Using business data to grow profits in SMBs

Using business data to grow profits in SMBs

Top-notch technology and even the best marketing and sales people can't help you maximize revenue if they're only guessing what consumers do on your website and in other interactions with your company, or assuming that things work the same today as in the past. Yet, well over 50% of small businesses, especially in the manufacturing sector, don’t have strategic data collection efforts, relying on perceptions and subjective situational observations instead.

Strategic data analytic efforts can give you an enormous market advantage – there is a reason why all the large companies do it. All you need is some understanding of how data can be used, and a few tools to collect and organize information.

Here are some factors to consider when thinking of analyzing data on your won and lost sales.

What type of sales do you generate?

When creating a plan to use your sales data, you must take into account the type of sales you generate.

  • You might generate a single sale of a larger, more expensive product or service after multiple contacts with the customer. This is often the case with B2B selling. This type of sale might generate a one-time sale or start a relationship that evolves into a long-term contract with repeat business.
  • You might sell with a high-volume, low-margin strategy, or offer more expensive items, making your profits on margins.
  • Some sales require extensive post-sale support, while others require no further customer contact.

Depending on what type of sales you do, the data you need to collect will be different.

Sales data organization

You probably know what GIGO stands for: Garbage In, Garbage Out

If you want the best information for creating marketing plans, you must first organize the data you collect in the most effective way possible.

This means that before you implement data collection using the metrics your sales and marketing department have asked for, you must first discuss how that data will be used once it's collected.

Your sales and marketing team must think about and request all the pieces of data they need cross-referenced before you set up your data collection and analysis technology. If you wait until after your data is collected to think of new cross-referencing categories, you may find that your data structure doesn’t accommodate them.

Each day your sales and marketing teams delay the collection of data, you experience delays in responding to problems and opportunities. Each customer you lose to a competitor is more likely to develop a preference for that competitor and will be harder to re-capture in the future.

The key to optimizing data usage is to start by asking the right questions, not by simply collecting lots of answers.

How much sales data is being analyzed?

What type of raw data will you need? This requires looking at each data category by itself with no cross-referencing (e.g. number of widgets sold).

Obviously, the next need you'll have is the ability to cross-reference data (e.g. number of red widgets sold to first-time buyers in April).

Can all data be reviewed by a human to reach conclusions, or must it be simplified?

Depending on the data you generate and how it's returned, long reports filled with individual pieces of data will probably be too complicated for a sales or marketing person to understand. This means that simply collecting the requested information and presenting it to sales and marketing managers is not always enough.

How you display your data in reports and the tools you use for that is important. For this reason, sales and marketing should be specific in what they want. Start with determining the data and report specifications. To do analytics, you can’t simply review disconnected numbers on units sold, prices, dates and types of customers. Something like the number of units sold to repeat customers who used a specific coupon code would provide more useable insight.

How many factors can be analyzed against each other?

Your data capabilities should provide you with the ability to analyze all the different pieces of data against each other, or at least those that have been requested by the sales and marketing team.

Looking at our red widgets sold to first-time buyers in April example above, we can see that we might need many single pieces of data to be collected in order to provide us with the true information sales and marketing teams need.

Your data collection technology should be robust enough to collect many pieces of data and be able to simplify them or present them individually.

Data scalability

It's unlikely that your company will get your data analysis needs correct on the first go-round. As you start reviewing and analyzing your data, sales and marketing will likely find that they need more information and specific details. This means your data collection technology must be scalable and able to handle increasing quantities of data collection, processing and delivery.

Perhaps you’d start with sales by product type, then implement reporting on conversions based on margins, and then add sales vs. spend on direct mail brochures for the month, etc.

Data granularity

You should consider the granularity of the data you will need. Do you need a customer's name to appear in two fields (first and last) or will one suffice? Do you need your widget sales data displayed by model number, price, color and margins, or will your sales team be satisfied knowing just the total sales of widgets, knowing that they are $30 each and that the color isn’t an important factor for shopping decisions?

Sometimes you will need to analyze raw data, while other times you'll want data that's already simplified. In some cases, raw data is easy to digest and provides valuable information with no cross-referencing. In other cases, you will want to automate cross-referencing logic to provide data sets that give you the answers you need based on more than one piece of information.

Setting the right goals

Your use of data should focus on both proactive and reactive marketing. With proactive marketing, you set sales and revenue goals and use data to enhance their achievement. With reactive marketing, you look for trends and opportunities you did not know existed to drive new sales.

For example, a proactive marketing strategy might set a goal to upsell a second product to a repeat consumer shopper. You would look for data that could help you do that, such as determining if the buyer is sorting search results by reviews or by price.

An example of a reactive marketing strategy might be to capitalize on the fact that your data reveals that B2B buyers who purchase blue widgets usually opt for overnight shipping (more so than customers who buy red or green widgets). Your marketing department might use this information to set a goal of increasing sales of blue widgets by offering free overnight shipping when a buyer purchases more than one blue widget.

Won vs. lost sales analysis

It's not enough to know why you gained sales, it's also important to know why you lost sales. This requires you to analyze the data of users who did not purchase to see where you might have lost them in the sale process.

You can compare their activity to their counterparts who purchased to see what might have triggered an exit (e.g., price, shipping time, delivered keyword results). You can analyze the keywords they used in product searches and/or their sort preferences. You might find that you are selling more products to people who search by reviews and losing more sales among shoppers who sort by price low to high.

Other sales insights that can be obtained without complicated analysis tools include:

  • Profit margins plot showing lost/won quotes to determine optimal price
  • Won/lost by product type or client type
  • Won/lost by sales person
  • Won/lost by speed of response

Data can also be used to find cause-and-effect relationships, as well as correlations. For example, if you reduce the shipping price on a product on random days, and sales increase each of those days, you can see a cause and effect.

Sales teams can be confused by correlations, which are relationships between two events, but one does not necessarily cause the other. For example, if you offer free shipping on the weekends and see a rise in sales, it might be that people simply shop more on the weekends.

To determine whether your data is showing a cause or correlation, you'll need to run more models to see if you can replicate the conditions and results over longer time periods (usually it is done via A/B testing). When you embark on these efforts, you will need to ensure that your ROI can be worthwhile and that your data can be statistically significant – A/B testing done by small companies is often a half-hearted effort with conclusions based on chance data.

The three types of analytics

Your data should help you achieve three different types of analytics:

Descriptive analytics

Descriptive analytics help you determine what happened in the past, including the immediate, recent past (e.g., this week's sales).

Predictive analytics

Predictive analytics help you get an insight into what might happen based on different data scenarios you run using your past visitor data. Note that it is unlikely that you’ll be able to predict exactly what will happen in the future.

Prescriptive analytics

Prescriptive analytics provides suggestions for future sales strategies and tactics based on the results of descriptive and predictive analytics. Prescriptive analytics suggest possible outcomes if you follow one or more tactics.

For example: your predictive analysis shows a likely increase in market demand for green widgets in March, while the past data also suggests that serving a pop-up promoting a coupon or free shipping increases sale conversions; thus, it may be a good idea to setup a pop-up coupon promoting green widgets in March with the goal of capitalizing on the likely upcoming market trend.

This type of data analysis can help not only with sales, but also with inventory management and forecasting.

Testing sales methods

Smart marketers have been testing their sales methods since long before the internet and ecommerce entered the scene.

Companies used different prices at different locations to see their effect on sales. They sold the same product under different names using different packaging to test the results. Direct mailers ran small samples of different brochures to see which one provided the best results before rolling out the entire mailing.

Broadcasters tested different TV and radio stations before committing their budgets.

Online marketers today do the same, but now have real-time data that allows them to react much faster, and on a smaller budget.

In a world where more and more companies are finding ways to map out their vast business data and gain insights from it, business analytics is not only about ROI, it might even determine who stays around in the years to come.

Other information to consider when evaluating your technologies and data needs

— by Andrey Kolesnikov, business development manager at Steersman

Request a demo

Required
Required
Must be a valid email address
Must be a valid US number

Help us get to know you

Required
Required