In particular their manufacturing overhead costs, which in some weeks could be as high as 90% of their total costs. The CEO and CFO would spend several hours gathering and cleaning data to bring it into a spreadsheet. Then they would run the overhead calculation. With an easy to use manufacturing overhead tracking app, they have a ready made tool. Which monitors key metrics in real time, is more reliable, and less prone to simple data entry errors. Clearly the executives’ time is better spent on more value added work than on such mundane but critical calculations.

No responses yet

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

simafore.ai